TECHNICAL PROGRESS

GEOPHYSICAL DATA

The 3-D seismic data were acquired between 1994 and 1995 in an area of approximately 352 mi². The data were merged from two surveys: OCS 310 (southwest) and SL 340 (northeast), covering five of Texaco's offshore fields: Starfak, Tiger Shoal, Mound Point, Lighthouse Point, and North Lighthouse Point. Both surveys are oriented NW–SE in the inline direction and SW–NE in the crossline direction. Western Geophysical Corporation (WGC) of New Orleans conducted the acquisition by using a cable crew with airgun source. Both WGC and Texaco's New Orleans office were involved in the data processing. Although some acquisition problems (dead cable, time breaks, etc.) resulted in difficulties and delays, there is no evidence of major quality problems in the final product.

Coherency time slice generated at 2,000 ms revealing the areal distribution of faulting throughout the 3-D seismic study area.

From an interpreter's point of view, the 3-D seismic data set is of good quality. Dominant frequency varies from 40 Hz in the shallow section to 20 Hz in the deep section, many of the gas reservoirs or reservoir groups being clearly resolved. Visible direct seismic indicators of gas-bearing zones include bright spots on structural highs and against faults and significant velocity sags observed in the gas-bearing areas. The signal-to-noise ratio is high, with no multiples or other coherent noises, and no migration problems are apparent. The merged data volume, however, does show some subtle differences in dynamic characteristics between the OCS310 subvolume and the SMI subvolume. The potential effect of this difference on the project should, however, be minimal because both Starfak and Tiger Shoal fields are well within the OCS310 subvolume.

Particular attention was paid to the accurate tying of wells to seismic data. Five check shots in Starfak field were used to generate synthetic seismograms. Those synthetics were used to match the well information to seismic. Available log-interpreted picks of sequence boundaries and tops of main reservoir units (mainly in Starfak field) were then loaded into the data base and checked for consistency in correlation.

Three types of post-stack processing were applied to the 3-D volume to improve interpretability of the data. First, a 90-degree phase shift was applied to the original, approximately zero-phase data. The resulting 90-degree phase data coincide better with impedance logs and therefore with gamma-ray, SP, and resistivity curves—those with which geologists are most familiar. We then calculated continuity (Landmark© ) cubes from the original data to aid in fault interpretation and identification of stratigraphic features. The primary benefit of these cubes to the project is that they have imaged numerous faults of different scales (from regional [tens of miles] to local [hundreds to thousands of feet]) and resolved important depositional features (for example, channel systems and slope fans).

Further, a processing of spectral balancing on deep data (3 s and deeper) was conducted to improve vertical seismic resolution. The deep section is an important hydrocarbon producing zone (U-Y sands, 12,000 sands and Robulus sands) and one of the critical targets of this research. However, compared with the higher frequency, shallower data (30- to 40-Hz dominant frequency in interval of 1 to 3 s), deeper data are characterized by a significantly lower dominant frequency (~20 Hz) and therefore much lower vertical resolution. To improve the interpretability of the seismic data, the amplitude level of higher frequency components in the data got enhanced, while keeping the lower, originally dominant frequency components intact. The range and enhancing scale of the higher frequency components were carefully selected and tested to avoid overamplifying the noise level of the data. The processing moved the dominant frequency of the data to the 30- to 40-Hz level (comparable to the shallower data). The resulting data reveal more stratigraphic details vertically.

Judging by the good and sometimes even better correlation between well logs and seismic events in the processed data, these improvements are real and come from originally masked higher frequency signals.

Tremendous potential for deep gas prospects is apparent. Many interpreted deep-seated faults have been identified in the structural saddle between Starfak and Tiger Shoal fields and to the northwest of Starfak field.