Multiscale method and flow diagnostics for reservoir management

Abstract

This work is a continuation of the PhD thesis of Omar Duran Triana [DURAN] who developed the computational tools to apply the Multiscale Hybrid-Mixed (MHM) [PAREDES] and Reduced Order Modeling (ROM) of geomechanic deformation [DURAN], to simulate two phase reservoir flow. The contribution of Dr Duran will be applied to a field as a fast prototype solver. As such the focus of this contribution will be: develop a computational mesh to model an oil field and recognize flow patterns that leadto accurate MHM approximations; develop advanced well models for considering wellbores as boundary conditions insteadsource terms; develop Reduced Order Models (ROM to accurately represent the geomechanics coupling of the reservoir. Single-phase flow diagnostics will serve as an efficient alternative to full physics flow simulationsfor optimization and characterization purposes. By developing a combination of these technologies in a single implementation, we obtain a workflow that makes it possible to quickly evaluate and optimize multiple realizations, while still retaining error control; mesh adaptivity will be used to adjust the accuracy of the simulation as a function of the flow pattern; we will evaluate the use of MHM approximations as a preconditioner for a full field simulator in order to obtain more eficiente results. (AU)