Understanding salt tectonics and salt geometry has always been important to petroleum exploration in the Gulf of Mexico (GOM). Initially, prospects were drilled almost exclusively above salt.
Today, more and more prospects are being drilled beneath the upper layer of salt in the deep water. With the high cost of drilling, it is important to properly image the prospective sedimentary layers below salt.
Seismic data acquisition and seismic data processing have improved dramatically over the years.
In most cases in the deepwater GOM, seismic data can image the top of the allochthonous salt. In some cases, often after significant effort and cost, seismic data can also image the base of allochthonous salt and subsalt reflectors.
Still for many areas in the deepwater GOM, the base of allochthonous salt and the top of the deeper autochthonous salt is poorly imaged by seismic. Significant opportunity exists for those companies that can highgrade these subsalt areas for further investigation.
For the purpose of defining the base of salt in those areas poorly imaged by seismic data, Conoco has developed and patented robust 3D inversion methods that utilize gravity and magnetic data to define the volume of salt in the subsurface.
The inversion method can simultaneously invert on gravity, magnetic, and gravity tensor fields, while accounting for lateral and vertical density and susceptibility variations within the sediments.
The input data to the inversion are the bathymetry, top of salt, density function, gravity, gravity tensor, and magnetic fields. The end result includes base of salt prediction from the inversion and salt thickness information.
The inversion method can be applied at the prospect, minibasin, or regional scale. Knowledge of the base of salt provided by the inversion method has improved Conoco's ability to highgrade prospects, reprocess seismic data, and drill better wells in the deepwater subsalt province of the GOM.
This article presents examples of the nil zone, inversion for base of allochthonous salt, inversion for top of autochthonous salt, constrained base of salt inversion, and comments on the value of tensor gravity data.
Fig. 1 is an example of a seismic line where the base of salt is poorly imaged. The top of salt has been well imaged and can be used as input into the gravity and magnetics inversion program.
Performing gravity and magnetics inversion for base of salt prediction can help the geoscientist decide if it is worthwhile to pursue further investigation of subsalt potential in this area. Further investigation can include reprocessing the seismic data with the additional base of salt information provided by the inversion.
Gravity, magnetics inversion
Fig. 2a is an example of a seismic line where both the top and base of allochthonous salt are relatively well imaged. Posted on Fig. 2b are the base of salt prediction from a gravity only data inversion and base of salt prediction from a joint gravity and magnetics data inversion. The base of salt prediction from the combined gravity and magnetics inversion is better than gravity data only inversion in this case because of what is known as the nil zone.
The nil zone is defined as that region wherein the salt and sediment densities are nearly identical. In these regions, salt masses produce insignificant gravitational signals and the base of salt prediction from the gravity data only inversion will be in error.
This is the case in Fig. 2b where the gravity only base of salt prediction diverges from the seismic base of salt on the right hand side of the figure. The nil zone is an example of one of the real world complicating factors that can impact the inversion results.
The primary focus of our gravity and magnetics inversion to date has been for base of allochthonous salt prediction.
Fig. 3a is an example of a seismic line where the base of allochthonous salt is poorly defined seismically and the prediction for base of salt has been posted from the inversion. Once the top and base of allochthonous salt have been defined, an inversion can then be performed to define the top of autochthonous salt.
Fig. 3b shows the same seismic line as Fig. 3a with the addition of the deeper top of autochthonous salt from a subsequent inversion. Knowing the salt volume and geometry can impact many aspects of assessing prospect risk, including seal, trap, hydrocarbon migration, reservoir, and geopressure risk.
The prediction for base of salt can also be constrained in the inversion. Fig. 4a is a seismic image with an incomplete base of salt reflector. A geoscientist could easily interpret a flat base of salt across the line with seismic only data.
Fig. 4b indicates a base of allochthonous salt prediction that has been constrained to the right and left of the keel by the seismic data. Knowledge of the thick keel of salt makes a significant difference in the interpretation of the prospectivity of the subsalt anticline seen on this line.
Gravity tensor data
In addition to the use of gravity and magnetics data, gravity tensor data can be used in the inversion. In the past, gravity data acquisition was limited to measuring the vertical component of the gravity field (gz).
