Enhancing Gas Lift Optimization in Oil Fields: Integrating Data-Driven and Physics-Based Approaches for Improved Decision-Making

Gas lift optimization (GLO) is essential in the oil and gas industry to improve production rates, operational efficiency, and economic viability of declining oil and gas fields. However, determining the optimal gas lift injection rate (GLIR) while considering the back pressure effect of multiple wells is challenging. This paper explores the integration of data-driven GLO with traditional physics-based models to enhance decision-making. The methodology involves gas lift performance curve (GLPC) fitting and lift gas allocation optimization. GLPCs exhibit nonlinearity and discontinuous behaviors, which are approximated using piece-wise linear functions for efficient solution methods. The GLO problem formulation incorporates various constraints, including operational, economic, and safety factors. The hybrid data-driven and physics-based approach leverages advanced analytics and automated data collection for frequent optimization and timely decision-making. This combination enhances production and operational efficiency, offering real-time recommendations for maximizing oil and gas production. The use of data-driven GLO complements traditional methods and facilitates continuous improvement of gas lift operations.