Elastic waveform inversion of microseismic data in anisotropic media

Abstract

Event location, estimation of source mechanisms, and anisotropic velocity model-building are the most important goals of microseismic monitoring of hydraulic fracturing. Waveform inversion (WI) could provide higher-resolution, improved estimates of the source and velocity parameters compared to conventional microseismic processing methods. Such an improvement is possible because WI employs both the phase and amplitude information contained in seismic data. The first problem addressed in this thesis is estimation of the source parameters of microseismic events recorded by downhole geophones. The developed 2D WI methodology is designed to recover the location, origin time, and moment tensor of microseismic sources embedded in 2D VTI (transversely isotropic with a vertical symmetry axis) media. VTI models are typical for unconventional shale reservoirs which often exhibit strong velocity anisotropy. The algorithm operates with multicomponent wavefields modeled using a finite-difference code, and the gradient of the objective function is calculated with the adjoint-state method. Synthetic tests for data from layered VTI media recorded by vertical receiver arrays confirm that WI can constrain all source parameters, if the velocity field has been estimated and a sufficiently accurate initial model is available. Then I develop an elastic WI algorithm designed to reconstruct the 2D anisotropic velocity model along with the source parameters from microseismic data. In this implementation the source coordinates and velocity parameters are updated sequentially at each stage of the inversion to minimize parameter trade-offs and improve the convergence. Synthetic examples demonstrate the accuracy and stability of the inversion for horizontally layered VTI media. Next, the WI methodology is extended to velocity analysis of microseismic borehole data from 3D VTI media. The performance of the algorithm is evaluated on synthetic data from layer-cake and “hydraulically fractured” (i.e., containing anomalies that simulate hydraulic fractures) models. I also present initial inversion results for a microseismic data set acquired during hydraulic fracturing in a shale reservoir. Finally, the proposed methodology is generalized for 3D estimation of the microseismic source parameters. The algorithm is designed for downhole data acquired in arbitrarily heterogeneous media with orthorhombic or transversely isotropic (TI) symmetry. Shale reservoirs become orthorhombic in the presence of natural or induced fracture systems. I apply the methodology to synthetic data from layered orthorhombic media simulated for typical acquisition geometries. The tests confirm the ability of the algorithm to recover the event locations, moment tensors, and origin times provided that the 3D velocity model is known with sufficient accuracy. The thesis results should be instrumental in implementing high-resolution microseismic methods for monitoring hydraulic fracturing.