Characterization of Shale Softening by Large Volume-Based Nanoindentation

A novel statistical nanoindentation technique is presented that obtains massive data on the basis of shale sample's large volume (LV) to assess water-induced softening in terms of Young's moduli of both individual minerals at the micro/nano-scale and the bulk rock at the macroscale, with the latter extracted by a newly proposed surround effect model. Distinguished from traditional statistical nanoindentation that only examines the material at shallow depths of up to a few micrometers, this LV-based method obtains successive measurements to much larger depths of up to similar to 300 mu m via sacrificial removal of the previously indented surface layer, enabling assessment of changes in mechanical properties over a large volume. Natural shale sample was first hydrothermally treated to cause softening, followed by statistical indentation with continuous stiffness measurement (CSM) on successive layers of increasing depth upon removal by polishing of the prior tested layers. For each tested surface, cumulative distribution function (CDF)-based deconvolution was performed to analyze multiple subsets of the massive data (i.e., similar to 1000 curves), each of which was extracted from the CSM curves via segmentation at a fixed depth. Such results on different segmentation depths were then fitted by the surround effect model to extract the Young's modulus of individual minerals and bulk rock. The clay matrix is highly sensitive to rock-water interactions and its Young's modulus decreases significantly from 29.2 to 16.5 GPa upon 30 days' treatment. The average rate of softening advancement was estimated to be similar to 10.5 mu m/day, suggesting that softening advancement is intrinsically controlled by permeability, despite various physical and chemical softening mechanisms.