An efficient method for smart well production optimisation

A method for dynamic optimisation of water flooding with smart wells is presented. The method finds the optimal injection and production rates for every well segment of the smart wells. We formulate the problem as a constrained optimisation problem and state this problem as an augmented Lagrangian saddle point problem. Comparisons are made with a more traditional optimal control method based on solving the adjoint systems of equations. In the examples tested the method obtains same maximum profit while using less computational effort.