AI and Digital Twins: The Future of Geothermal Energy Management

The global energy transition is accelerating, and with it, the pressure on heavy industry to reduce carbon emissions from process heating and cooling. While traditional renewables such as wind and solar dominate headlines, geothermal energy provides the foundational, uninterrupted baseload thermal supply that industries require to sustain around-the-clock operations.

However, harvesting geothermal energy efficiently is not without complex challenges. Subsurface exploration carries inherent geological uncertainties, while surface-level industrial integration requires managing highly dynamic thermal loads.

To transition geothermal energy from an underutilized source into a mainstream industrial asset, we need to look beyond physical hardware. The future of clean thermal energy lies at the intersection of renewable energy expansion and digital transformation. Within the EU-funded GeoS-TECHIS project, this digital evolution is already taking shape. By leveraging Artificial Intelligence (AI) and Hybrid Digital Twins, the project is advancing the application of data-driven approaches to geothermal energy management, demonstrating how digital tools can support subsurface exploration and improve operational efficiency.

Geothermal management has historically relied on static physical models and periodic manual assessments. While functional, these methods struggle to handle the highly variable demands of modern industrial processes, such as food processing, chemical manufacturing, or paper production.

When an industrial facility changes its production volumes, its thermal load fluctuates. If a geothermal-supported-system cannot adapt effectively to these fluctuations, energy is either wasted or supplementary fossil-fuelled boilers need to step in to bridge the gap. Digitalisation addresses this exact friction point. By integrating AI and digital twins, thermal installations shift from rigid infrastructure into more flexible and data-informed energy systems.

• Predictive Capability: AI-based methods can integrate thermal, structural, alteration, and auxiliary deformation-related indicators to support geothermal prospectivity assessment and early-stage exploration planning.

• AI-Supported Subsurface Insight: Deep learning models can support the analysis of geothermal indicators and surface-to-near-surface patterns, providing insights for resource planning and long-term geothermal system assessment. 
• Surface-Level Optimization: Intelligent automation matches geothermal heat output with industrial workflows, reducing reliance on supplementary fossil-fuelled backup systems and improving the use of thermal storage where available.

AITOWN and the AIGEM Methodology: Smart Exploration and Mapping

A primary barrier preventing widespread industrial adoption of geothermal energy is subsurface uncertainty. Drilling exploratory wells is a capital-intensive process, and a lack of precise data often deters risk-averse industrial operators from making the initial leap.

To de-risk this crucial stage, project partner AITOWN has contributed to the development of AI-driven methodologies within Work Package 6 during the first 18 months of the project (M1-M18). As a digital technology partner in GeoS-TECHIS, AITOWN focuses on developing AI-driven applications for digital transformation and industrial process digitisation in smart cities and energy sectors. Within WP6, AITOWN has spearheaded the development of the AIGEM methodology for geothermal exploration and assessment.

The AIGEM approach applies AI and machine learning techniques to support geothermal resource mapping and assessment. Rather than relying on traditional, fragmented geological estimates, AIGEM integrates open-access remote-sensing and geoscientific datasets, including land surface temperature anomalies, structural/fault-density indicators, hydrothermal alteration patterns, and auxiliary InSAR-derived deformation information, to support a more integrated approach to geothermal resource assessment. This work provides a basis for the upcoming Deliverable D6.4: AI-Driven Geothermal Mapping Assessment Report (due in Month 30), which will formally document findings on the long-term effectiveness of these AI-driven resource mapping methodologies.

Key Advancements of the AIGEM Methodology:

• AI-Assisted Indicator Integration and Prospectivity Mapping:  AI techniques are used to integrate thermal, structural, and alteration-related indicators to support geothermal prospectivity mapping. 

• Risk Reduction in Exploration: By analysing multiple geospatial and geological datasets, the model can support the identification of prospective geothermal zones and inform exploration planning. 

