| Title | Transforming Geothermal Exploration: The Integration of Physics-Informed Machine Learning and Digital Twins |
|---|---|
| Authors | Mahmoud AlGaiar |
| Year | 2026 |
| Conference | World Geothermal Congress |
| Keywords | Machine Learning, Digital Twins |
| Abstract | This comprehensive review examines the synergistic integration of physics-informed machine learning (PIML) and digital twins (DT) in geothermal energy exploration and development, with a focus on improved resource characterization, reservoir modeling, and production optimization. The objectives are to evaluate current PIML-DT applications, identify emerging trends, assess challenges and limitations, and outline future research directions. The methodology involves a structured literature review approach that summarizes existing research on PIML-DT integration in geothermal energy. Relevant publications, conference proceedings, and industry reports are reviewed to identify key elements, methodologies, and applications. A multidisciplinary perspective integrates insights from geoscience, data science, engineering, and energy research. The review focuses on current initiatives to identify knowledge gaps and challenges for widespread adoption in geothermal applications. The results demonstrate significant advances in PIML-DT integration, leading to improvements in reservoir characterization and modeling accuracy, enhanced production optimization and forecasting, reduced uncertainty in geothermal resource evaluation, and increased computational efficiency. Successful applications include the GOOML framework, the Texas A&M Institute's geo-engineering digital twin, GeoThermalCloud models, and various digital twins adopted by major energy companies. In addition, the involvement of data giants such as Google in geothermal projects paves the way for further applications of data-driven approaches to geothermal energy. These examples demonstrate the transformative potential of PIML-DT integration in geothermal energy development, enabling data-driven decision making, improved resource management, and reduced exploration risk. This research presents one of the earliest reviews of PIML-DT applications in geothermal energy, identifying emerging trends and challenges and outlining future research directions. |