Charlie Jing

CSEM inversion: Impact of anisotropy, data coverage, and initial models

While 3-D inversion is a particularly effective tool for interpreting marine CSEM data, 3-D data acquisition is needed to image the resistivity structure of an anisotropic subsurface. Tests on synthetic models show resistivity artifacts when anisotropic data are forced through isotropic inversion and when offline data are excluded from anisotropic inversions. Similar artifacts appear also when the inversion starting model differs too much from the actual subsurface. As a result, inadequate attention to anisotropy or inadequate data coverage can lead directly to misinterpretation of the subsurface resistivity structure. Anisotropic imaging with good data coverage, accurate receiver orientation, and good initial resistivity models are necessary to quantitatively image resistive anomalies in an anisotropic earth.

Inversion study of a large marine CSEM survey

The Brazil RC Marine CSEM survey was collected in April of 2004 for the ExxonMobil Remote Reservoir Resistivity Mapping (RM) Project. The portion of the survey reported here consisted of a total of 735 km of transmitter towlines arranged approximately on a 5 km x 5 km rectangular grid (see Figure 1). Vertical and horizontal electric field measurements were recovered at a total of 23 seafloor locations from a deployment of 36 seafloor instruments. Imaging of these CSEM data into full three-dimensional conductivity volumes represents a formidable challenge due to the subtle effects of reservoir targets, the volume of data and its large dynamic range. In this presentation, we report on an initial round of inversion results obtained using both isotropic and anisotropic (VTI) imaging methods. Our results support the need for an anisotropic model to accurately represent subsurface resistivity.

A comparison of streamer and OBC seismic data at Beryl Alpha field, U. K. North Sea

Compressional P-wave ocean-bottom-cable (OBC) seismic data from the Beryl Alpha field in the U. K. North Sea provide a superior image of the subsurface compared to heritage streamer seismic data. To determine the reason for the superiority of OBC data, the results of a detailed comparison of these OBC and streamer data sets are compared. The streamer and OBC data sets are reprocessed using a strategy that attempts to isolate the roles of processing, fold, azimuth, PZ combination, and hydrophone and geophone data have on the improved OBC image. The vertical component of the geophone (OBC Z) provides the major contribution to the improved OBC image. The imaged OBC Z datacontain fewer multiples and have a higher signal-to-noise ratio than the streamer. The OBC data have a lower level of multiple contamination because of the contribution from the OBC Z component, together with an effective suppression of receiver-side water-column reverberations as a result of the combination of the OBC hydrophone and geophon...

Resolvability Analysis For Rock Property Inversions of Multicomponent Seismic Data

A Singular Value Decomposition (SVD) technique has been used to investigate the resolvability and reliability of rock property parameters from linearized inversions of multicomponent pre-stack reflection seismic data. Resolution maps of rock property parameters have been generated and plotted as functions of maximum incident angle and background shear-to-compressional wave velocity ratio for different components of the seismic data: pure compressional wave PP, compressional-to-shear converted wave PSv, and pure shear wave SvSv. The sensitivity to noises of each parameter has been illustrated by the variation of the resolution matrix with the damping factor used in the SVD. Our analyses have shown that Pand S-impedance can be resolved reliably from separate inversions of PP and PSv (or SvSv) data, respectively. The S-impedance from PP data is much more sensitive to noises than that from PSv (or SvSv). Density should be able to be resolved reliably from SvSv data. Density from either PP or PSv data is likely to be unreliable on real seismic data. In order to resolve velocity and density or Lame constants and density reliably, joint inversion of PP with PSv and SvSv, especially with the inclusion of SvSv data, is needed.

Sensitivity Study of PP and PS AVO on Azimuthal Anisotropy

Multicomponent seismic data offer not only seismic reflections of pure compressional wave (PP) but also of converted shear waves (PSV and PSH) resulting from conversions at stratum interfaces in the subsurface. Different types of waves have different responses to subsurface geology. The reflection amplitude responses of different types of waves have been investigated and compared through a numerical modeling study. The shot-receiver offset and azimuthal angle have been varied relative to the subsurface anisotropy symmetry direction. Our results show that PSV data are more sensitive to azimuthal anisotropy than PP data. This makes the converted wave a better tool for subsurface azimuthal anisotropy (or fracture) detection. The study also demonstrates that the higher-order term in PP-wave amplitude versus offset (AVO) expression (beyond the traditional intercept and slope terms) can be important for fracture detection. Feasibility modeling study before field application is essential to better understand the relative importance of each term in the azimuthal AVO response and the appropriate incidence angle range for azimuthal anisotropy detections.