The Study On Equivalent Characterization Of Fine Geological Modeling Of Heterogeneous Sandstone Reservoirs

With the development of oil field discover and the deepening geological understanding, where heterogeneity obviously controls remaining oil. Especially in reservoirs fine geologic model building process, in order to ensure an accurate three-dimensional description of the reservoir heterogeneity distribution characteristics, the size of the grid is limited in a certain range, because the finer the mesh the more accurate description of geological information. This makes the number of mesh nodes in a fine reservoir geologic model reach tens of millions level or more. Moreover, reservoir numerical simulation is one of the most effective means to design and test development programs. But due to the problems of calculation convergence and computing capacity constraints grid nodes can not be more than one million level, this contradiction results in the need for a certain degree roughening of fine geological model to reduce the number of grid nodes before numerical simulation. Therefore, scale-up algorithm is greatly concerned, but also great progress is made, but so far, equivalent characterization of roughening before and after is still not fully realized, which means regardless of any mathematical algorithm, the original geological information is difficult to ensure the effective retention in roughening progress.For the current situation, this paper takes Sa Midwest Portuguese group as an example, uses the data of special high-density well network (280 wells/km2),determines the deposition microfacies and studies the heterogeneity, establishes the high-precision point bar side laminate three-dimensional geological model, On this basis, this paper combines the conceptual model with the actual geological model to roughening limit research on their characteristics, and ultimately establishes numerical simulation results and fine geologic model of rough approximation model.This paper proposes a comprehensive methodology to identify the complex fluvial reservoir sedimentary faces. Based on high density well network, contributing factor analysis method and feature extraction method are adopted, GPTMap software automatically determines the correction phase module results, and then rationally combines the flat sedimentary micro-faces in the guidance of single channel combined mode.This paper proposes a set approach of dot dam recognition and its internal configuration mode anatomical. According to special high-density well network data and point bar deposit mode, the side of the laminate occurrence and density is determined by using a variety of empirical formula. Based on this, the key study area bar configuration is anglicized, and a fine point bar architecture 3D geological model is established.This paper presents a fine geological model boundary coarsening methods, and corresponding coarsening threshold value of study area reservoir characteristics. A variety of different degree of rough conceptual model simulation results are used to make comparison, and the seepage barrier and the coarsening effect of high-low permeability conversion interface are the main factors affecting the upscaling results. In this paper, the actual geological model coarsening effect is compared, under different circumstances high-low permeability conversion interface permeability difference are derived, high-low permeability distribution plane conversion interface is drawn by combining with plane anisotropy parameters.On the basis of reservoir structure and roughened boundary study, this paper establishes scale-up equivalent Representations under different sedimentary microfacies combined conditions, clears and distribution of high seepage barrier and high-low permeability transformed surface and geological origin, and uses the actual geological model and conceptual model to simulate the results of comparative analysis, finally points out the advantages and direction of fine geological model scale-up equivalent characterization techniques.