Adaptive POD-Galerkin Technique for Reservoir Simulation and Optimization

In this work, a novel method with an adaptive functional basis for reduced order models (ROM) based on proper orthogonal decomposition (POD) is introduced. The intended to be applied in particular hydrocarbon reservoir simulations, where range of varying boundary conditions must explored. proposed allows us update the POD constructed specific problem setting match conditions, such as modified well locations and geometry, without necessity recalculate each time whole set functions. Such technique significantly reduce number snapshots required calculate new basis, hence computational cost simulations. was two-dimensional immiscible displacement model, simulations were performed using high resolution classical model whose adapted location geometry. Numerical show that approach by few orders magnitude compared scheme, noticeable loss accuracy calculated fluid production rates. scheme can therefore provide significant gain efficiency problems multiple or iterative are required, optimization design optimization.