Bio
Research focuses on developing coupled flow–geomechanics upscaling techniques to predict the strength parameters of the heterogeneous McMurray formation (oil sands). This is an innovative approach that leverages a three-dimensional convolutional neural network (3D CNN) to derive upscaled geomechanical properties directly from geological realizations. By integrating detailed geological heterogeneities into reservoir-scale simulations, this method dramatically reduces the computational cost of coupled fluid-flow and geomechanical analyses while faithfully reproducing stress and strain behavior, volumetric and plastic strains, and pore-pressure evolution under partially drained conditions. This study offers a robust, efficient pathway toward more accurate, high-resolution 3D simulations of oil-sand reservoirs, opening new possibilities for predicting geomechanical responses in complex subsurface environments.
