Understanding seismic reflectors within shallow tabular salt

Distribution, origin in the Gulf of Mexico

Jul 1st, 2000
Th Bos1

The effect of BOS mis-interpretation on subsalt strata.
Click here to enlarge image

Examination of a 3,600-sq-km 3D seismic volume along the west central portion of the Walker Ridge OCS area (Gulf of Mexico) reveals many seismic reflectors internal to shallow allochthonous trans-ported salt sheets which can be misidentified as the base-of-salt surface on time seismic data and post-stack depth migrated data.

By mis-interpreting these events in the pre-stack depth migration (PSDM) modeling process, sub-salt seismic structures may be incorrectly positioned, changing the amount of closure and thus affecting geologic and economic risk assessments. For the most part, these seismic reflectors are only locally continuous and occur near the base of large allochthonous salt masses and in association with high-angle rises in the base-of-salt surface (BOS).

While many of these internal reflectors are seemingly randomly distributed and are of limited lateral extent, several examples are extensive enough to warrant further investigation. The use of 3D PSDM data clearly identifies these events and suggests a more complex lateral extrusion history.

Location of observed internal reflectors (blue polygon) covering approximately 55 sq km on base-of-salt (BOS) surface. Orientation represents southeastward viewing oblique perspective; approximate BOS area is 80 km wide and 50 km long. Illumination angles and colors are displayed in insert.
Click here to enlarge image

Data processing classes generated from a "top-down" PSDM method as well as 3D visualization techniques (Liro et al., 2000) are used to distinguish these internal seismic reflectors (IRs) from the true BOS surface. These data include 3D post-stack depth migrated (PoSDM) sediment and salt flood volumes and target oriented PSDM lines. Seismic character and velocity "pull-up" and "push-down" effects in the subsalt strata are iteratively tested, compared and corrected in the velocity model-building process before finalizing the final salt model.


(A) Pre-stack time migrated seismic line displaying mis-leading appearance of an internal reflector (B) PoSDM line (same as A) extracted from 3D volume generated in the "top-down" pre-stack depth migrated (PSDM) model building process;(C) PSDM line (same as A and B) illustrating correct positioning of subsalt strata, confirming correct interpretation of BOS surface.
Click here to enlarge image

The most extensive and coherent example of these internal seismic reflectors occurs along the southwestern portion of the data area. Utilizing only pre-stack time migrated data the internal reflectors can be easily mistaken as the true BOS due to their proximity to, and parallel orientation with the BOS. Since sand/shale velocities are approximately 1/3-1/2 of salt velocities at these depths, mis-interpreting this internal reflector as the BOS surface would create depth imaging distortions, artificially "pulling-up" the subsalt strata by several hundred meters, thus creating or enhancing a false anticlinal structure. The magnitude of structural discrepancy would significantly affect economic evaluations in high-risk, deepwater rank exploration areas such as Walker Ridge. Hence, each viable base-of-salt interpretation is tested to weigh its impact on subsalt structural imaging. The final interpretation seeks to optimize the integrity of the subsalt image by way of an accurate, geophysically and geologically sound PSDM velocity model.

Several criteria have been established to distinguish internal seismic reflectors from the true base-of-salt. These criteria are applied primarily to stacked data as part of the model building process (Liro et al., 2000) however, the distorting effect of invoking insufficient salt in the velocity model can be seen on depth gathers as well. If the internal reflectors are interpreted as the base-of-salt, the velocity model below the incorrect BOS is seriously flawed. The seismic events below these features are not their true NMO (normal move-out) position and therefore are too shallow in the depth domain. This implies the model velocities are too slow throughout the interval between the internal reflector and the true base-salt surface. The polarity of these events is also important. The internal reflectors do not change polarity throughout the data volumes, are consistent between the different processing classes of data, and have the same polarity as the BOS, which is another attribute that can lead to BOS misinterpretations.

Various visualization techniques such as sub-volume detection and opacity decimation allow us to quickly and thoroughly characterize internal salt reflectors. We note that they are approximately parallel to the true base salt surface, are always present in the lowest 1/3 of the salt body and occur near high-angle rises in the base salt surface. Internal salt reflectors can be extensive; the example, shown in the first figure, covers approximately 55 sq km, exceeding the surface area of two OCS blocks.