Following the declassification of the gravity gradiometry (tensor) technology by the US Navy, tensor data are now available for oil and gas as well as mineral exploration.1 Gravity gradiometry records the spatial rate of change of the acceleration due to density anomalies in the subsurface. In gravity tensor surveys, the spatial changes of the gravity field are measured in all three directions simultaneously.
This provides a total of nine tensor components of which five components (Txx, Txy, Txz, Tyy, and Tyz) are independent of each other. These five tensor components along with the traditional vertical component gz are the survey data provided by Bell Geospace Inc.
A key value of tensor data lies in the ability to achieve a high level of enhancement of the signal-to-noise ratio for each of the tensor channels and gravity channel. This enhancement is achieved by operating on all channels simultaneously using proprietary techniques developed at Conoco.
The following images (Fig. 5) show the results of Conoco's proprietary tensor multichannel processing method compared to a standard filtering approach. A noticeable improvement in signal coherency is observed.
The better the data that go into the inversion, including bathymetry, top of salt interpretation, density, gravity and magnetics data, the better the probability of a successful, more accurate inversion result.2 3
Conclusions
Exploration efforts in the deep water can be made more cost effective with the use of potential field data. The costs involved in gravity and magnetics data purchase and analysis, weighted against the value of geologic insights that can be obtained, warrant the investment. Salt geometry can be obtained by carrying out simultaneous inverse modeling of the data.
With regard to the inversion process:
- Joint gravity and magnetic inversion can be more accurate than gravity only inversion, especially in the "nil zone."
- Inversion for base of salt prediction can be done with regular gravity and magnetics data.
- Multichannel tensor gravity data help in the pre-processing noise removal stage and also provide extra information not available in traditional gravity data.
- Before the data are inverted it is critical to pre-process the data to reduce data error caused by nongeologic sources.
- The joint inversion of gravity, gravity tensor, and magnetics is recommended for best results.
In this brief article, we have presented several examples of subsalt mapping problems in the deepwater GOM. A more detailed account of case histories from the deepwater GOM can be found in Jorgensen et al.2
This inversion technique for base of salt prediction has been applied successfully both onshore and offshore. The inversion technique has also been applied to top of basement prediction in the North Sea with successful results.
Acknowledgments
The authors thank Conoco management, particularly Alan Huffman and Randy Thompson, for their assistance in preparing this article. We are indebted to Veritas Marine Surveys, WesternGeco, and Bell Geospace for permission to show portions of their data.
References
- Brett, J., Hammond, S., Macfarlane, J., Humphrey, D., Selman, D., Watkins, J., Coburn, G., Mumaw, G., Murphy, C., Stalin, F., and Pritchett, S., "Three-dimensional, full tensor gravity-gradiometer (3D-FTG)," Petroleum Frontiers, Vol. 17, No. 4, 2001, pp. 8-17.
- Jorgensen, G.J., Kisabeth, J.L., and Routh, P.S., "The role of potential field data and joint inverse modeling in the exploration of the deepwater Gulf of Mexico mini-basin province," Petroleum Frontiers, Vol. 17, No. 4, 2001, pp. 18-35.
- Routh, P.S., Jorgensen, G.J., and Kisabeth, J.L., "Base of salt imaging using gravity and tensor gravity data," SEG annual meeting 2001, San Antonio, pp. 1,482-84.
The authors
Partha Routh is a geophysicist in the Seismic Imaging Technology Center of Conoco Inc. His research interests include inversion methods, potential fields and pre-stack inversion. He joined Conoco in 2000 after receiving his integrated MSc in geophysics from IIT, Kharagpur, India, and PhD in geophysics from UBC, Canada.
Greg J. Jorgensen is a senior geophysical advisor in SITC-Conoco. He joined Marathon Oil in 1984 after receiving his MSc in geology from Brigham Young University. He joined Conoco in 1989 as a potential methods specialist.
Jerry L. Kisabeth is a senior research associate in SITC-Conoco. His research interests include potential methods, magnetotellurics, and the polar and equatorial electrojets. He attended Texas A&M University, University of Alaska, and University of Alberta, earning degrees in geophysics, upper atmospheric physics, and space physics.
W. Rex McKinley is leader, seismic analysis group, SITC-Conoco. He worked in Conoco E&P offices in the US Gulf Coast and Norway before transferring to SITC in 1999. He joined Conoco in 1979 after graduating from the University of Texas.