Abstract: After years of exploration, reservoir prediction in Block W has entered a bottleneck period. In order to improve the inversion accuracy, facies-controlled inversion experiments are required. Facies-controlled modeling is a key step in facies-controlled inversion. To make the facies-controlled model approximate the actual reservoir, the modeling is divided into three parts: selection of facies-controlled data, selection of lateral interpolation method, and selection of parameters combining facies-controlled data and lateral interpolation. Through the analysis and parameter selection of three stages, the technical process of facies-controlled modeling is finally established, providing key input for facies-controlled inversion. Facies-controlled inversion based on the facies-controlled model improves the reservoir prediction accuracy and achieves the purpose of facies-controlled modeling. With the assistance of the latest reservoir prediction results, exploration wells have been deployed, and the application effect has been verified by newly drilled wells.