Geology-geophysics-data mining solution enhances the identification of deep fault-karst carbonate reservoirs

Abstract

As the seepage channels of surface atmospheric fresh water and deep hydrothermal fluid, the large-scale strike slip faults in Tarim basin are easy to form high-quality carbonate reservoirs in the fault damage zone. They can connect the deep source rock and traps to form vertical migration channels during the hydrocarbon accumulation periods, which formed fault-karst carbonate reservoirs. However, the reservoir has a deeply buried depth that exceeds 6000m, with the complex spatial distribution and coupling relationships. Geophysical interpretation methods such as seismic attribute calculation, seismic inversion, conventional logging recognition, etc. are still unable to solve the contradiction between resolution and detection results. At present, it is really needed to integrate geological data with geophysical exploration data to improve the fault-karst reservoir identification accuracy. Based on the carbonate sequence stratigraphic framework, the mapping relationships of multitype geophysical data sets were established by machine learning algorithms, forming the geological-logging-seismic data fusion technology, and a set of fault-karst carbonate reservoirs prediction workflows was proposed. Note: This paper was accepted into the Technical Program but was not presented at IMAGE 2022 in Houston, Texas.