Dr. Hafssa Tounsi

Biography

Hafssa Tounsi is a geomechanics engineer at Itasca Consulting Group, where she contributes to geomechanics projects in the mining and energy sectors. Her work includes stockpile stability analyses and modeling of enhanced geothermal systems.

She holds a PhD in Geosciences and Geoengineering from MINES ParisTech. Her doctoral research on thermo-hydro-mechanical-chemical (THMC) modeling of artificial ground freezing for underground mining received the Pierre Londe Award for the best PhD thesis in rock mechanics in France.

Following her PhD, Hafssa conducted postdoctoral research at Lawrence Berkeley National Laboratory under the supervision of Dr. Jonny Rutqvist. There, she worked on THMC modeling of nuclear waste disposal, CO₂ sequestration, and subsurface thermal energy storage.

Introduction of the Lecture

What a DFN-Based Hydro-Mechanical Model Reveals About Hydraulic Stimulation at Utah FORGE

Natural fractures can strongly influence hydraulic stimulation in enhanced geothermal systems, yet this influence is difficult to infer from field observations alone. Numerical models that explicitly represent fracture geometry and its coupling to fluid flow and rock deformation are needed to interpret how stimulation propagates through fractured rock.

We present a field-constrained, discrete-fracture-network- (DFN) based coupled hydro-mechanical model of hydraulic stimulation at Utah FORGE (Frontier Observatory for Research in Geothermal Energy), built using the XSite modeling framework. XSite is a discrete element code developed specifically for modeling hydraulic fracturing in naturally fractured reservoirs. It uses the synthetic rock mass (SRM) method and a lattice-based approach to represent rock deformation, new fracture propagation, and shear reactivation of pre-existing fractures.

We back-analyzed Stages 8-10 of the 2024 stimulation campaign at FORGE and compared the simulated results with geophysical and wellbore monitoring data. Reservoir stimulation was controlled by pre-existing natural fractures and dominated by hydro-shearing, rather than simple tensile fracture propagation. Across the three stimulation stages, the model predicted intersections of created and reactivated fractures with the production well at locations far from those expected from a planar-fracture model, consistent with field observations.

These results show how DFN-based hydro-mechanical modeling can translate field signals into a mechanistic interpretation of fracture reactivation and flow-path development during geothermal stimulation, providing a basis for predictive modeling and future reservoir development at FORGE..