Generation-to-Storage Data Framework for Large-Scale Carbon Sequestration Simulation Datasets

Carbon sequestration is a promising CO2 removal technology, particularly for the oil and gas industry. Consequently, monitoring the CO2 injected into deep geological formations is crucial for containment and environmental safety. This paper presents an application of data engineering principles to managing large datasets from numerical flow simulations of underground CO2 injection. By integrating data lifecycle phases with a Medallion-like architecture, the proposed framework delivers consumable assets for physics-informed machine learning experiments. One expects it provides insights into gas plume motion tracking, supporting reservoir monitoring in the post-injection phase of Carbon Capture and Storage (CCS) programs.