Three interpretations

We offer three possible interpretations of these base-sub-parallel internal reflectors:

  1. They may represent earlier tops-of-salt in a multiple-intrusion-event salt feature. The internal reflectors record sediments deposited over the top of an earlier lateral salt extrusion event, where they reached sufficient thickness to be seismically resolvable. These sediments, or carapace, were subsequently overridden by a later salt extrusion event. The resultant internal reflectors are regionally discontinuous due to the nature of salt carapace deposition combined with carapace removal during salt overthrusting. Where no carapace was deposited or preserved, the lack of seismic reflection coefficient at the salt-on-salt contact gives the appearance of a vertically continuous salt mass.
  2. These sub-parallel internal reflectors may represent sediments that were sheared off at topographically high salt-sediment interfaces and are subsequently entrained within the salt mass as the salt and suprasalt materials move basinward. Thus, the BOS acts as a decollement surface and sediment entrainment occurs as a result of subsalt scouring, as similarly observed at the base of ice glaciers. These seismic reflectors from the entrained sediments occur downslope from the proposed sediment source and are typically observed in positions no shallower than the shallowest (highest) BOS surrounding the features.

Both hypotheses attribute internal seismic reflectors to the presence of sedimentary inclusions within thick allochthonous salt sheets. Analogous evidence of entrained sediments exists from borehole data at Mahogany and Mesquite in the Ship Shoal and Vermilion, Gulf of Mexico OCS areas, respectively. Moore et al. (1995), using sidewall cores and FMI/FMS logs from several wells in these areas, encountered early-mid-Tertiary aged compacted sediments within thick salt masses.

These sedimentary inclusions range in thickness from a few inches up to 88-ft thick. Salt mine observations by Kupfer (1990) also describe a classification of anomalous zones and beds within salt structures. Estimations from the internal seismic reflectors in the Walker Ridge dataset yield sedimentary thicknesses of up 150 ft, well within the magnitude of inclusion thicknesses encountered in the field analogies.

  1. A third interpretation of these reflectors could be that the base-of-salt paralleling internal events represent anisotropic properties of deforming salt. Certain orientations of salt movement may invoke crystallographic deformation within the moving salt mass that will preferentially align salt crystals. The stress inducing the crystallographic distortion is maintained at the time of seismic acquisition since the allochthonous salt in the Gulf of Mexico continues to move towards the present-day Sigsbee Escarpment. Hence, crystals within the re-alignment interval do not restore to their original shape and internal reflectors are imaged when the zone of re-crystallization is thick enough to be seismically resolvable. In-situ anisotropy measurements made at the Mahogany oil field (Raymer et al., 1999) were sourced from sonic logs, VSP's and simple seismic models. At Mahogany, azimuthal variations were noted from all three sources, thus confirming the existence of salt anisotropy in this field analogy.

In summary, the advent of regional 3D data volumes and high-resolution seismic imaging allows detailed definition of the regionally extensive salt features in the deepwater Gulf of Mexico. Identification of seismic reflectors internal to salt is critical to subsalt imaging efforts. When correctly identified, these events indicate that the lateral emplacement history in the deepwater Gulf of Mexico is more complicated than simple lateral extrusion models, requiring more careful attention to seismic interpretation of lateral salt features in order to develop an accurate subsalt image. ;


The authors thank Veritas Marine Surveys for permission to publish this paper. Discussions with Dan Knupp, Leslie Hester, Jerry Young, Sashi Devan, Robert Hobbs, Mark Murphy, Bill Skinner, Julie Oliver, David Brookes, Xiaoguang Meng, and Ruth Kurian contributed to this paper. We also wish to thank Mark Rowan of Rowan Structural Consulting for technical input. A complete list of references is available from the authors.

Distinguishing between salt internal reflectors and true base-of-salt surface

  • Internal reflectors are discontinuous
  • Base-of-salt (BOS) event is a more regionally continuous event
  • At the terminus of an internal reflector, the underlying seismic facies are continuous with the adjacent salt seismic facies
  • Targeted PSDM lines and PoSDM volumes do not image sedimentary reflectors below an internal reflector
  • Use of an internal reflector as BOS event results in seismic "pull ups," suggesting an inadequate salt volume in model
  • Velocity analyses suggest that the material between internal reflectors and BOS events is salt rather than sediment

More in Geosciences