Automated High-Pressure Gas Lift Optimization: Coupling Surface Constraints With Subsurface Dynamics
High-pressure gas lift (HPGL) is a practical artificial lift option for unconventional wells because it enables deep single-point gas injection, eliminates the need for staged gas-lift valves and packers, and allows operators to produce through either the annulus or the tubing, depending on well conditions and well life (McNeilly et al., 2024; Dalamarinis et al., 2024). However, many HPGL wells are operated at fixed gas injection rates or fixed surface choke settings, with limited guidance on when to increase injection, reduce injection, or change the operating objective. This study presents an automated HPGL optimization workflow that treats the reservoir, wellbore, and compressor as a coupled system rather than a one-time nodal-analysis design. A dynamic Python-based model was developed using PROSPER-generated Vertical Flow Performance (VFP) curves, Python-coded Vogel Inflow Performance (IPR) relationships, an empirical reservoir pressure update and digitized compressor performance curves. The model was used to evaluate standard Delaware Basin and Midland Basin well production profiles under annular-flow and tubing-flow HPGL configurations over a one-year period. Two optimization objectives were evaluated. First, production optimization targeted the maximum oil rate by adjusting compressor speed and gas injection rate while honoring compressor feasibility limits. Second, economic optimization used a target gas utilization factor to select the injection rate where the expected oil gain justified the lift-gas usage. Results show that the optimized production cases increased first-year oil recovery by approximately 780 to 5,600 STB/yr [SL3.1][JO3.2]compared with a fixed 1.0 MMSCFD injection case, with the largest gain observed in the Delaware Basin annular-flow case. The economic optimization reduced uneconomic gas injection, with gains ranging from $171,000/yr to $1,400,000/yr, depending on the basin and flow configuration. This workflow also provides a practical method for detecting over-injection by monitoring the compressor discharge-pressure response to gas-rate perturbations. Overall, the results show that HPGL optimization can be directed toward either maximizing barrels or maximizing economic value, giving operators a flexible workflow to improve production, control lift-gas usage, and avoid friction-dominated operating conditions as the well depletes.