The work presented in this paper focuses on the application of an anisotropic velocity model in determining microseismic event locations from surface-acquired passive seismic data. The Thomsen parameters ε and δ were determined to accurately locate calibration shots to their known location. Hydraulic fracture events where then imaged and compared to their locations derived from processing incorporating an isotropic velocity model.
Velocity models used in the processing of surface microseismic data are in many cases initially derived from sonic logs and subsequently adjusted based on calibration shots (typically perforations or string shots). A scalar shift is usually applied to the velocity model to locate events at depth. Although calibration shots can be located with sufficient accuracy, this method does not directly account for the anisotropic nature of shales. As determined by Thomsen (1986), anisotropy for nearly vertical wave propagation, is mostly governed by the parameter δ, which is "an awkward combination of elastic parameters" (Thomsen, 1986), and appears to be sensitive to the conformity of the contact regions between clay particles, as well as to the extent of disorder in their orientation (Sayers, 2005). However, the importance of ε increases with with an increasing horizontal component of the propagation path.
Event location accuracy in surface microseismic monitoring is known to be fairly robust using the regularly assumed isotropic velocity model (Thornton, 2011), but this can be further improved in some instances by determining ε and δ to account for velocity anisotropy (Eisner et al., 2011). When compared directly to calibration shot locations derived with an isotropic velocity model, we showed that the absolute average error in calibration shot positioning in all directions was improved by almost 30% and hypocenter events from the hydraulic fracturing treatment depicted a more dense and confined zone of microseismic activity.