Geologically Relevant Framework for Adaptive Fractured Reservoir Modeling

It is a great challenge to accurately model the flow through subsurface reservoirs. The fact that they are located a few kilometers below the earth’s surface means that it is very difficult to locate flow affecting phenomena. Fractures are considered to be such phenomena. They can either enhance or block flow which has a substantial effect on the flow through the reservoir. This research addresses the flow properties of fractures and the location of fractures. This research introduces analogue fracture maps to get a grasp on possible fracture locations and introduces a classification of fractures based on their orientation with respect to subsurface stress. Using this classification, fractures are assigned different permeability values, a flow affecting property. Using this approach a fracture map is obtained that can be used to model flow through a reservoir. To model flow, the projection based Embedded Discrete Fracture Model (pEDFM) is implemented. This method uses independent grids for both the matrix and fracture and couples them using a transfer function. Even when using the pEDFM method, the computational costs for running simulations through a field scale model will be too large. Therefor the concept of multiscale is introduced and developed further in this research. This research uses algebraically calculated basis functions to prolong the coarse scale solution to finescale. Besides this, flexible coarsening is introduced for a static Multilevel Multiscale (MMs) modeling approach and for Adaptive Dynamic Modeling (ADM). This allows the user to independently choose coarsening ratios for matrix and fractures. This research shows that fractures can be represented accurately using only two vertices on a coarse grid. This can reduce computational costs. Besides the flexible coarsening, the research shows how the grid refinement scheme can be used in efficiently in multiphase simulations. Using a grid re-finement scheme enables the user to capture a saturation front and to use a multiscale grid at the same time. The grid refinement scheme works most efficiently when the cells belonging to high permeable fractures are coupled to the cells belonging to the matrix. The finescale solution This research finally shows the impact of geological uncertainty. The sensitivity analysis shows how geological uncertainty can have an influential impact on the problem that has to be solved. In the sensitivity analysis it is researched how a change in subsurface stress affects the finescale solutions.