عنوان مقاله [English]
The process of upscaling is an active area of research, due to the advances made in geostatistical techniques, which allows reservoir properties, such as absolute permeability, to be described by more and more fine models. However, these detailed reservoir models cannot be used directly for numerical simulation because they require high amounts of time and memory storage capacity. As a consequence, it is necessary to coarsen detailed geological models to a scale more appropriate for reservoir simulation. The first step in upscaling is the upgridding of the original geological model. Once upgridding is completed, the reservoir properties are reassigned to the simulation grid using various property upscaling methods. An important issue in upscaling is that the upscaled model must maintain the ssential eterogeneity details of the original geological model. Moreover, the flow characteristics of the upscaled and eological models must be the same.In this study, a new and simple method for the upscaling of absolute ermeability geological models is presented. This method, based on quadtree decomposition, upscales the geological model such that the grids of the upscaled model are unstructured. In other words, in the regions of the geological model where permeability variation is high, the grids of the upscaled model remain small, but, in regions where permeability is almost uniform, the grids of the upscaled model become large. So, this method reduces the total number of grids, and results in large savings of the time and memory required for flow simulations in the reservoir. The grids of the upscaled model have different sizes and each grid may be adjacent to several grids with several arrangements. Simulation of flow through such complex upscaled models is very hard. To solve this problem, we present a new method, in which, each side of all grids reduces the number of neighbors to a maximum of 2, and facilitates discretization of the flow equations.