With advances in characterizing the reservoir heterogeneities and modeling the complex reservoir-geomechanical behavior using modern geological modeling software and simulators, the impact of geological uncertainties on reservoir-geomechanical responses has drawn increasing attention for an improved engineering design and decision makings in subsurface developments. Robust and efficient upscaling techniques for heterogeneous reservoirs have emerged as the key to enable field-scale reservoir-geomechanical assessments while consider the impact of lithological heterogeneities like weak bedding planes. The proposed upscaling technique is based on local continuum geomechanical simulations with user-defined boundary conditions on each upscaled region, thus, it can reproduce the anisotropic stress-dependent failure criteria caused by sub-grid scale heterogeneities. This upscaling technique is applied in a reservoir-geomechanical simulation of a steam-assisted gravity drainage (SAGD) project and reduces the computational time by 94%.

eomechanical assessment using coupled reservoir-geomechanical simulation is becoming increasingly important for analyzing the potential geomechanical risks in subsurface geological developments. However, a robust and efficient geomechanical upscaling technique for heterogeneous geological reservoirs is lacking to advance the applications of three-dimensional (3D) reservoir-scale geomechanical simulation considering detailed geological heterogeneities. Here, we develop convolutional neural network (CNN) proxies that reproduce the anisotropic nonlinear geomechanical response caused by lithological heterogeneity, and compute upscaled geomechanical properties from CNN proxies.

An integrated coupled THM modeling methodology is proposed here to improve the modeling of reservoir deformations and caprock integrity in a heterogeneous oil sand reservoir with interbedded shale barriers. The pressure and temperature front are found to propagate at different speed and that dominate the elastic and plastic deformations caused by changes of shear and mean effective stress. Therefore, four stages are divided in the SAGD process that can be interpretations of changes in stress paths including buildup of pore pressure, generation and dissipation of thermal induced stresses.

A robust and efficient upscaling technique is essential to model the impact of heterogeneity on the deformation and failure of oil sands and caprock shale. Although conventional analytical and numerical upscaling techniques are available, many of these methods consider oversimplified assumptions and have high computational costs, especially when considering the impact of spatially correlated interbedded shales on the shear strength. A machine learning enhanced upscaling (MLEU) technique that leverages the accuracy of local numerical upscaling and the efficiency of artificial neural network (ANN) is proposed here.

Understanding the strength and failure modes of overburdens and reservoirs is a critical component in safety assessments for oil sands surface mining and in situ thermal recovery operations. The purpose of this work is to propose a numerical characterization workflow that helps predict the failure mode and shear strength of heterogeneous oil sands interbedded with shale beddings during thermal recovery. Heterogeneous models are generated through sequential indicator simulation with a calibrated constitutive model and geomechanical parameters for each lithology. Numerical simulations with boundary conditions reflecting in situ stress changes are conducted to study the impact of shale beddings on stress–strain response, failure modes, and shear strength of the sheared zone.