固体地球物理研究生学术报告

For satisfying the ever-growing demand of present production, recorded seismic data volumes are becoming bigger and bigger, resulting in the traditional location methods based on picking arrival times manually not applicable. As a result, in microseismic monitoring, migration-based location methods are widely used to detect and locate weak seismic with low signal-to-noise ratio (SNR), which with no need for manually picking P- and S-wave arrivals. With different brightness functions, the source image quality has a deal of difference. We propose a new stacking method to image source precisely. The aim of this study is to compare the stacking of different waveform attributes including the actual waveform (called linear stack), the envelope of waveform (called envelope stack), the short-time average to long-time average ratio (STA/LTA) of the waveform (called STA/LTA stack) and stacking the waveforms by multiplication (new stacking method). We test these methods using both synthetic and real datasets. Through synthetic data test that simulates a dataset of downhole monitoring added by different random noise intensities, all the methods can locate the event in acceptable resolutions with a high SNR. But when SNR decrease, those methods failed except our new method. To further test the ability to locate microseismic of these methods, we have located 30 induced seismic events with the multi-well data in an oil/gas field in Oman. The results show STA/LTA stack images more exact source location than stacking the actual waveform and the envelope of the waveform, and the new method multiplying the waveform proves the highest image resolution compared with summation.