Davis W. Ratcliff
Amoco Production Co.
Houston
Proper use of post stack depth migration can improve exploration success rates in areas characterized by intrusive salt.
The technique provides an accurate image of salt overhang reflections, as seen on Fig. 1, where the reflections are nearly coincident with sediment terminations. This type of image helps determine updip limits of possible reservoir sands and can help determine best possible well locations along the salt-sediment interface.
Amoco Production Co. has used the technique successfully in the Gulf of Mexico off Louisiana.
This paper will show how to recognize salt overhang signatures on the seismic record and will compare 2D conventional migration with 2D post stack depth migration.
It will briefly discuss quality control of the output of 2D depth migration, then compare 3D conventional with 3D post stack depth migration and describe the 3D velocity model.
OVERHAND SIGNATURES
Overhang reflection signatures important to this discussion are the ones that have traveled through salt and been recorded on the back side of the salt structure.
Figs. 2a and 2b show how this occurs. Seismic energy travels from the earth's surface through sediment and salt, reflects off the salt-sediment interface of the overhang, and is recorded on the back side of the salt body (Fig. 2a).
Fig. 2b shows a sketch of what an unmigrated stacked seismic section would look like from this structure given that overhang reflections survived the stacking process. The geophysicist has to recognize the different overhang signatures that may exist on the unmigrated seismic data in order to apply the appropriate migration algorithm.
Proper migration of the seismic energy reflected from the salt overhang requires depth migration. This requires a more complicated velocity field than more-common time migrations, which are usually based on sediment velocities.
Depth migration uses sediment velocities outboard the salt as well as salt velocities.
There are many areas in the Gulf of Mexico where signatures such as these have been recorded. Fig. 3 shows an unmigrated seismic line shot over a salt overhang structure off Louisiana.
Fig. 4 is an unmigrated time slice over this salt structure with the position of the seismic line shown in black. It shows a primary wave front forming the < 90 salt face signature. It also shows two other wave fronts, which are salt overhang reflections.
The line is perpendicular to the outer and innermost salt reflections but is not perpendicular to the middle wave front. 2D depth migration, which assumes that the seismic line is perpendicular to structure, should properly migrate the outer and innermost salt reflections but incorrectly image the middle salt reflection since the line is not perpendicular to it.
MIGRATING THE DATA
Fig. 5 is the product of conventional migration of this seismic section, incorporating a single sediment velocity. This is called a V(z) migration since velocity varies with depth.
The overhang reflections have been incorrectly migrated. They appear as odd-looking events inside the salt body.
Fig. 6 shows the post stack depth migration result. This is called a V(x,z) depth migration because the migration velocities vary laterally and with depth. This migration uses both salt and sediment velocities to drive the migration.
The V(x,z) depth migration has properly migrated the outer salt face and some of the salt overhang energy. The salt overhang and sediment reflections are nearly coincident, indicating that a correct migration velocity was used.
This salt structure appears to be detached, with the base of salt visible near the bottom of the section. Use of depth migration technology is showing that many of the so-called salt domes in the Gulf of Mexico are really detached salt bodies, which can have implications for the industry's subsalt exploration efforts.
Although the image of the structure is good, an odd-looking event remains inside the salt body in Fig. 6. It is the salt reflection that was nonperpendicular to the seismic line, visible in the time slice. Proper imaging requires 3D depth migration.
Many migrated data sets in the Gulf of Mexico have such odd-looking events inside otherwise properly imaged salt bodies. Some observers say these events are multiples, some say they are out-of-plane events, and some say they are migration artifacts.
Those statements are sometimes valid, but many times these odd-looking events turn out to be salt overhang and base-of-salt reflections. Determining what the events are requires some type of depth migration as well as forward and inverse modeling over the structures.
With the advent of the supercomputer, 2D post stack depth migration can be done very rapidly and economically.
Amoco routinely runs 2D post stack depth migrations on lines believed to be somewhat perpendicular to structures where drilling is contemplated along salt-sediment interfaces.
If the energy of an odd-looking event does not migrate to form a salt overhang or base of salt, then we can conclude that it is an out-of-plane event or some type of multiple.
Figs. 7a-c show another seismic line over a different salt structure. The outer salt face reflection and the internal overhang reflection appear on the unmigrated data in Fig. 7a.
Fig. 7b shows the conventional migration of the seismic line. Again, the salt overhang reflection is poorly imaged. It appears as a "smiling" event inside the salt body.
Fig. 7c shows the post stack depth migration result: an excellent image of the salt overhang reflection on the depth migrated data.
