Detecting and locating microseismic events with stacking velocity analysis for surface monitoring

Microseismic surface monitoring may suffer from the signal-to-noise ratio (SNR) in the data and lack of datasets to constrain a realistic velocity model. Therefore, an 1D depth velocity, which is derived from the well log and calibrated with perforation shot, is commonly utilized in microseismic monitoring. In this study, we simultaneously detect and locate microseismic events by applying a stacking method without using a depth velocity model. Since the 1D layered velocity model could be approximately equivalent to a RMS velocity, we borrow the velocity analysis concept in traditional exploration seismic data processing and apply it on mocroseismic surface monitoring. We stack the perforation shot waveform along a hyperbolic curve and pick the maximum stacking energy to obtain a stacking velocity above the target zone. The microseismic events are automatically detected and located by a stacking procedure with the known stacking velocity in time domain. The events are then converted from time to depth using the average velocity calculated from the perforation shot. Synthetic examples show that our approach can successfully recover the true locations of events with a large amount of noises added to the data. The real data results also suggest that this method is applicable for effectively detecting and locating events with low SNR. Although the approach with stacking velocity is limited to the layered media, this restriction is acceptable for many unconventional reservoirs.