Research

Primary Research Thrusts:

Geothermal resource assessment and reservoir creation:

We advance geothermal energy across EGS and sedimentary resources, focusing on how geomechanical processes control fracture permeability, flow distribution, and heat extraction. In parallel, we integrate temperature mapping, heat-transport modeling, and economic evaluation to support geothermal exploration and development.

ML/AI for drilling and stimulation optimization:

We develop and apply ML/AI tools to improve drilling and stimulation decisions using field data while maintaining physical consistency. Our work targets practical, decision-relevant tasks, including near real-time prediction of fracture and flow behavior, early detection of drilling anomalies and issues, and optimization of stimulation parameters.

Experimental rock deformation:

We develop and use biaxial, triaxial, and true-triaxial loading frames and core-flow systems to quantify rock and fracture deformation, permeability, and injection-driven seismic response under realistic reservoir conditions. These experiments produce mechanistic constraints that can guide stimulation design, injectivity management, and reservoir development.

Induced seismicity: mechanisms and risk mitigation:

We study the physical controls on induced seismicity during fluid injection and how seismic observations can be used to characterize fracture and fault systems. Induced seismicity is inherently coupled: pore-pressure changes, fracture deformation, frictional behavior, and permeability evolution interact and co-evolve during injection. We investigate how these processes govern when and where slip initiates, how seismicity migrates in space and time, and how injection protocols (e.g., rate, pressure, volume) shape the seismic response.

Sensors and sensing for rocks and minerals characterization:

We design and develop sensing methods and prototypes to improve rocks and minerals characterization. We focus on integrating sensing with geomechanical understanding to support practical workflows for mineral identification, fracture characterization, and decision-making during subsurface exploration and development.

Geomechanics for Critical Minerals Extraction:

We investigate the coupled geomechanical and geochemical processes involved in extracting critical minerals from deep reservoirs. Through integrated experimental and numerical approaches, we analyze rock-fluid interactions, mineral dissolution and transport, and rate-dependent deformation to enhance fracture connectivity and mineral recovery efficiency.

Geothermal Potential in South-Central South Dakota

South-central South Dakota hosts significant yet largely untapped geothermal potential, with regional geothermal gradients commonly exceeding ~100°C/km. We compiled the most comprehensive temperature dataset to date, developed a higher-resolution geothermal map, and evaluated the techno-economic potential of both direct-use (hydrothermal) and EGS applications. This work provides a strong first-order assessment to support future development of this unique geothermal resource in the Upper Midwest (Ye et al., 2025; 2026). We also conduct geophysical analyses and heat transport modeling to explore the heat resource mechanisms driving the geothermal anomaly.

Physics-informed Machine Learning for Ahead-of-Bit Fracture Detection

We develope a depth-based, physics-informed machine learning approach to predict fracture intensity and aperture ahead of the bit using near-real-time drilling and logging data. Unlike conventional machine learning methods that rely on random train-test splitting, our depth-based framework trains on shallower intervals and predicts fracture characteristics in deeper, unobserved sections, thereby enabling true ahead-of-bit fracture detection and characterization.

Fluid-Driven Fracture Shearing Using a Custom Biaxial Loading Frame

We design and develope a custom biaxial loading frame for experimental investigation of injection-induced fracture slip and seismicity.

flow-front.mp4

Flow-front evolution during injection and shearing.

Numerical Simulations (DFN-FEM and PFC)

Left: We apply a coupled DFN-FEM approach to analyze cavern stability for underground excavations at the Sanford Underground Research Facility (SURF), with explicit consideration of rock anisotropy. Right: We also use Particle Flow Code (PFC) to simulate fluid-driven fracture propagation, investigating the effects of rock texture and injection scenarios.

Page updated

Report abuse