| With the further development of waterflooding,the phenomenon of high water consumption is becoming more and more common,so it is necessary to explore an economical and effective method to identify the high consumption layer.In view of this,this paper briefly reviews the development process of high water consumption layer identification technology,and expounds the definition and genesis of high water consumption layer.Then,the performance characteristics of the high water consumption layer are illustrated with the area data.On the basis of fully investigating the data of regional geology,logging and development,this paper analyzes the problems faced by water injection development in the area with actual data.The Basic Law of Reservoir is analyzed,the reservoir parameters are explained,and the data construction is constructed for the subsequent high water consumption layer model.Then,based on the absorption profile,the high water consumption layer is picked up,and the characteristic parameters of the corresponding layer are extracted as the model input,and the machine learning recognition model is constructed to identify the high consumption layer.In this method,the layered information of high water consumption layer is extracted based on the data of dynamic suction profile.Then we construct the multidimensional static parameter sample set with the layer data as the constraint,and the dynamic combination.The recognition model of high water consumption layer is exported by data mining algorithm.This model is used to verify the samples with high water consumption,and the accuracy rate is up to 95%.The high water consumption layer identified by this method in practical work area corresponds well with the porosity and permeability belt distribution and lithologic distribution in Petrel work area,and the feasibility of this method is illustrated. |