Multiscale characterization and connectivity analysis of 3D seismic attributes of ultra-deep carbonate conduction system in Shunbei, Tarim Basin

The 7000-meter ultra-deep carbonate fault-controlled karst reservoir in the Shunbei area of Tarim Basin provides a good transport channel and reservoir space for oil and gas. In this paper, the three-dimensional internal structure of the fault-controlled karst conduction system is finely depicted at different scales. Firstly, fracture-enhanced filtering technology is used to improve the signal-to-noise ratio of ultra-deep seismic data. Secondly, according to the seismic reflection characteristics of the fault-controlled karst conduction structure, it is divided into three different scale structural modes: large, medium, and small. The large-scale structure includes large strike-slip faults, karst caves and accompanying karst collapses, with regular fracture strong reflection and "beaded" reflection seismic facies characteristics. The mesoscale structure is mainly a secondary fault that develops laterally near the strike-slip fault and the accompanying karst local collapse, which has the characteristics of chaotic strong reflection seismic facies. The small-scale structure is a background crack that disperses the entire conduction medium, which has the characteristics of chaotic and weak reflection seismic facies. It is difficult to comprehensively describe the fault-controlled karst conduction medium with a single seismic attribute. Through attribute analysis, the sensitive seismic attribute groups of the three types of structural patterns are clarified. The sensitive seismic attribute groups of the three structural modes are fused by Bayesian clustering and fusion technology. Finally, the seismic attributes fusion of the three types of structural models are sampled into the three-dimensional conduction medium grid for quantitative analysis. Finally, the fracture-cavity connected structure model is established, which provides a three-dimensional transmission system connectivity model for subsequent lithology prediction and fluid identification.