Rapid Workflow for Assessing Geothermal Potential of Existing Oil & Gas Sites

Assessing geothermal feasibility requires linking subsurface characterization, fracture behavior, reservoir performance, and economic analysis into a unified workflow. In this work performed by Baker Hughes, an integrated modeling approach was applied to evaluate the potential for repurposing late-phase oil and gas sites for geothermal energy production in the southern United States.

The workflow combines geomodeling, Discrete Fracture Network (DFN) simulation, dynamic reservoir modeling, and full-physics Dynamic Fracture Modeling (DFM) using XSite from ITASCA Software to assess potential geothermal performance of fossil fuel sites.

Key Takeaways

• This novel workflow delivers an integrated pathway for geothermal feasibility assessment through physics-based fracture modeling that:
  • Bridges subsurface modeling and economic evaluation.
  • Enables early-stage assessment of potential development scenarios.
  • Supports decision-making for repurposing oil and gas assets to geothermal energy.
• Integrated modeling is essential to capture the coupled physics governing geothermal systems, made possible with XSite.
• Natural fractures significantly enhance sustainable long-term energy production and cost-effectiveness.
• Flow control strategies such as zonal isolation are critical for long-term performance.

Approach

A three-phase workflow was implemented to assess existing fossil fuel sites for geothermal potential.

1. Site Screening and Characterization

Five candidate sites were evaluated using Play Fairway Analysis (PFA) to select two sites for study using the following criteria:

• Low-risk geology
• High heat potential
• Existing power grid connections
• Data availability for modeling
• Market demand

This phase also involved developing site geomodels (structural and stratigraphic, geomechanical, temperature, natural fracture assessment, and geochemistry), as well as the conceptual design of optimal well trajectories and stimulation zones.

Three Class 5 well configurations were used for the analyses. The pre-feasibility DFN and EGS scenarios both used a doublet of lateral wells of approximately 1,500 m in length located 300 m apart at the same depth, along with another DFN scenario with a vertical injector perforated 250 m above a lateral producer. The full-physics XSite-based EGS feasibility case scenarios used lateral injectors at fixed depth with producers placed at 3 different depths above the injector.

2. Pre-Feasibility Modeling

The second phase of the study involved pre-feasibility studies on the two sites that were selected, broken into 5 steps:

1. Generation of Discrete Fracture Network (DFN) models using fracture data.
2. Design of hydraulic stimulation to model hydraulic fracturing of intact rock and/or “hydroshearing” of preexisting natural fractures to create Enhanced Geothermal System (EGS).
3. Upscaling of EGS and DFN models to a gridded 3D permeability field.
4. Forecasting resource potential and production using Reservoir Dynamic Modeling, along with sensitivity scenarios to test model uncertainty.
5. Preliminary economic analysis including the Levelized Cost of Electricity (LCOE).

DFN Modeling

Natural fracture systems were simulated using finite element software to optimize hydrothermal fluid flow by incorporating the geometry and properties of discrete fractures. A full stochastic DFN model was developed, and properties of the DFN were upscaled into the geomodel grid to model flow and transport over time.

EGS Representation

To avoid computationally intensive simulations of fracture propagation, several Enhanced Geothermal System (EGS) scenarios were modeled stochastically to represent perforation clusters along the lateral wells. Cluster spacing was found to be a controlling factor in the residence time of fluid flow through fractures, optimizing heat extraction and minimizing temperature decline. Variations in uniform and non-uniform cluster spacing, fracture half-lengths, and fracture intensity were modeled to mimic the variability that is frequently found in hydraulic fracturing operations.

Reservoir Simulation

Results of the DFN permeability cases modeled for both sites showed that fracture intensity strongly influences performance, with denser fracture networks yielding higher thermal output. Modeling of the various EGS cases revealed that the optimal configuration for both sites was a uniform, dense cluster spacing.

Economic Outcomes

Economic analysis included surface plant conceptual design, drilling and well completion design, risk analysis, a cost estimate of DFN/EGS/AGS options, and LCOE. Findings included:

• DFN (naturally fractured) systems provide the lowest LCOE, but only slightly.
• EGS systems show higher variability due to a large decrease in thermal output over time and stimulation costs in some cases.

3. Feasibility Assessment Using XSite

To capture realistic fracture behavior during stimulation, XSite was used for full-physics Dynamic Fracture Modeling (DFM). XSite simulates fracture initiation and propagation using a lattice-based Synthetic Rock Mass (SRM) approach, enabling:

• Fully coupled thermo-hydro-mechanical modeling.
• Simulation of hydraulic fracturing with or without natural fractures.
• Representation of near-wellbore and field-scale processes.

Performance Assessment

Various well-engineering scenarios were analyzed using XSite, involving a combination of hydraulic fracture propagation and hydroshearing of pre-existing joints, to select the optimal production plan over a period of 30 years.

The models incorporated:

• Multi-stage hydraulic stimulation along lateral wells.
• Interaction between induced fractures and pre-existing fracture networks.
• Injection-driven fracture propagation and fluid flow.

Results of the scenario with hydraulic fracturing without natural fractures showed mostly upward growth of approximately 2–3 relatively isolated fractures per stage that exhibit stress shadowing effects. For maximum effectiveness in this case, wells must be stacked vertically.

The scenario with a naturally fractured reservoir indicates a more complex fracture propagation path and, in some stages, stimulation extends greater distances. Fracture growth is still primarily upward, but downward fracture growth also occurs due to pre-existing natural fractures.

Sensitivity studies for sites with and without natural fractures were performed with different well separation distances, and with and without zonal isolation. The optimal scenario among these studies was characterized by a natural fracture network with well separation of 250 m and zonal isolation, which resulted in 7% thermal decline over 10 years and 25% decline after 30 years of production.

Economic Assessment

Economic analysis involved calculating the Levelized Cost of Electricity (LCOE), including drilling, stimulation, and operational costs for the various cases. Zonal isolation had the strongest effect on LCOE, then well separation distance.

Case without natural fractures (black line at base of fractures represents the injection well). Left: Stage-by-stage hydraulic fractures along the lateral injector. Right: Resulting hydraulic aperture. Horizontal black line represents the injection well.

Case without natural fractures (black line at base of fractures represents the injection well). Left: Stage-by-stage hydraulic fractures along the lateral injector. Right: Resulting hydraulic aperture. Horizontal black line represents the injection well.

Results & Impact

The streamlined, integrated workflow addresses increasing global demand for geothermal energy, supporting efforts toward decarbonization and reducing greenhouse gas emissions.

This Baker Hughes project highlights a new integrated workflow that combines reservoir modeling, fracture network (DFN) simulation, Dynamic Fracture Modeling with XSite, and economic analysis to guide decision making in repurposing fossil fuel sites for geothermal energy production. Study of multiple scenarios of sites with and without natural fracture networks, zonal isolation, and with varying well separation distances found:

• Sites with natural fracture networks are more sustainable, with more consistent energy production over 30 years than scenarios with only hydraulic fractures, and more cost-effective operations.
• Repurposing oil & gas sites for geothermal energy is viable for sites with naturally fractured systems and high temperatures (above 150°C), in terms of both economic and technical performance.