Analysis of 4-C data for AVO effects at Teal South, Eugene Island 354

Abstract

Abstract A recent 3-D, 4-C seismic survey at Teal South in the Gulf of Mexico poses a number of challenges for AVO analysis, some which are not normally encountered in conventional marine streamer surveys. Among these, some are associated with the use of a sparse grid of ocean-bottom receivers. Before embarking on AVO analysis, therefore, we believe that it is necessary to demonstrate whether the survey sufficiently samples the AVO response of the subsurface and also to determine how best to process the survey so as to preserve that response. The former question is related to how evenly the subsurface is illuminated and also to how evenly the upcoming reflected wavefield is sampled at the seafloor. Ideally, both sources and receivers would be finely spaced on regular grids. Since this is certainly not the case for our receiver grid, we have less flexibility in dealing with azimuthal variations. That is, the distribution of azimuths and offsets in CMP bins will not only be limited, but will exhibit significant changes spatially. Regarding the processing question, tests support applying a conventional AVO-preprocessing sequence. Further, they indicate that AVO analysis of this survey might better be conducted after prestack migration. Using a 3-D PSPI depth migration of common-receiver gathers, we obtained an image volume that compares favorably to that obtained from conventional CMP stack and 3-D poststack time migration. Since irregularities in the source grid can affect amplitudes in the output image, we devised a regridding procedure as an alternative to binning. Our next step is to evaluate whether the proposed processing strategy outperforms conventional processing in conditioning prestack P-P and P-S reflection data for AVO analysis. Introduction Teal South is located in the Gulf of Mexico, 160 miles southwest of New Orleans, where the water depth is about 270 ft. Oil is produced from unconsolidated Tertiary sands at depths in the 4000-8000-ft range. Small reservoir volumes and high flow rates result in rapid depletion, hence possible rapid changes in seismic response over time. Texaco's 1997 ocean-bottom survey at Teal South (Ebrom, et al., 1998) was, to our knowledge, the first 3-D, 4-Ccommercial seismic survey recorded offshore in the Gulf of Mexico. It was designed as a vehicle for learning (1) how to acquire and process useful multicomponent data in a shallow, soft-seafloor area and (2) how to conduct cost-effective timelapse reservoir monitoring in such an environment. The 1997 survey will serve as a baseline for the time-lapse study; the first repeat survey is expected to be complete in the spring of 1999. (The repeat survey is funded by a consortium of companies, facilitated by the Energy Research Clearing House.) Although the 1997 survey was recorded after the start of production, poststack amplitude comparisons to an earlier pre-production streamer survey demonstrated large changes in the seismic response of some reservoirs that appear to be related to production-induced water intrusion. This paper describes progress in developing tools and a methodology for conducting time-lapse AVO analysis of the 4-C data set at Teal South. These are needed as a consequence of the significant differences between this type of survey and others (marine streamer surveys and land multicomponent surveys). Initial questions included:In view of the sparse receiver grid, will it be possible to exploit static AVO information in a single 3-D survey?If so, will time-lapse comparisons of repeated 3-D surveys be useful?Will we have to modify or abandon our current processing approach in order to exploit AVO effects?Can we exploit converted-wave (P-S) AVO using the OBC horizontal com