Use of Ocean-Bottom Seismic for Improved Imaging at the Clair Field

Abstract

Abstract One of the keys to establish an efficient depletion plan is good seismic imaging. In some cases it is difficult to obtain quality seismic imaging using conventional streamer seismic acquisition. However, multi component ocean bottom seismic (OBS) data acquisition may offer an alternative which has the potential of producing significantly better P-wave imaging. The paper describes the use of multi component OBS as a tool for improved seismic imaging based on data acquired over the Clair Field. We will outline the use of OBS data for P-wave imaging, from the initial 2D test lines to acquisition of a full 3D survey, and show data examples. Introduction The Clair field is located 75 km west of the Shetland Islands; it was discovered in 1977 and put in production in 2005. The field covers an area of about 220 square km, and the reservoir is fractured, Devonian sandstone. The difficulty in obtaining good seismic imaging has been one of the reasons why the sanctioning of the development took this long. Smith and McGarrity's (2001) work on fracture characterization, which was based on borehole seismic, concluded that seismic data with full azimuth range would be advantageous for fracture characterization, and consequently three 2D OBS lines were acquired in May 2000. It turned out that the sea floor conditions were favorable and that the OBS acquisition produced good quality P- and converted S-wave data. In particular the processed P-wave data gave imaging which was clearly an improvement over existing streamer data. Encouraged by this, a 3D OBS data was acquired during the summer of 2002. Four multi component cables (one hydrophone and three orthogonal, gimbaled geophones at each receiver station) of 6km were deployed with a line spacing of 355m. Data was then recorded from shot lines 245m apart and perpendicular to the cables, with 2.5 km maximum offset from the cables. Receiver and shot intervals were 25 meters. In order to cover the survey area twenty deployments were used, and with the receiver cable area covering about 170 sq. km, over the central portion of the field, see figure 1. This acquisition geometry produces high fold and uniform azimuth - offset distribution for the cdp's. Variations in fold, sometimes called the acquisition footprint, appears mainly at offsets of 2.5 km and greater, and at lower offsets of 2-400 m, due to cable and shot line spacing. For the initial processing of the data, offsets up to 4.5 km were used, which means that the average fold is about 140. Figure 1. Map of the Clair field with outline of the 2002 4C OBS survey.(available in full paper) Data Processing The processing flow for this data contains several steps that are different from what is used for streamer data. They are dual sensor summation, noise and multiple suppression in the cross spread domain, and azimuth dependant velocity picking; processing steps which all take advantage of the acquisition geometry of OBS surveying. For imaging, pre-stack migration was used, both time and depth domain solutions have been processed.