Deep learning for geomechanics
Nanoindentation provides a promising approach for characterizing geomechanical properties when large samples, such as core plugs, are unavailable. However, common models for its interpretation are limited to simple (ideal) patterns that are absent in subsurface samples, as shown by more than 10,000 tests conducted in our group. This research focuses on the interpretation of complex fractures.
Machine learning for capillary pressure
In the subsurface, the natural porous medium has a complex topology, making two-phase displacement simulations challenging; as a result, researchers have proposed various models to simulate this by interpreting key features of capillary pressure measurements. However, only a few models map pore-scale processes (~10 nm–1 μm) to centimeter-scale behavior, and most require parameter tuning. More importantly, these models depend on researchers’ interpretation of the measurements. This research aims to interpret capillary pressure using machine learning.
Interpretation of subsurface signals
Subsurface sources (fossil and geothermal) are indispensable for human beings and the subsurface can also serve as a storage region. But we do not usually have direct access to the subsurface so as to accurately characterize it. Our goal in this research is to quantitatively predict how fluids, petrophysical properties, geologic features, and stress are distributed within the subsurface by analyzing the signals. Our earlier investigation focused on predicting permeability enhancement based on acoustic emission (AE) events.
The geomechanical properties of a reservoir are determined from triaxial tests on large samples (core plugs or blocks). It is difficult (time consuming and expensive) to recover large samples, especially from shale formations that are mechanically unstable. As a result, the required information is scarce for a majority of horizontal wells and for some vertical wells. Our goal in this research is to determine the geomechanical properties of shale and carbonate formations from drill cuttings based on two-scale workflows with the potential for real-time analyses.
Transport properties in the shale matrix
One of the distinguishing features of the shale formation is that its characteristic pore size (pore throat or pore body) is on the order of nanometers in the matrix. This means that the pore space is a natural nanofluidic system. Nanofluidics, a discipline which analyzes fluid transport in sub-100-nm conduits, shows that fluids behave differently in nano-size conduits than in larger structures, such as those of micrometer dimensions. Our goal in this research is to determine the transport properties in the shale matrix by accounting not only for the main principles of nanofluidics but also for the effective pore connectivity at the 1-cm scale.
Multiphase flow in shale with complex fractures
Energy recovery from unconventional resources often entails formation stimulation by hydraulic fracturing. Accurately predicting the multiphase-flow behavior in the presence of multiple intersecting curved (complex) fractures is challenging because of the inherent nonlinearities in the physics of flow and in the fracture geometry. Most assume that the induced fractures are planar for simplicity, or re-mesh and couple the existing models. The former, however, leads to unrealistic results, while the latter is computationally expensive and relatively hard to implement. Our goal in this research is to quantitatively predict the multiphase-flow behavior by developing analytical models.