QUALITY CONTROL
Two quality control techniques that can be used to increase the reliability of the 2D depth migrated data will be described here.
The first technique is a migration velocity analysis involving two or three different depth migrations in which the data are migrated with a different salt velocity.
Fig. 8a is a depth section migrated with an anomalously low salt velocity of 10,000 fps. Fig. 8b is a depth migration of the same line migrated with a salt velocity of 12,500 fps. Finally, Fig. 8c is the depth migrated section migrated with the correct salt velocity of 15,000 fps.
The overhang reflections within the salt advance toward their correct positions as the salt velocity is increased to its true value of 15,000 fps. This imaging exercise increases confidence in the result of the depth migration.
A simpler, less expensive way to control quality of the output of the depth migrated data is to digitize all the salt face and salt overhang reflections on the stack of the unmigrated seismic line. Those events can be migrated via an event migration routine and then compared with the output of the depth migrated data.
3D IMAGING
Assumptions of a two dimensional world limit 2D depth migration. Since salt structures in the Gulf of Mexico are three dimensional in nature, 3D depth migration is required to enhance imaging of salt overhang reflections and surrounding sediments.
The processing sequence depicted in Figs. 9a-e shows how 3D depth migration overcomes the limits of 2D technology.
Fig. 9a shows a 2D unmigrated stack of a seismic line shot over a Gulf of Mexico salt structure. The results of conventional 2D time migration appear in Fig. 9b. Fig. 9c shows the same line processed with 3D conventional migration.
The 2D data show what appears to be an excellent outline of the salt body, but the 3D migration gives the correct image of the salt structure. The 3D conventional migration focuses all the sediments outboard of the salt, properly images the faulting, and properly images the salt face, which has traveled through the sediments.
The image of the salt face overhang, however, remains an odd-looking event inside the salt. It has traveled through both salt and sediments and requires 3D depth migration.
The next step in sophistication is 3D post stack depth migration (Fig. 9d). The switch to depth migration has properly positioned the salt overhang reflection.
Of course, 3D depth migration depends upon a three dimensional velocity field to drive the migration. Obtaining the correct velocity field is an iterative process that involves several 3D migration runs.
Fig. 9e shows the second iteration of the 3D post stack depth migration. The salt overhang reflection and the sediment terminations show better agreement than they did in the first iteration (Fig. 9d).
THE VELOCITY MODEL
The 3D velocity model that was used to drive the depth migration in the previous example consisted of a V(z) velocity function outboard of salt combined with a 3D interpretation of the salt body, with salt velocity of 15,000 fps.
Figs. 10a-b show velocity cross sections pulled from the first iteration and final iteration, respectively. The interpretation of the salt structure in the velocity model changes significantly.
The final velocity model in Fig. 10b was responsible for the excellent imaging results of Fig. 9.
Figs. 11a-b show deep and shallow velocity depth slices, respectively.
The salt body expands in areal extent from the shallow to deep velocity depth slices. It is important to remember that the 3D velocity model is the key to obtaining correct imaging of the salt structure and surrounding sediments.
The velocity model building is an iterative, time consuming, and interpretive process. Once completed, the 3D velocity model can be used in a migration program in order to properly image complex salt structures.
PRODUCTS, CONCLUSIONS
Fig. 12 shows examples of 3D post stack depth migrated data for four lines over the same salt structure, migrated with the 3D velocity model previously described
Imaging is excellent for both the salt structure and the surrounding faults. The depth migrated data are well focused and can be used as a tool to assist the explorationist in the search for hydrocarbons.
Imaging salt overhang reflections via 2D and 3D post stack depth migration is an important exploration tool.
The overhang reflections, if preserved and properly migrated, can aid in determining optimum well locations for salt overhang prospects. They also can be used to determine updip limits of possible reservoir sands.
3D depth migration technology is an iterative and time consuming process, but it can result in excellent imaging of complex salt structures. The improved images can be used by explorationists in order to more effectively explore for oil and gas.
ACKNOWLEDGMENTS
The author wishes to thank N. D. Whitmore and S. H. Gray, both of Amoco Research, for their excellent depth migration algorithms and J. D. Garing, C. A. Jacewitz, and J. A. Lynch, of Amoco Research, for their assistance in the implementation of the 2D and 3D migration algorithms. He also wishes to thank Fred Lockett for his salt interpretation and T. J. Kunz for his velocity model building work.
Copyright 1992 Oil & Gas Journal. All Rights Reserved.
