- PNNL, Fervo Energy and NVIDIA are developing an AI-powered digital twin platform for geothermal energy operations
- The Enhanced Geothermal System Twin aims to give operators near real-time visibility into underground reservoir performance
- The EGS-Twin platform will combine geothermal field data with advanced AI and simulation technologies
The Department of Energy’s Pacific Northwest National Laboratory is partnering with Fervo Energy and NVIDIA to develop a digital twin platform intended to help geothermal operators enhance power generation and manage underground reservoir performance in real time.
What Is the EGS-Twin Project?
PNNL said Monday the initiative will create the Enhanced Geothermal System Twin, or EGS-Twin, a virtual platform that simulates the behavior of geothermal reservoirs using artificial intelligence and computer modeling. Researchers said the technology is designed to help operators quickly assess reservoir conditions and make operational decisions that maximize electricity production.
According to PNNL, geothermal energy accounted for approximately 0.4 percent of U.S. electricity generation in 2023, despite growing interest in the resource as electricity demand increases.
How Will the Collaboration Work?
The project will combine field data from Fervo Energy’s geothermal sites in Nevada and Utah with NVIDIA’s AI and accelerated computing technologies to develop the EGS-Twin system. PNNL researchers will train AI models on production data and integrate them into NVIDIA Omniverse-based simulations to create a digital representation of subsurface geothermal systems.
The platform is intended to provide operators with near real-time insights into reservoir conditions, reducing reliance on modeling tools that can take weeks to generate results.
What Are the Expected Outcomes?
Researchers said the platform is intended to help operators respond more quickly to issues affecting wells, reservoirs and pipelines while providing visibility into how injected water moves through fracture networks to maximize heat extraction. PNNL said EGS-Twin, funded by DOE’s Hydrocarbons and Geothermal Energy Office, is expected to be ready for deployment by 2029 and will be available to other geothermal operators using anonymized data.
