APPLIED GEOPHYSICS Depth maps from seismic velocities help Wilcox exploration

C.E. Guzman, M. Ramaswamy, B.K. Wright, K.P. Lawler
Shell Western Exploration & Production
New Orleans

Applied aggressively in conjunction with new processing techniques, depth maps generated in this fashion can identify prospects not visible on time maps.

Since depth conversion depends on velocities, the method relies on precision velocity measurements derived from an event-oriented algorithm. The velocities are carefully analyzed for misties and then contoured within geologic boundaries. Well information, if any, is also incorporated into the velocity map, and then the time map is converted to a depth map.

Two case studies presented here calibrate the method. Other examples show how the technique helped locate a drillsite and predict geologic horizon depth and how it identified a structure not visible on time maps.

Wilcox characteristics

The Tertiary Wilcox in Shell's Onshore Division covers a wide strip of Texas and Louisiana (Fig. 1 [40211 bytes]).

The Wilcox sands are thought to be deposited in a deltaic environment in the greater part of the trend. Ancient barrier island systems provide a model for fields such as Fandango in South Texas.

The structural style is that of growth faulting towards the Gulf of Mexico. Structural complexity varies from simple low-relief structures of the Louisiana Wilcox to the highly faulted steep-dip traps in South Texas.

Depth-conversion method

The depth-conversion method, developed in the 1980s and tested in years since then, begins with time maps prepared from interpreted seismic lines. At a horizon of interest, or at the map level, detailed velocity analysis is conducted with a horizon-oriented algorithm.

Using prestack data, the velocity-analysis package generates MCS velocities along horizons. This enables the interpreter to track velocity variations at geologic interfaces in a lateral sense.

The next step is to post the velocities on a map, interpret, conduct a mistie analysis, and smooth velocities before contouring them within geologic boundaries. In general, velocities vary with azimuth, lithology, pressures, fluid content, and other interval characteristics. A rigorous treatment is not practical for the explorationist. The statistical redundancy of velocity measurements and the ability to examine velocities areally are key aspects of the approach.

In the last step, a depth map is made from the time and velocity maps by point-to-point multiplication. Ray-tracing through the velocity model is a preferred alternative.

Constant-velocity calibration

A site with well control was chosen to gauge the validity of the method. Analysis of the wells at Amoco's South Harmony Church field in Allen Parish, La., indicated that a southwest-northeast view of the average velocity to the top of the Cane River Marl formation is nearly constant (Fig. 2 [14335 bytes]).

The method should confirm this finding; that is, seismically derived velocities should be nearly constant in this case. Any deviation from constancy would serve as a calibration for the method.

A southwest-northeast seismic profile across the field shows a strong reflector near the Cane River Marl formation (Fig. 3 [131820 bytes]). Fig. 4 [31413 bytes] shows the result of velocity-scanning this seismic reflection.

The constant seismic velocity profile is consistent with the constant average velocity profile for the wells. The seismic velocity is greater than the well average velocity profile, as is to be expected. The seismic velocity measures the trend successfully in this case.

Varying-velocity calibration

A prospect in Goliad County, Tex., tested the method in a regime of lateral velocity variation.

The Shell Frantz well, drilled in 1981, tested a time anticline in the Middle Wilcox package. The well was surprising in that it came in structurally lower than the previously drilled Shell Bego.

A depth profile indicates that because of lateral velocity changes, the structural picture has changed drastically: The Shell Bego is downdip of the Frantz well only if a constant velocity is assumed in constructing the depth map.

The method described earlier was applied here as a calibration. A horizon velocity scan and seismic profile, shown together in Fig. 5 [100351 bytes], indicate a strong velocity gradient at this level. This trend in velocity is observed in all of the lines comprising the prospect. The depth map obtained is clearly different from the time map, and the seismic-derived results fit the well control (Fig. 5).

A defensive application

In a different application, the technique began with a time map and predicted depth to a geologic horizon of interest.

Fig. 6 [66626 bytes] is a time map of a prospect in Louisiana. Since even slight lateral velocity changes in this low-relief time structure could significantly alter a drilling recommendation, the depth-conversion method was applied prior to drilling.

