Production optimization through water-front control using adjoint gradient-based techniques
The optimization of oil production is a tedius and computational intensive process that requires the solution of time dependent nonlinear set of partial differential equations describing the flow of hydrocarbons in anisotropic porous media. Optimization of production is usually performed using either gradient free techniques like genetic, particle swarm algorithms, or gradient-based techniques where the gradients are computed through the solution of the adjoint problem. Optimization using gradients converges much faster than gradient-free techniques resulting in significant saving in computational time but it usually gets trapped to poor local optima. It is known that the optimal solution of the production optimization problem in homogeneous reservoirs requires equal arrival times of the water-front from the injector wells to the production wells. The aim of this project is to achieve production optimization by a redefinition of the objective function, which is usually defined to be the cumulative oil recovery, so that water-fronts can be controled directly to arrive simultaneously at the production wells. This project requires a highly motivated student that will co-design, develop, and implement the adjoint gradient-based method for the particular objective functions in a compositional reservoir flow simulator. Strong C++ programming skills are required as well as experience in reservoir simulation and compositional flow models.