Determination of seismic anisotropy parameters from multicomponent vertical seismic profiles for improved seismic imaging and reservoir characterization.

摘要

Multicomponent vertical seismic profile (VSP) data can be used to determine seismic anisotropy more accurately. First, I modify the slowness-polarization method by including both P- and SV-wave data for estimating the anisotropy parameters delta and eta of VTI (transversely isotropic with vertical symmetry axis) media. Then I apply the technique to a multicomponent VSP dataset from the Wattenberg Field in Colorado, USA.;The importance of the derived anisotropic velocity model from the joint P- and SV- wave slowness-polarization method for reservoir characterization at the Wells Ranch VSP area is: 1) identifying the possible existence of open fracture networks in the Niobrara Formation at the VSP well location, 2) improving the quality of the Niobrara Formation image which is vital for future drilling programs, 3) accurately depicting the structure in the well vicinity and finally 4) determining elastic properties of the Niobrara reservoir. To identify the existence of open fracture networks, azimuthal AVO response of top of the Niobrara Formation at the VSP well is analyzed. To correct the azimuthal AVO response for propagation phenomena, using the anisotropic velocity model from the joint slowness- polarization method, I modified the moveout-based anisotropic spreading correction (MASC) technique for the VSP data.;The azimuthal AVO analysis shows very weak azimuthal anisotropy at the top of Niobrara Formation near the VSP well. This result indicates the lack of open natural fractures at the Niobrara Formation in this area and explains the low production associated with the well. In addition, I used the anisotropic velocity model obtained from the joint slowness-polarization method to build a 2D VSP image. Comparing the final VSP images using the isotropic and anisotropic velocity models with well data shows that the anisotropic image is more accurately depicted and if inverted would give more robust elastic parameter definition.