Numerous deepwater, subsalt depth-imaging projects have shown the value of low frequencies. In this study, that observation was exploited through a series of acquisition experiments that were conducted over four fields in the Gulf of Mexico. The purpose was to see how much added benefit could be derived in processing and interpretation when rich low-frequency content was provided at the outset. The experimental methodologies consisted of an integration of steps that generated, preserved, and recorded more low-frequency signal while attacking the accompanying high-amplitude noise. In addition to giving attention to the low end of the spectrum, some of the experiments successfully focused on providing high frequencies as well.
Deepwater subsalt plays are among the most sought-after prizes in the Gulf of Mexico. Rewards can be great, but risks are high. Seismic surveys play an important role in finding the reservoirs, but unfortunately, data quality is often dismally poor. This has provided keen motivation for improving the seismic product. This effort has focused mostly on data processing – with special emphasis on depth imaging, tomography, and multiple suppression. However, although such efforts are surely necessary, they are probably not sufficient. That is, a more substantial approach to improvement needs to include development of data acquisition techniques that provide a better starting point for processing and interpretation. To this end, a study was performed that included a series of 2D swath-style, streamer acquisition experiments. The experiments were conducted over four fields in the Gulf of Mexico in late 2003 and early 2004.
From sister studies, it was concluded that key causes of degradation in subsalt data quality included poor illumination and insufficient accuracy in overburden velocity models. For best illumination, wide-azimuth geometries should have been used in the field experiments. However, sufficient time and resources were not available, so better illumination was sought simply by extending the lengths of the streamers. Consequently, in some experiments the inline far offset exceeded 12,000 m. With reference to the velocity issue, in a controlled numerical experiment, Christof Stork (personal communication and Kapoor et al., 2005) showed that when the correct velocity field is input to migration, a WEM algorithm is capable of preserving frequency content in the final image - even in zones beneath complicated salt bodies. When the migration velocity field was smoothed, frequency content in the final image was maintained fairly well outboard of salt, but not beneath the salt. In that case, high frequencies were sacrificed. One implication is that in order to image acceptable bandwidth beneath salt, accurate interpretation of the topsalt boundary must be performed when generating the migration velocity field. This is sometimes difficult. For this reason and others, the signal that is present in subsalt migrated images is very often characterized only by low frequencies. So a key objective in this study was to devise acquisition strategies that were faithful to low frequencies. In some experiments, the approaches were in the same spirit as those discussed by Ziolkowski et al.
