GeoTwin CCS — Intelligent Digital Twin for Real-Time Monitoring and Optimization of Geological CO₂ Storage in Deep Reservoirs

Technology area / topic:
Carbon-free energy supply — CO₂ Capture, Utilization and Storage (CCS/CCUS), Digital Twin, AI/Machine Learning for reservoir monitoring

Summary:
GeoTwin CCS is an R&D project developing and validating, over 24 months, a Digital Twin and AI-based platform for monitoring and managing permanent geological CO₂ storage in deep saline aquifers and depleted reservoirs. The platform combines Physics-Informed Neural Networks (PINNs) that embed porous-media flow physics (Darcy's law, multiphase transport) with multimodal data assimilation from distributed fiber-optic sensing (DAS/DTS), surface microseismic monitoring, and bottom-hole pressure data, enabling near-real-time CO₂ plume forecasting with automated integrity alerts. A Reinforcement Learning module actively optimizes injection rates and bottom-hole pressure across multiple wells to maximize storage capacity while keeping pressure below the caprock fracture threshold.

Technical scope:
The project is structured in two phases: Phase 1 (Months 1–12) covers geological/geomechanical characterization (facies modeling with CNNs, triaxial geomechanical testing under reservoir P/T conditions with supercritical CO₂, dynamic geocellular modeling) and development of the Digital Twin core (PINN training, RL-based injection optimization, integrated software architecture), raising the technology from TRL 3 to TRL 5. Phase 2 (Months 13–24) is a strategic field pilot: the platform is deployed and operated with real field data in partnership with an O&G operator or public institution (data-sharing discussions underway with Petrobras, ANP and CPRM in Brazil), reaching TRL 6 with an externally validated technical report.