Fracture size scaling of hydraulic fracture stimulations in shale reservoirs

Wednesday, 17 December 2014
Ted Urbancic and Adam M Baig, Engineering Seismology Group Canada Inc, Kingston, ON, Canada
It is becoming widely evident that hydraulic fracture stimulations in shale reservoirs can result in the generation of events with magnitudes M>0. These events are of concern both to the public as potential geo-hazards possibly affecting groundwater conditions and surface infra-structure, and to engineers for optimizing productivity and engineering design. Typically, in these environments, recording bandwidth limitations has resulted in a bias towards the consideration of events with M<0. This in turn has limited the observable fracture sizes to those constrained within lithological units. By extending the recording bandwidth to lower frequencies, the dimensions of the observable fractures are also extended to include larger fractures/faults activated during the stimulation. Our observations suggest that these larger-scale events can contribute upwards of 80% of the overall seismic budget or energy release associated with the stimulation process. Effective analysis of scaling relations independent of recording further suggests that breakdowns in scaling can be related to the presence of barriers to growth such as contrasts in rock properties associated with different lithological units. Generally, detected larger-magnitude events are associated with smaller-magnitude events, M<0, suggesting that these latter events can be used to characterize aspects of the rupture process whereas their associated signals observed with the low-frequency network can be used to characterize the overall fracture/fault behavior. By accounting for the presence of larger events, additional activated fracture surface area within the reservoir results in a significant increase in surface area. In an example provided, these events account for a further ~10 km2 of additional activated fracture surface area than estimated based on only utilizing high-frequency band-limited recordings. Overall, the identification of the actual discrete fracture network over many size scales allows for a better understanding of the fracturing processes and size scaling associated with stimulations and their impact on production.