1 Optimization of Production from Mature Fields P. Wang, Stanford University, USA; M. L. Litvak, BP, USA; Khalid Aziz, Stanford University, USA Abstract In many mature fields, the production of oil, gas, and water is facility constrained. For such fields, optimal use of existing surface facilities is the key to increasing well rates and/or reducing production costs. Here we propose solution procedures for such nonlinearly constrained production optimization problems. The objective function of the optimization problem is oil or gas production from the field. Production is subject to multiple flow rate constraints at separators, pressure constraints at specific nodes of the gathering system, total gas-lift volumes, and maximum velocity constraints for pipelines. The control variables are the well rates, gas-lift rates, and well allocations to flow lines. The problem is formulated as a mixed integer nonlinear optimization problem and is solved by a heuristic nonlinear optimization method. The optimization algorithm is coupled with models for multiphase fluid flow in the reservoir and surface pipeline network in a commercial reservoir simulator + . The proposed procedure was tested in a Gulf of Mexico oil field and a published example, and then applied to the Prudhoe Bay field in the North Slope of Alaska. Results demonstrate the effectiveness and business value of the developed tools. Introduction In some mature fields, oil production is constrained by the gas and/or liquid handling capacities of surface facilities. While facility expansion may be an option to increase rates, it may not be the optimal choice. An economic alternative is to make optimal use of existing production facilities. In this study, we address the following operational decisions to enhance production: 1. How to control well rates with chokes? 2. How to distribute available lift-gas among specified wells? 3. How to route fluids by switching well connections to flow lines? These operational decisions are constrained by multiple capacity constraints in production facilities and wells, along with velocity constraints in flow lines to avoid excessive corrosion/erosion. Various aspects of the gas-lift optimization problem have been studied by Kanu et al. 1 , Buitrago et al. 2 , Nishikiori et al. 3 , and Martinez et al. 4 using various optimization techniques, i.e., the equal-slope method, a Quasi-Newton method, and a genetic algorithm. Fang and Lo 5 proposed a linear programming model to optimize lift-gas subject to multiple nonlinear flow rate constraints. In these studies various gas injection scenarios were evaluated using gas-lift performance curves for individual wells, ignoring interactions among wells. Dutta-Roy et al. 6 analysed a gas-lift optimization problem with two wells sharing a common flow line. They pointed out that when flow interactions among wells are significant, nonlinear optimization tools are needed. They applied a Sequential Quadratic Programming (SQP) method to a linearly constrained gas-lift optimization problem with 13 wells and showed the advantages of their method. Oil production in the Prudhoe Bay and Kuparuk River fields is constrained by the gas handling limits of surface facilities. Barnes et al. 7 presented the Western Production Optimization Model (WPOM) for the Prudhoe Bay field. This model maximizes oil production while minimizing the need for gas processing. Their model allocates the oil rate and gas rate to surface facilities and wells based on the “incremental GOR” concept, which produces the next incremental barrel with the lowest GOR. Litvak et al. 8 built an integrated reservoir and gathering system model of the Prudhoe Bay field, and employed some heuristic methods to allocate well connections to manifolds. Gas-lift rates were allocated based on gas-lift tables of gas-liquid ratio, liquid well rate and water cut. Lo et al. 9 developed a linear programming (LP) model that + For the work reported here the VIP simulator of Landmark Graphics Corporation was used.