Mixed integer linear programming approaches to optimising oil production in offshore petroleum fields

In this work, mathematical models were developed to maximise oil production in offshore petroleum fields. The wells considered can be equipped with continuous gas-lift, electrical submersible pumps and naturally flowing, and connected by manifold or satellite. Platforms' fluid processing capacity and operational condition were considered as well. Two nonlinear models were developed: one model that considered the pipeline network as a fixed parameter, analysing the system configuration; and another that allows the assignment of wells to manifolds, analysing production improvement with network changes. Using piecewise linearisation for multidimensional functions, four mixed-integer linear formulations were proposed. Two-variable functions were linearised using the diagonal method and triangle method and three-variable functions were linearised using plane method. Using a real scenario, there were investigated the piecewise refinement, computational complexity, and solutions error. Results shown the model's applicability for daily adjustments and indicates the necessity of new investments in pipeline connections.