Abstract: With the development of tight gas exploration,fault plays an important role in controlling the distribution of gas reservoirs and improving the physical properties of reservoirs. Restricted by many factors,such as random noise and seismic footprints,seismic data have low signal-to-noise ratio and poor imaging quality. The clarity of variance or coherence attributes commonly used to identify faults cannot meet the requirements of guiding microfault combination in plane. In order to solve the problem of seismic data interpretation,the method of likelihood attribute controlled by dip-steering filter was used to improve the accuracy of fault identification. Based on the likelihood attribute cube,the threshold of random noise and fault was determined. Combining with regional geological analysis,it is considered that when the value of likelihood attribute is less than 0.3,it is noise,otherwise,it is fault. Finally,fracture density and connectivity can be calculated by using the results of fracture prediction,which can semi-quantitatively or quantitatively describe the development of fracture,and comprehensively understand the distribution of fracture in plane,profile and three-dimensional space. This method breaks through the traditional fault identification method in seismic interpretation, solves the problem that fault and lithofacies changes cannot be distinguished by using conventional coherence method,and improves the accuracy and reliability of seismic interpretation and provides technical support for horizontal well trajectory design and fracturing scheme optimization.

Key words: dip-steering, media filtering, fault likelihood attribute, fracture density, fracture connectivity