Horizon velocity scan measurements were posted on a map. A mistie analysis was conducted with attention given to signal-to-noise ratio, multiplicity, data gaps, distance from end of line, strike or dip bearing of the line, and so forth. Seismic velocities were then adjusted and contoured within fault blocks (Fig. 7 [65647 bytes]). A depth map is shown in Fig. 8 [56402 bytes].

The six wells in the area were used to calibrate the results. Predicted or estimated depths are consistent with depths from well control if the bias due to use of seismic velocities instead of average velocities is removed. The exception is at Well F, where adequate seismic control was not available.

The depth map was used to predict the depth to the Cane River Marl formation at a wildcat location. It also provided the basis for selection of the well site to minimize risk. The predicted depth to Cane River Marl of 12,660 (150) ft was confirmed by the well result of 12,571 ft, a deviation of 90 ft.

A conventional method to predict depth to Cane River Marl by using the nearest check shot survey would have resulted in a deviation of approximately 500 ft.

Aggressive application

The depth-conversion technique identified a prospect not visible on time maps in Shell's Fandango field, South Texas.

The 400 bcf field, discovered in Zapata County in 1979, has strong geopressure conditions near the Queen City formation; the Wilcox section exhibits rapid lateral velocity changes. A lateral slowdown in interval velocity in the Wilcox section can be observed in the average velocity profiles from the wells (Fig. 9 [19785 bytes]). The velocity slowdown masks the depth high as a time saddle.

Results from a velocity scan of a Wilcox seismic reflector and a scan of well average velocities measured from check shot surveys are shown in Fig. 10 [34925 bytes]. As can be seen, near the top Wilcox formation, the average velocity slows down markedly (400 ft/sec/mile). The seismic measurements reveal the same trend.

Time, calibrated MCS velocity, and depth maps can be used to prospect for untested depth leads remaining in the area.

Coarse lithology prediction

It is possible to infer coarse lithology through use of seismic velocity measurements.

An example from Texas shows a seismic velocity speed-up due to a high-velocity channel fill known from well measurements (Figs. 11 [49513 bytes] and 12 [146993 bytes]).

Straightforward method

The method is very straightforward: horizon-oriented seismic velocity measurements coupled with areal interpretation of the measurements to aid in derivation of the velocity model. The depth conversion is made to inspect the area of interest for depth leads. Prestack depth migration using an initial interval velocity model derived from the seismic velocity field may be used to explore further.

Seismic velocities are not vertical velocities and may fail to predict correct depths. This method minimizes measurement errors with horizon seismic velocity scans and focuses on the extraction of velocity trends from the seismic. When there are many seismic lines and wells (and therefore many seismic-seismic ties and seismic-well ties), the redundancy in measurements enables one to get calibrated velocity and depth trends with reasonable accuracy.

The method provides a practical solution to the explorationist even though it may seem simplistic and not rigorous. Interpretation of the seismic velocity areally and along geologic boundaries is a powerful and not necessarily simple requirement.

Acknowledgments

The authors thank management of Shell for permission to publish this article. We thank the management team at Shell Western Exploration & Production Inc. for supporting this work 12 years ago.

Bibliography

Al-Chalabi, M., Velocity Determination from Seismic Reflection Data, Developments in Geophysical Exploration Methods, Vol. 1.

Mastoris, Susan, "3D seismic interpretation techniques define shallow gas reservoirs," OGJ, Jan. 9, 1989, pp. 69-73.

Sheriff, R. E., Geldart, L. P., Exploration Seismology, Vol. 2, Data Processing and Interpretation.

The Authors

Carlos E. Guzman has worked with Shell Oil Co. since 1977. He holds a BS in physics from the University of New Orleans and an MS in physics from Purdue University.

Murali Ramaswamy worked at Shell Oil Co. during 1981-91 and Houston Advanced Research Center in 1991-96. He now works for Exxon Exploration. Ramaswamy holds a BS in electrical engineering from Indian Institute of Technology, an MBA from Indian Institute of Management, and an MS in physics from the University of New Orleans.

Brad K. Wright has worked at Shell Oil Co. since 1965. He holds a BS in electrical engineering from New Mexico State University.

Kevin P. Lawler, now with ARCO, worked with Shell Oil Co. in 1985-94 and Veritas Seismic during 1994-96. He holds BS and MS degrees in geology from Rensselaer Polytechnic.

Copyright 1996 Oil & Gas Journal. All Rights Reserved.