Microstructural controls on elastic anisotropy of finely laminated Mancos Shale

Shales commonly exhibit anisotropy in their elastic wave velocity, which directly impacts the accuracy of seismic imaging and their geomechanical response to drilling and completions. Anisotropy is often caused by mineralogical layering, fractures and rock fabric (i.e. oriented grains and intrinsic anisotropy of clay sediments). However, the relative impact of each of these features on macroscopic shale properties is not well understood. We combined scanning electron microscopy (SEM) and X-ray microcomputed tomography to image the mineralogical and structural heterogeneity of Mancos Shale and converted the acquired computed tomography (CT) and SEM images into heterogeneous 2D elastic models. We used wave propagation numerical simulations to understand the effects that layering and fractures have on elastic wave velocity anisotropy. Consistent V-p/V-s ratios around 1.46 for modelled and measured velocities validates SEM observations of a quartz-dominated shale lithology. CT-derived models containing layering and fractures exhibit 28.6 per cent and 58.8 per cent of the V-p and V-s anisotropy observed in the laboratory, whereas SEM derived models exhibit 74.5 per cent and 73.2 per cent of the anisotropy, respectively. The increased anisotropy of SEM-derived elastic models is a result of the ability of the SEM to discern individual mineral grains and microstructural features, whereas the CT models require the use of an effective medium theory to model variations of lithology. Overall, modelled wave propagation perpendicular to bedding more closely captures the experimental velocities than parallel to bedding. Therefore, subresolution rock fabric anisotropy likely accounts for the relatively larger velocity mismatch in the parallel direction, and is likely responsible for the decreased anisotropy in coarse rock models. Future modelling would require higher resolution images to structurally constrain these features and/or anisotropic elements to account for fabric anisotropy. Despite some limitations, our study provides a reliable procedure to estimate anisotropy of dynamic mechanical properties of laminated shales using SEM and CT imaging combined with numerical simulation of wave propagation.