Gas Lift Optimization in the Permian Using Machine Learning and Artificial Intelligence

Abstract

Summary Gas lift is among the most prevalent artificial lift methods in the Permian Basin, accounting for over 50% of total oil production. Despite that, gas lift optimization has historically been a time-consuming process, which limited optimization efforts to cases with the most significant potential for improvement. Automated gas lift optimization can add incremental volumes without the need for major OPEX/CAPEX expenditures. In this paper, we present a closed-loop iterative well-by-well Gas Lift Optimization Workflow deployed to more than 1300 ExxonMobil wells in the Permian Basin. The workflow conducts multi-rate tests via remote control of gas lift injection rate setpoints in combination with automated well data acquisition. Optimal injection rate setpoints are determined by quantifying the relationship between gas injection rate and downhole pressure, and automatically maximizing drawdown/production. This is achieved with minimal disturbance to the surface facilities network. A machine learning model provides optimization recommendations for wells without a downhole pressure gauge. Production data undergoes a rigorous quality control process to ensure that measurements are accurate and representative of current well conditions. Incremental uplift is tracked using a model-based approach based on well productivity and steady-state well tests. The optimization workflow has been applied to more than 1300 wells, with an average oil production uplift of approximately 2.0%.