Development and Calibration of New 3-D Vector VSP Imaging Technology: Vinton Salt Dome, LA

Abstract

ABSTRACT:Vinton salt dome is located in Southwestern Louisiana, in Calcasieu Parish. Tectonically, the piercement dome is within the salt dome minibasin province. The field has been in production since 1901, with most of the production coming from Miocene and Oligocene sands. The goal of our project was to develop and calibrate new processing and interpretation technology to fully exploit the information available from a simultaneous 3-D surface seismic survey and 3-C, 3-D vertical seismic profile (VSP) survey over the dome. More specifically the goal was to better image salt dome flanks and small, reservoir-compartmentalizing faults. This new technology has application to mature saltrelated fields across the Gulf Coast.The primary focus of our effort was to develop, apply, and assess the limitations of new 3-C, 3-D wavefield separation and imaging technology that could be used to image aliased, limited-aperture, vector VSP data. Through 2-D and 3-D full elastic modeling, we verified that salt flank reflections exist in the horizontally-traveling portion of the wavefield rather than up-and down-going portions of the wavefield, thereby explaining why many commercial VSP processing flow failed. Since the P-wave reflections from the salt flank are measured primarily on the horizontal components while P-wave reflections from deeper sedimentary horizons are measured primarily on the vertical component, a true vector VSP analysis was needed.We developed an antialiased discrete Radon transform filter to accurately model P-and S-wave data components measured by the vector VSP. On-the-fly polarization filtering embedded in our Kirchhoff imaging algorithm was effective in separating PP from PS wave images. By the novel application of semblance-weighted filters, we were able to suppress many of the migration artifacts associated with low fold, sparse VSP acquisition geometries. To provide a better velocity/depth model, we applied 3-D prestack depth migration to the surface data. The reflector dip calculated from these images were used to further constrain the depth images from the less well sampled VSP data.In spite of the above technical success, we were less than pleased with our final VSP images. Since no extra sources are used, simultaneous recording of the surface and VSP data were believed to provide a cost-effective means to acquire 3-D VSP data. However, the subsurface sampling associated with the VSP is quite different from that associated with the surface seismic data. After our analysis, we find that considerable shot infill would result in a better, unaliased subsurface image. We were able to ascertain that the subsurface illuminated by the VSP was extremely small, with the PS image being even smaller than the PP image. One-way wave equation extrapolators do not work well for the VSP geometry, where we wish to extrapolate energy sideways (from the VSP well towards and away from the salt dome) as well as vertically (away from the shots on the earth surface). Merging separately-generated images proved to be both ...