Davis W. Ratcliff, David J. WeberCareful application of traditional processing and interpretation methods to new 3D seismic data is identifying exploratory prospects in a mature part of the Gulf of Mexico. The methods have highlighted large structures on the upthrown side of the Corsair fault-the area below the fault plane. The Corsair, a listric growth fault system, extends for 320 km along the Texas shelf. Although the downthrown side of the system is a mature producing province, the upthrown side of the fault has received little attention from explorationists. Figs. 1(43882 bytes)and 2(32280 bytes) show how structures are beginning to appear in the upthrown sedimentary volume, often considered to be incomprehensible in seismic data. They are 3D time-migrated automatic gain control sections strike to the fault plane and dip to events below the fault. On these sections, structures and faulting appear above the Corsair fault, as they have in past seismic representations. What is new is the clear reflectivity below the fault. The figures highlight several structures as well as faulting below the main fault. None of these structures have been drilled. Few wells have been drilled through the fault plane into the upthrown side, and those that have penetrated deeper have bottomed in shale. This has led some operators to postulate a sand liability below the main fault-plane reflection. Yet sampling to date of sub-fault sedimentation is far from sufficient to condemn the entire upthrown area on that basis.
Diamond Geophysical Service Co.
Houston
Seismic problems, solutions
Drilling below the fault has been further discouraged by difficulties in mapping drilling targets. In a sense, structures in the upthrown side of the Corsair fault are analogous to those now yielding discoveries beneath horizontal salt bodies elsewhere in the Gulf of Mexico: Until recently, they have been impossible to image with seismic energy because of the fault plane, which, like salt bodies, severely degrades the data. Lessons from subsalt seismic work help in solutions to problems related to the upthrown side of the Corsair fault. Velocity changes introduced by the fault can lead to errors in stacking of the seismic data, for example. Techniques used in solving the many velocity problems created by salt bodies can make velocity picking beneath the fault plane accurate enough to prevent misstacking (OGJ, Oct. 24, 1994, p. 43). In addition to velocity problems, multiples at 2.5-5 sec two-way travel time plague seismic data in the Corsair fault region and elsewhere off Texas, complicating interpretations of data deeper in time. And, largely because of depth, reflection data from below the fault have been thought to be dominated by low frequencies. Dealing with these challenges has not required development of sophisticated new technologies but rather careful and, in some cases, unusually rigorous application of existing methods. One broad solution is the recent increase in 3D seismic surveying in the area. Until now, there simply hasn't been much 3D shooting below the Corsair fault. For suppression of multiples, acquisition of long-offset data has proven to be effective. And careful interpretation and processing of data in the frequency domain enhance the high-frequency content of the data and provide resolution evident in the sections presented here.Long-offset data
Diamond Geophysical Service Corp.uses split-spread shooting geometry to acquire data with offsets of 6,000 m-twice the length of offsets of other surveys in the Corsair fault region. The shooting technique uses two vessels, a recording vessel in front and a shooting and recording vessel in the rear. Each vessel tows three 3,000 m streamers, with a 3,000 m gap between the last recorder groups on the front streamers and the stern of the trailing vessel (Fig. 3)(17114 bytes). This geometry produces an asymmetrical, split-spread shot record. The rear boat collects near-offset and mid-offset data, and the front vessel records mid- and far-offset data. Fig. 4 (38580 bytes) shows how the long offsets suppress multiples. It is a series of common depth point gathers processed with a 3D dip moveout procedure (a partial prestack migration technique used to preserve energy from dipping reflectors that stacking would otherwise weaken). Offsets, from left to right in each gather, are from 300 m to 6,200 m. In these gathers, primary reflections appear flat, while multiples retain some curvature associated with moveout-the gain in reflection time resulting from increasing offset. (Basic seismic processing makes great use of the fact that multiples and their associated primaries do not respond identically to moveout correction since the energy that produces them does not travel through identical velocity layers in the subsurface). Stacking, or combining of the traces in the CDP gather, strengthens the aligned amplitudes of the primaries and weakens, or smears, the misaligned multiples. This is one of the advantages of CDP stacking. As these gathers show, however, multiple reflections on traces out to about 3,000 m (from the left edge of each gather to about halfway across to the right) nearly align and thus will not attenuate the multiples when stacked. From 3,000 to 6,000 m, curvature of the multiple (residual moveout) exists. When all offsets are stacked out to 6,000 m, these multiples will be attenuated. Long-offset data in this area is critical to multiple suppression and is one of the main acquisition parameters that have led to improved imaging below the Corsair fault. Data with the necessary offset lengths have been scarce or nonexistent in the region.
