Nonlinear AVO Inversion Based on Zoeppritz Equations and Its Applications

Abstract

Amplitude Variation with Offset (AVO) inversion is an effective method to estimate elastic parameters of target subsurface layers. However, most conventional AVO inversions are based on linear approximations of Zoeppritz equations assuming weak contrasts and seismic isotropy. For this reason, conventional AVO inversions cannot correctly estimate elastic parameters of some geologically important formations (e.g., organic-rich shale, formations beneath salt-dome or carbonate, etc.) which are highly anisotropic and often surrounded by hard layers generating strong contrasts. To overcome this challenge, I apply a non-linear AVO inversion based on exact Zoeppritz solutions for seismic reflection amplitudes. In the dissertation, I will describe how the non-linear inversion can efficiently be implemented and how it can be beneficial for both conventional and unconventional reservoir characterization. Specifically, I will show inversion results applying the inversion with models that represent strong contrasts and strong anisotropy. Direct outputs of the inversion are three contrast parameters between overlying and underlying layers and one background values. However, with an assumption that the model parameters of the overlying (or underlying) layer are known, the proposed AVO inversion can accurately determine horizontal and vertical P-wave and S-wave velocities of target layers of the models. Therefore, seismic anisotropy of the models can also be determined. In addition, this inversion provides a benefit to differentiate density from acoustic impedance. Based on the results, I will propose a workflow to define geomechanical properties (i.e., Young’s modulus and Poisson’s ratio), and total organic carbon (TOC) content of target layers. Furthermore, I develop a new inversion method to improve the Zoeppritz AVO inversion by jointly using PP- and converted PS-seismic reflections together. This joint inversion provides better estimations of S-wave velocity, shear impedance and the ratio of background P-wave and S-wave velocities. Lastly, I apply the inversion into more complicate field seismic data in order to demonstrate the superiority of the method compared to linear inversions. With the given field example, I also introduce a new seismic attribute, ∆VP, in order to more practically estimate seismic anisotropy from real seismic data. The ∆VP attribute is well correlated with values of gamma-ray (GR) log, which can assess the amount of shale contents and that should be highly correlated to seismic anisotropy. Therefore, the effectiveness of the attribute for inference of seismic anisotropy can indirectly be verified. In summary, the non-linear Zoeppritz AVO inversion provides better estimations of elastic parameters for such challenging situations: strong contrasts in properties, and strong anisotropy in seismic velocities. With the proposed workflow and attribute based on the inversion results, further estimations of seismic anisotropy, geomechanical properties, and TOC can also be possible. On the basis of the study, the inversion can consequently contribute to the well placement, stimulated reservoir volume (SRV), and the completion design for both conventional and unconventional reservoirs.