• Sustainable Reservoir Management: Beyond initial mapping, the generated prospectivity outputs can support future site-specific investigations and more detailed reservoir assessment when combined with additional geophysical, geochemical, and well data.

Through AIGEM, AITOWN is contributing to improved understanding of geothermal resource assessment, supporting evidence-based decision-making by researchers, industry stakeholders, and policymakers involved in geothermal energy development.

ITG’s Hybrid Digital Twins: Optimizing Surface Thermal Assets

Once a geothermal resource is securely accessed, the focus shifts to the surface, where the energy needs to be integrated into industrial workflows. This is where project partner ITG steps in, leading the technical developments within Work Package 4.

During the M1–M18 timeframe, ITG has focused on industrial digitalisation activities related to Work Package 4, advancing the implementation of Hybrid Digital Twins designed specifically for real-time thermal asset management.

A digital twin is a virtual, real-time mirror of a physical asset. A hybrid digital twin goes a step further by merging traditional physics-based thermodynamic equations with real-world sensor data and machine learning algorithms. This dual approach means that the digital model is not simply a static representation; it aims to provide a closer representation of system behaviour by combining physical models with operational data. 

Enhancing Operational Performance and Asset Management

ITG’s hybrid digital twins provide comprehensive monitoring across the system’s operational lifecycle, tracking critical parameters, environmental variables, and equipment performance indicators.

• Optimising System Efficiency and Energy Management: The digital twin framework simulates diverse operational scenarios, enabling operators to evaluate the most effective strategies for energy routing, resource allocation, and minimizing overall system losses..
• Advanced Predictive Analytics and Asset Optimization: By identifying subtle anomalies in operational data before they impact hardware integrity, the twin facilitates proactive maintenance scheduling, maximizing asset lifespan and significantly reducing unplanned downtime.
• Dynamic Integration and Demand Response: The system balances real-time energy availability alongside various storage solutions and shifting demand profiles, ensuring optimal utilization across fluctuating production schedules and operational requirements.

The collaborative innovations of AITOWN and ITG represented an important project review point during the recent GeoS-TECHIS Month 18 General Assembly Meeting (GAM) held in Barcelona. The assembly served as a vital alignment point for the consortium partners to review the technical progress achieved during the first 18 months of the project.

The real-world data and digital frameworks presented by ITG and AITOWN demonstrated how the integration of AI and physics-based modelling is being explored through practical industrial applications. The M18 meeting established a clear, structured action plan for the next phase of the GeoS-TECHIS project, supporting the transition from model development towards prototype implementation, data integration, and subsequent validation activities planned for the remaining project period. 

As these digital twins and AI systems begin absorbing live field data from the project’s demonstration site, they are expected to contribute to an empirical baseline of operational data. As the project’s digital frameworks and data streams are progressively integrated, they are expected to support validation, operational learning, and improved decision-making across demonstration activities. Together, these developments are expected to provide valuable insights into the economic and environmental viability of high-temperature geothermal systems across European industrial sectors.

Supporting the EU’s Digital and Green Transition 

The deployment of AI and digital twins within GeoS-TECHIS supports the European Union’s twin priorities: the green transition and digital transformation. Smart energy infrastructure can help reduce operational waste, directly helping industries lower their emissions under the EU Emissions Trading System (ETS) and assisting Member States in meeting their National Energy and Climate Plans (NECPs).

By combining AI-driven geothermal mapping with hybrid digital twin technologies, GeoS-TECHIS is exploring how digital tools can support the wider adoption of geothermal energy and contribute to industrial decarbonisation efforts across Europe.

GeoS-TECHIS is revolutionizing industrial energy systems through innovation and collaboration. Explore our open-source tools, contribute to our mission, or share your ideas at the GeoS-TECHIS homepage. Tweet us @GeoSTECHIS with your thoughts on sustainable industry! Together, we can build a greener, more resilient industrial future.

Let’s shape a sustainable tomorrow!

The GeoS-TECHIS Team