Application of Lithology Intelligent Identification Method in Geosteering

Abstract

The geosteering in the drilling filed is mostly to analyze the logging curves of adjacent wells or pilot wells. The identification of the formation is achieved by manually comparing the microscopic features of the curve during the drilling process. The artificial stratum identification method requires geologists to have abundant experience, and this method needs complex multi-source data analysis with low efficiency. This article researches the principle and solving method of Support Vector Machine classification algorithm. The identification of stratigraphic lithology is realized by the intelligent algorithm of Support Vector Machine. The actual data of multiple horizontal wells in the Junggar Basin are validated to verify the effectiveness of this method. When the recognition result continuously appears in the lithology of the marker layer, it can be considered that the well trajectory has been drilled to the marker interval. The research shows that the intelligent identification method of stratum lithology can be applied to the horizontal well geosteering, assisting the geologists to identify the marker layer and reduce the manpower workload.