Robert E. Wellborn1
Search and Discovery Article #40065 (2002)
*Adapted for online presentation from the Geophysical Corner column in AAPG Explorer November, 2000, entitled “Finding Limits of 3-D Seismic,” and prepared by the author. Appreciation is expressed to the author, to R. Randy Ray, Chairman of the AAPG Geophysical Integration Committee, and to Larry Nation, AAPG Communications Director, for their support of this online version.
1Chief geologist, Wold Oil Properties, Casper, Wyoming
Some of you may recall the early 1950s, when the reflection seismograph was making great advances. Some advisors said conventional petroleum geology exploration was being replaced and it might be wise to change college majors to some other field, perhaps geophysics. At the same time, I recall graduate school at Berkeley, with professor C.M. Gilbert musing that “geology is an art, not a science,” referring to the importance of an educated guess. The talk about being replaced was flat-out wrong, but Gilbert's idea still remains true.
Admittedly the giant strides in the science of seismology, including 3-D and 4-D seismic, coherency cubes, complex processing parameters and the like, are mind boggling to many of us. It might be easy for some individuals to succumb to the final seismic product as being “gospel” -- and perhaps being able to replace old-fashioned geology. However, there are circumstances where the seismic can begin to fail us, particularly in steep-dip areas where reliable reflection migration becomes more difficult. And even worse than recording “no-data” in these steep-dip areas, we sometimes see erroneous events “sneaking in,” which can lead to an incorrect structural interpretation.
Various geophysicists have told me they have encountered a similar problem, as illustrated by the 3-D seismic program over Crooks Gap Anticline in Fremont County, Wyoming.
Figure 3. Detailed comparison of structural cross section using well data (a) to a 3-D seismic section (b). Note the location of steep dips and faulting based on wells vs. the interpretation of seismic reflectors. Kf is Cretaceous Frontier, Kd is Cretaceous Dakota, Jn is Jurassic Nugget and Pt is Permian Tensleep. Line of section in Figure 2.
Figure 5. Close-up comparison of a structural cross section using well data (a) to a 3-D seismic section. Seismic does not image the steep dips at the Dakota and Nugget levels. Line of section in Figure 2.
Crooks Gap Field is in the northeast corner of the Great Divide Basin (Figure 1), in a structurally complex area thought to be suited to 3-D seismic exploration. In 1994, a 27-square-mile seismic to drilling option survey was conducted to define further the Crooks Gap structure for possible additional drillsite locations and to check for other structural leads.
As a result of this survey (Figure 2), three different operators with their own geophysical departments drilled three unsuccessful Crooks Gap option wells. All drilled close to prognosis to the Mowry Shale, which then became anomalously thick due to a fault repeated and overturned section caused by multiple bedding plane detachments. Also, all of the beds on the southwest flank rapidly roll over into very steep to vertical dips.
Figure 3a is a geologic cross section through some earlier wells across Crooks Gap Field, and defines an anticlinal axial plane dipping at 18 degrees from vertical. Significantly, the #12 well has dipmeter dips changing from the northeast to the southwest within the Frontier Formation.
The first indications of problems with the 3-D seismic interpretation show up with the seismic cross section (Figure 3b), which follows this same line of wells. It became obvious that erroneous seismic events were continuing updip beyond the known crest of the structure for Dakota and deeper formations. Also, the structural axes for all of the seismic horizons were stacked up vertically, oblivious to the inclined axial plane.
Of particular interest is the strong southwest continuation of the deep Phosphoria-Tensleep (Pt) seismic events. In two wells, which penetrated the Paleozoic at Crooks Gap, E-log correlations indicate overturned beds in the Phosphoria, contrary to the flat seismic reflections.
Although the farmees were aware of the geologic discrepancies, they decided the seismic data were too strong to the southwest to be ignored. They programmed the #25 well to penetrate both the Dakota and Nugget reservoirs in a structural position high to the #6 well (Figure 4b, dashed orange and yellow lines). Formations came in as expected to the top of the Mowry, followed by a "wadded up" Mowry Shale, and the hole ended up on the steep southwest flank of the structure (Figure 4a), missing the crest at the Dakota level.
Undaunted by the previous failure, another company took a farmout to drill a Nugget test just 800 feet southeast of the #25 dry hole. With velocity problems seemingly resolved, the #22 location was chosen on what was thought to be an antithetic forelimb detachment thrust (Figure 5b). Both the Dakota and Nugget formations were interpreted to be unusually high (dashed orange and yellow lines). Unfortunately, this attempt was even farther off-structure, and drilled near-vertical Mowry Shale for nearly 1,000 feet (Figure 5a).
With a prior structural analysis from existing well data, this failure might have been avoided. The postulated antithetic thrust was large enough that it should have been recognized in nearby wells, particularly the #25 hole; however, evidence of a fault is lacking. On the seismic section, this fault appears to die out too rapidly upward and does not displace or even fold the Frontier seismic events.
A third operator drilled yet another dry hole at the north end of the anticline based on similar migrated seismic data, believing they could get higher than the previous wells. Instead, the hole ended up structurally low, in 60-degree southwest dips, much like the other examples. A review of previous hole deviations in this area could have ruled out this third attempt, since wellbore deviation plots indicate these earlier wells were on the structural axis at Dakota level.
Now one might ask, "what happened?" I was advised by one geophysicist that the problem lies in the complex velocities associated with the thrust faults and in the migration of extreme dips beyond the ability of the software to handle the data properly. Prestack depth migration, the next logical step, was not conducted at the time because of the difficulty associated with building an accurate velocity model.
At this point a structural geologist should have been involved, using perhaps as much art as science -- and an occasional “educated guess.” Detailed structural studies, utilizing all possible data, such as field mapping, air photos and well control, should be done prior to the seismic survey. Structure cross sections should be constructed to compare to the seismic sections (using the same horizontal scale) to determine where and why they might differ. One should honor the dipmeters as they are usually more accurate in the immediate area than the seismic. Lacking dipmeter data, one can use directional surveys or even non-directional hole deviation surveys, in locating a structure's crest.
Most importantly, when an exploration company is gathering seismic in a complex structural area, their geophysicist should also be an experienced structural geologist, or the structural geologist should be a competent geophysicist. Since most of us lack dual expertise, it is essential the geophysicist and structural geologist work together to avoid incorrect interpretation and costly mistakes.