Micro-seismic Reverse Time Localization Based On Image Correlation Technology

In unconventional oil and gas develop such as shale oil and shale gas,microseismic monitoring technology can help to develop oil and gas,increase production,and guarantee production safety and environmental safety through describe hydraulic fracturing effects and fracture properties.Micro-seismic records usually have a large amount of observation time,a large amount of data and the signal-to-noise ratio is relatively low,so that the conventional method of localization micro-seismic by picking up first arrival is limited.Drawing on the idea of offset imaging in reflective seismic exploration,some focal location methods that do not require pick-up first-arrival have been rapidly developed in recent year.In this paper,we simulate the propagation process of elastic waves in homogeneous media and layered media by staggered grid finite difference method through learning the finite difference of staggered grids which be used in forward modeling of elastic wave seismic waves and CPML boundary conditions,and analysis of characteristics about micro-seismic wave field.Based on the idea of migration imaging in seismic exploration,we compared several imaging conditions in the inversion positioning problem.We found that the existing reverse time positioning method has characteristics that fast calculation efficiency but low precision,or high precision but slow calculation efficiency.At the same time,we analyzed and compared the effects of auto-correlation imaging conditions and cross-correlation imaging conditions.Combined with the two methods and ideas,the method of image correlation method which have both the computational efficiency of auto-correlation imaging conditions and imaging accuracy and noise immunity of cross-correlation imaging conditions is proposed.It improves the resolution of reverse time imaging without excessively increasing the amount of computation.Finally,obtained a method with both computational efficiency and imaging accuracy by the stochastic strategy based on the imaging method of packet cross-correlation is grouped by random selection detector.