An anisotropic viscoplasticity model for shale based on layered microstructure homogenization

Jinhyun Choo, Shabnam J. Semnani, and Joshua A. White

International Journal for Numerical and Analytical Methods in Geomechanics

Abstract

Viscoplastic deformation of shale is frequently observed in many subsurface applications. Many studies have suggested that this viscoplastic behavior is anisotropic—specifically, transversely isotropic—and closely linked to the layered composite structure at the microscale. In this work, we develop a two‐scale constitutive model for shale in which anisotropic viscoplastic behavior naturally emerges from semianalytical homogenization of a bilayer microstructure. The microstructure is modeled as a composite of soft layers, representing a ductile matrix formed by clay and organics, and hard layers, corresponding to a brittle matrix composed of stiff minerals. This layered microstructure renders the macroscopic behavior anisotropic, even when the individual layers are modeled with isotropic constitutive laws. Using a common correlation between clay and organic content and magnitude of creep, we apply a viscoplastic modified Cam‐Clay plasticity model to the soft layers, while treating the hard layers as a linear elastic material to minimize the number of calibration parameters. We then describe the implementation of the proposed model in a standard material update subroutine. The model is validated with laboratory creep data on samples from three gas shale formations. We also demonstrate the computational behavior of the proposed model through simulation of time‐dependent borehole closure in a shale formation with different bedding plane directions.

Link to Full Paper

https://doi.org/10.1002/nag.3167

Highlight

highlight image
Simulation of creep closure of a horizontal borehole in a layered shale. Each column corresponds to different bedding plane orientations with respect to horizontal, while each row provides snapshots of the plastic strain evolution at increasing times. The proposed material model is able to capture both the impact of anisotropy and time-dependent creep.

Updated: