Abstract
Fractures that develop in coal seams threaten safety in many ways, but they can be predicted using fracture parameters derived from seismic data. However, the post-stack split shear waves are difficult to thoroughly separate by Alford rotation due to wavefield mixing. We propose a method of predicting fractures in a coal seam using multi-component seismic data, which was applied to coal seam 13-1 of the Huainan coalfield, China. We employed the Alford rotation to separate the split PS-waves (P-to-S converted waves) and perform interlayer travel-time inversion of the fast shear waves using geophysical logs, rock-physics parameters, and tunnel-excavation information as constraints. However, post-stack wavefield mixing of the coal seam interfered with the Alford rotation of the real post-stack seismic data. Therefore, we only performed the Alford rotation on radial and transverse component post-stack sections to derive fracture azimuths, which were then applied to the pre-stack separation of the split PS-waves. Using joint PP- and PS-wave inversion, anisotropy parameters were derived for use in fracture prediction. Finally, we predicted unsafe mining areas with a high probability of coal and gas outbursts. The application results were verified by excavation data from the mine tunnels. Our method contributes to fracture prediction on coal mine safety.
Highlights
The separation of split S-waves has been applied in many studies based on multi-component seismic data
There are thin fractured strata above coal seam [], the fracture azimuths of this coal seam can be derived directly based on layer stripping
We developed a combined method using the pre-stack Alford rotation and joint inversion for the fracture system prediction of coal seams
Summary
The separation of split S-waves has been applied in many studies based on multi-component seismic data. Fracture prediction based on split S-wave analysis is useful for both the exploration and exploitation of unconventional reservoirs and the improvement of coal mine safety. Prediction using formation micro-resistivity imaging logging and other petrophysical logs in conjunction has shown to be relatively accurate in actual cases[18,19,20] These methods have only proven to be accurate at a particular depth point and there are limitations to their application to three-dimensional geologic bodies. Additional P-wave exploration methods, such as AVO (azimuthal amplitude variation with offset) and seismic curvature attribute analysis, have been proposed to satisfy the requirements of fracture prediction[26,27]. It is possible to predict fracture development through the separation of split PS-waves, which is a more direct method than the conventional P-wave method
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
