A Gas Lift Optimization System with Variable Choke and Action Constraints

Gas lift systems are multi-scale industrial systems involving downhole actuators, surface equipment, processing facilities, and control systems connecting them. Gas lifting oil wells works by injecting high pressure gas into the flow stream, decreasing bottom-hole pressure, and increasing oil flow to the surface. From there, fluids move through pipelines, separators, compressors, and other components in the facility. If the gas available for injection is insufficient to optimize every well, then the available gas needs to be optimally allocated to maximize field production — this is the gas lift optimization problem. This paper proposes a mixed-integer linear programming formulation using gas injection and wellhead choke as optimization variables. Surrogate well performance and choke models are piecewise linear. New constraints called action constraints and an enumeration procedure are introduced to make the solutions more easily implemented in the field. Recently published precedence constraints are also included. Numerical results demonstrate the efficacy of this approach, which has been successfully implemented in the field.