Estimation of the Focal Mechanisms using Moment Tensor Inversion in Viscoelastic Anisotropic Media

Abstract

Abstract Estimation of the Focal Mechanism using Moment Tensor Inversion in Viscoelastic Anisotropic Media

Myungsun Kim Doctoral Dissertation Dept. of Earth Resources and Environmental Engineering The Graduate School Hanyang University

Recently, many studies have been conducted on the extraction of the focal mechanism of microseismic events. The focal mechanism of a microseismic event yields very useful information for understanding fracture behavior, improving production, and reducing seismic hazards in the shale gas field. Moment tensor inversion, which provides information about the focal mechanism, is the most commonly used method to calculate source parameters, such as source magnitude, fracture orientation, and fracture type. The shale gas layer is generally an anisotropic medium and has low quality factor (Q-factor), which leads to high attenuation of the amplitude of seismic waves because the pores are occupied with gas. Therefore, for accurate moment tensor inversion of microseismic data in the shale gas reservoir, the Green’s functions should be constructed considering the viscoelastic and anisotropic properties of the medium. If the Green’s function is calculated correctly using a proper modeling algorithm with an accurate velocity model and material properties, moment tensor inversion using the full waveform method provides the most reliable results. In previous studies, full waveform moment tensor inversion algorithms have been developed on the assumption that the shale gas layer is isotropic or anisotropic, not both viscoelastic and anisotropic. Therefore, first, in this study, a 3D finite difference modeling algorithm that simulates microseismic events occurring during the production of the shale gas reservoir was developed for viscoelastic anisotropic media by using the staggered grid method. The algorithm that was developed was verified by comparing the modeling results with analytic solutions for isotropic, anisotropic, and viscoelastic media. Second, with the developed modeling algorithm, the Green’s function used in the moment tensor inversion was calculated, and a full waveform moment tensor inversion algorithm for estimating the source parameters of a microseismic event occurring in the shale gas reservoir was developed. The synthetic microseismic data were generated with a velocity model containing a shale gas reservoir that had the low Q value of 75 and the property of vertical transverse isotropy (VTI). As microseismic event sources, an isotropic (ISO) source that has volume change, a double couple (DC) source that describes shear faulting, a compensated linear vector dipole (CLVD) source that represents a tensile fault with ISO components, and a source with the combined three components that represents a shear–tensile fault were used. The moment tensor inversion algorithms for isotropic, anisotropic, and viscoelastic anisotropic media were applied to these data. Then, the inversion results were analyzed by using moment tensor decomposition and the T–k plot, from which the source type can be identified visually. The results showed that the moment tensor inversion algorithm considering viscoelastic and anisotropic properties provided the most accurate results compared with the other inversion algorithms that did not consider viscoelastic and anisotropic properties for the shale gas reservoir. The inversion algorithms without consideration of anisotropy produced distorted inversion results because the seismic velocity varies depending on propagation direction, and the result without consideration of viscoelasticity provided incorrect fault orientation for the DC source because of the amplitude attenuation and dispersion effect due to the property of viscoelasticity. The moment tensor decomposition results also confirmed that the ISO and CLVD sources could be misinterpreted as a shear or a shear–tensile source.

Keywords: Focal mechanism, Viscoelasticity, Anisotropy, Microseismic monitoring, Moment tensor inversion, Shale gas.