Innovative Processing Strategy for 3D Shallow Marine OBC Seismic Data of Giant Offshore Oil Field in Abu Dhabi, United Arab Emirates

Abstract

At a giant oil-field offshore Abu Dhabi, U.A.E., an ocean-bottom cable (OBC) seismic survey was conducted to acquire seismic data in areas of shallow water and intensely developed production infrastructure. OBC data are acquired utilizing two types of detectors: hydrophones (fluid-pressure change detectors) and single- or multi-component geophones (particle-velocity or acceleration detectors). The conventional seismic processing strategy is to sum the two sensors very early in the processing sequence. However, due to numerous factors, the differences between the physical measurement characteristics of hydrophones and geophones, their data character (including noise levels, multiple content, coupling effects) can be very dissimilar. In this project a strategy was employed which avoided summing of the separate sensor data until just prior to imaging. This allowed us to investigate the differences between the two data types, optimize the processing flow for the individual sensors and then sum the sensor data in a manner which maximized the primary signal content in the dataset. Investigations of the raw field data clearly defined significantly different responses of the separate senor data to the surface wave field. This was attenuated by the utilization of a “physics based” 3D surface wave mitigation algorithm applied to the separate sensor datasets. In addition, the separate sensors clearly demonstrated wavelet character changes beyond what would be predicted from conventional ghost filter modeling. As a result, different wavelet shaping filters and surface consistent amplitude compensation corrections were required to be applied to the two sensor types. In addition, the very small scale ( geophone data, again necessitating separate compensation for this phenomenon. Finally, interpretive driven deconvolution, residual amplitude compensation, attribute validation and zero phasing were applied to the data in order to maximally condition it for subsequent quantitative and qualitative analysis. In this paper, we show a step change in seismic imaging quality on a shallowwater Arabian Gulf dataset as a result of this processing strategy. We will compare our results with those obtained using more conventional approaches.