AI-driven rapid forecasting of CO2 plume migration and trapping efficiency in geological carbon storage

• Deep learning surrogate models reduce CO 2 plume forecasting time from hours to microseconds. • An enhanced sequence-to-sequence architecture captures spatiotemporal evolution of trapped and mobile CO 2 vol. • Hybrid LSTM–MLP framework tracks plume expansion across dynamic, multi-phase injection strategies. • Surrogate models accommodate variable-rate, start-stop, and intermittent CO 2 injection schedules. • Fast forecasting supports real-time optimization and regulatory compliance in carbon storage operations. Accurate and rapid forecasting of CO 2 trapping, mobility, and plume evolution under dynamic injection conditions is crucial for effective planning, operational optimization, and regulatory compliance in geological carbon storage (GCS) projects. Traditional analytical methods often rely on oversimplified assumptions, compromising accuracy, while numerical simulations, though precise, require extensive computational resources, limiting their utility for real-time scenario analysis. To overcome these challenges, this study proposes an advanced deep learning framework for forecasting trapped and movable CO 2 fractions and plume extent. An enhanced sequence-to-sequence (Seq2Seq) neural network with a composite loss function robustly predicts CO 2 volumetrics, while a hybrid Long Short-Term Memory (LSTM) and Multilayer Perceptron (MLP) model forecasts CO 2 plume extent. The models incorporate nine static geological and reservoir characteristics and dynamic injection profiles, capturing real-world injection complexities, including varying rates and intermittent schedules with multiple start-stop events. Training utilized several hundred high-fidelity numerical simulation realizations covering diverse geological and operational scenarios. The enhanced Seq2Seq model achieved an average Mean Absolute Error (MAE) of 0.016 for trapped and movable CO 2 fractions, while the LSTM-MLP model attained an average MAE of 42 meters for plume diameter. These deep learning-driven surrogates drastically reduce computational time, providing accurate forecasts within seconds per scenario compared to conventional methods requiring hours. This significant advancement facilitates rapid, reliable decision-making, optimized storage strategies, and rigorous regulatory compliance in GCS initiatives.