Objectives/Scope: The objective of this work is to develop an integrated digital framework that combines real-time digitalization and AI-driven interpretation of well logs with dynamic reservoir simulation to enhance Measurement, Monitoring, and Verification (MMV) for Carbon Capture, Utilization, and Storage (CCUS) projects. The framework aims to digitalize and interpret well logs integrated with digital MMV, continuously update subsurface models, improve prediction of CO₂ plume migration and pressure evolution, and enable early identification of containment, conformance, and well integrity risks through automated alerts and risk intelligence.

Methods, Procedures, Process: The proposed framework integrates well log digitalization, advanced visualization, artificial intelligence, and reservoir simulation into a unified MMV workflow. Historical and real-time well logs including resistivity, sonic, density, temperature, and pressure are standardized, quality-controlled, and streamed into a cloud-based digital platform. Interactive dashboards provide continuous visualization of formation properties, fluid distributions, and integrity indicators across injection and monitoring wells.
Machine learning and physics-informed AI models interpret multi-log signatures to estimate lithofacies, porosity, permeability trends, saturation changes, and stress indicators. These AI-derived properties are used to dynamically update reservoir simulation models through assisted history matching and data assimilation workflows. The reservoir simulator predicts CO₂ plume evolution, pressure propagation, and geomechanical response under current and forecasted injection scenarios.
Simulation outputs are continuously compared with real-time log-derived observations to identify deviations from expected behavior. This closed-loop feedback improves model fidelity and reduces uncertainty in plume tracking and containment assurance. Key MMV performance indicators such as pressure limits, plume extent, caprock integrity, and wellbore stability are quantified and monitored in real time.
A probabilistic risk engine links AI interpretations and simulation forecasts to predefined thresholds and regulatory criteria, enabling automated alerts for abnormal pressure buildup, unexpected plume migration, or potential leakage pathways. The framework is designed for interoperability with seismic, surface monitoring, and regulatory reporting systems, ensuring transparency, auditability, and regulatory compliance.

Results, Observations, Conclusions: The integrated digital workflow significantly improves subsurface understanding by continuously reconciling observed well log responses with reservoir simulation predictions. Real-time updates to the reservoir model enable earlier detection of anomalous behavior compared to static MMV approaches. Automated alerts support proactive operational decisions, such as injection optimization or targeted monitoring. The framework reduces interpretation latency, enhances confidence in containment assurance, and improves the robustness and traceability of MMV reporting while lowering operational effort through automation. Real-time model updating, visualization, and risk alerts strengthen storage assurance, regulatory compliance, and long-term operational confidence in CO₂ storage performance.

Novel/Additive Information: This work introduces a closed-loop digital MMV framework that tightly couples AI-driven well log interpretation with continuously updated reservoir simulation. By transforming well logs into real-time model updates and risk intelligence, it shifts CCUS monitoring from periodic validation to predictive, proactive assurance.