Joint Detection, Location and Source Mechanism Determination of Microseismic Events

Abstract

Imaging of seismic events is of utmost importance not only in seismology but also in exploration geophysics. For instance, detailed information about origin times, locations and source mechanisms of microseismic events is crucial for passive monitoring of hydraulic fracturing because it allows engineers to observe fracture growth, identify active faults and evaluate effectiveness of well stimulation. Microseismic monitoring greatly benefits from imaging events with low signal-to-noise ratio (SNR), but their detection on individual seismic traces is usually difficult. This work is devoted to a new imaging method based on stacking of seismic amplitudes along diffraction traveltime curves which facilitates noise suppression and improves imaging of microseismic events with low SNR. A key attribute of the proposed method consists in correction of polarities of stacked amplitudes according to the seismic moment tensor simultaneously determined for each set of amplitudes. This attribute allows taking into account source radiation patterns and significantly enhances imaging of typical seismic events generated in fault zones. The generalized imaging concept proposed in the thesis allows continuous data processing and involves original algorithms for joint detection and location of microseismic events, together with an algorithm for source mechanism determination. The detection algorithm based on automatic triggering routine enables joint determination of origin times of multiple consecutive microseismic events. In order to evaluate reliability of detected events, an additional semblance analysis of stacked amplitudes is proposed. Reliability of detection can be further improved by selection and correction of stacked amplitudes according to predicted source radiation patterns. The location algorithm involves representation of event’s image function in a form of statistical spatial distribution. This enables estimation of location uncertainty and determination of coordinates that are not restricted to the nodes of a spatial grid. The proposed method has been applied to a real dataset recorded during microseismic monitoring of hydraulic fracturing in a shale formation. Benchmarking of locations and source mechanisms of microseismic events with high SNR shows good agreement with the results of a standard imaging method based on manual picking of event traveltimes. Results of continuous data processing show that the proposed method is able to detect and locate a large number of real microseismic events in a vicinity of the stimulated well interval. Source mechanisms obtained for these events demonstrate fault geometries consistent with the orientation of natural fractures. The developed imaging method is fully automated and feasible for real-time microseismic monitoring.