Treffer: A smart injection optimization for CO₂ sequestration: ensuring storage integrity through adaptive control and real-time monitoring.

• Introduces a Smart Injection Field framework for CO₂ sequestration integrating real-time monitoring and adaptive control. • Compares PID, Reinforcement Learning (RL), and Genetic Algorithm (GA) strategies for dynamic CO₂ injection management. • Demonstrates up to 84 % reduction in CO₂ leakage through adaptive and optimized injection control. • Reinforcement Learning provides the most balanced approach, ensuring minimal leakage and stable reservoir pressures. • Framework enhances long-term storage integrity and offers transferability to EOR and gas storage operations. The oil and gas industry is undergoing a transformative shift towards digital and smart fields, driven by the integration of artificial intelligence (AI), machine learning (ML), and real-time data analytics. Within this context, the novelty of this study lies in the development of an adaptive smart injection system for CO₂ sequestration that integrates real-time monitoring with advanced control strategies. Unlike conventional injection schemes that rely on pre-defined injection plans, the proposed framework dynamically adjusts injection parameters to optimize storage efficiency while mitigating leakage risks. A fully three-dimensional reservoir model with three injection wells and one legacy well is simulated over 5 years of injection followed by 50 years of storage, using a commercial reservoir simulator coupled to Python-based supervisory control. Three control strategies Proportional Integral Derivative (PID) control, Reinforcement Learning (RL), and Genetic Algorithm (GA) based optimization are compared under a conservative bottom-hole pressure limit tied to the fracture gradient. In an uncontrolled case, 10.2 % of the injected CO₂ leaks through the legacy well. The smart injection framework reduces this leakage to 2.8 % with PID, 2.0 % with GA, and 1.6 % with RL, corresponding to an 84 % reduction for RL relative to the baseline. RL provides the greatest average leakage reduction and most adaptive response to changing reservoir conditions, whereas GA offers slightly higher leakage but the most consistent performance across realizations; PID serves as a simple benchmark with limited adaptability. These results demonstrate that AI- and optimization-driven control can substantially enhance CO₂ storage security and operational efficiency, with direct transferability to waterflooding, enhanced oil recovery, and underground gas storage operations. [ABSTRACT FROM AUTHOR]

Copyright of International Journal of Greenhouse Gas Control is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)