Hydraulic fracturing technology has been successfully used in the production of shale gas reservoirs. Hydraulic fracturing directly induces pore pressure changes in the reservoir and associated stress. The sudden release of such a stress build-up results in microseismic events that create new fractures or activate pre-existing ones. This study is undertaken to analyze induced microseismic events to map the fracture growth and understand fluid flow in the reservoir.
A well-established processing workflow was used to process data that included enhancement of signal-to-noise ratio, detection of events by auto-picking the arrival times of the P- and the S-waves, and identification of the various phases. Three different methods were explored to detect the arrivals including root mean square amplitude, envelope method, and the Akaike information criterion. Microseismic events were recorded from the downhole array during multistage hydraulic fracturing in a shale gas reservoir. The downhole array consists of twelve 3- component sensors deployed in a vertical well. The events are triggered during the 12- stages of fracturing and perforation.
The validity of the analysis depends on the accurate localization of events that require the identification and picking of waveform phases correctly and reliable knowledge about the velocity structure. A reliable velocity structure is not always available, in most cases, the velocity model used in microseismic processing is estimated from dipole sonic logs performed along the monitoring or treatment wells before the hydraulic stimulation. Therefore, these velocities have the limitation of being representative of the propagation in the vertical direction close to the wells, which can potentially be disturbed due to the drilling and other associated perturbations around the wellbore. The workflow followed in this study has overcome the above limitations. Initial results show that all located microseismic events are well correlated with injection rates. Note a significate increase in events at the end of the injection till 10 minutes after the injection. Changes in the length and width of the fracture indicate heterogeneity along the treated volume. The spatiotemporal pattern of the events helps distinguish fluid movements and compartmentalization within the reservoir.
This analysis serves to gain experience in processing microseismic data and the results can be correlated with numerical models to understand which variables play important role in fracture propagation. This study also highlights the importance of changes in local stress in determining the efficiency of hydraulic fracturing.