Imaging of the Salt Sediment Interface at the Top of Near-Surface Salt Domes
Presiding: G L Kinsland, University of Louisiana at Lafayette; R V Schneider, University of Louisiana at Lafayette
NS21A-01 08:30h
Imaging Shallow Salt With 3D Refraction Migration
In offshore West Africa, numerous salt walls are within 200 m of sea level. Because of the shallowness of these salt walls, reflections from the salt top can be difficult to map, making it impossible to build an accurate velocity model for subsequent pre-stack depth migration. An accurate definition of salt boundaries is critical to any depth model where salt is present. Unfortunately, when a salt body is very shallow, the reflection from the upper interface can be obscured due to large offsets between the source and near receivers and also due to the interference from multiples and other near-surface noise events. A new method is described using 3D migration of the refraction waveforms which is simplified because of several constraints in the model definition. The azimuth and dip of the refractor is found by imaging with Kirchhoff theory. A Kirchhoff migration is performed where the traveltime values are adjusted to use the CMP refraction traveltime equation. I assume the sediment and salt velocities to be known such that once the image time is specified, then the dip and azimuth of the refraction path can be found. The resulting 3D refraction migrations are in excellent depth agreement with available well control. In addition, the refraction migration time picks of deeper salt events are in agreement with time picks of the same events on the reflection migration.
NS21A-02 08:45h
A High-Resolution Seismic Survey to Image the top of Salt at Avery Island, Louisiana
We performed a shallow high-resolution 2-D seismic reflection study in an attempt to image the top of the salt dome at Avery Island, Louisiana. The survey was conducted using a 48-channel Geoder system with an Elastic Wave Generatorr (accelerated weight drop surface source) as the source. The purpose is to aid in mine planning by Cargill Salt and in shallow subsurface environmental planning by McIlhenny Co. The processing scheme included, 1) conversion to ProMaxr compatible format, 2) geometry establishment using both GPS and topographic map locations, 3) record truncation and resampling in time, 4) ground roll, air wave and refractor mute, 5) filter testing, design and application (dominant frequency is found to be about 100 Hz), 6) velocity analysis and 7) stacking. The line runs from near the mineshaft down the main road to near the water of the surrounding canal near the flank of the dome. The final stack of the data showed five reflectors at the end of the line nearest the flank of the dome, the most significant reflector having stacking velocities over 2,000 m/s. This indicates the possible presence of salt-cemented sediments similar to those found by studies on Weeks Island. No unambiguous "salt reflector" is identified. It is possible that although the salt is present, the salt-clastic sediment interface does not produce the expected strong reflector perhaps due to a gradational contact or, the salt is, in some areas, too near the surface causing the refraction to interfere with the reflection or the interface is too uneven to be imaged with our survey. Surveys and processing of other designs and/or well data might differentiate between these hypotheses.
NS21A-03 INVITED 09:00h
High-Resolution Seismic Investigation of a Surface Collapse Feature at Weeks Island Salt Dome, Louisiana
Seismic imaging techniques delineated the subsurface expression of an active sinkhole above a former salt mine at Weeks Island, Louisiana, which was used at the time by the U.S. Department of Energy's Strategic Petroleum Reserve. (The Weeks Island salt dome is no longer part of the Reserve.) The sinkhole, which at the time of the survey was approximately 12 m wide and 11 m deep, is directly over the edge of the upper storage chamber and approximately 60 m above the top of the salt dome. Surface seismic reflections imaged a dramatic bowl-shaped depression in a 28-m-deep reflector spatially consistent with the sinkhole. Two reflections (28 m and 60 m) on multichannel VSP data represent the only velocity and/or density contrasts detected above the top of the salt dome. The 28-m reflector identified on both VSP and surface seismic reflection data is at a depth consistent with the piezometric surface. Considering the high measured permeability and relative geometric severity of the reflection geometry, it is questionable whether this drape in the 28-m reflection is consistent with the water table. Localized velocity variations could account for some of the apparent geometry. The 60-m salt reflection, evident on VSP, can be interpreted on selected processed surface seismic shot gathers, but is difficult to confidently and consistently identify on stacked sections. The sinkhole lies along a northeast-trending acoustic lineament, possibly related to or associated with salt dissolution. The acoustic expression of the sinkhole suggests a localized, predominantly vertical feature. No evidence was discovered to confidently ascertain the mechanism responsible for exposing the salt to unsaturated meteoric water.
NS21A-04 INVITED 09:15h
Two Dimensional Shallow Imaging Seismic Survey to Determine the Top of Salt at Weeks Island, Louisiana
In the early 1990s we were contracted by Morton Salt Company to perform shallow-imaging, two dimensional seismic surveys in an attempt to determine the clastic sediment to salt interface over a portion of their mine at Weeks Island, Louisiana. Morton Salt Co. was considering mining shallow salt above their present workings and wanted to be assured that the thickness of salt required for safety existed in the area. We performed several tests over the area and collected two lines with an all terrain vehicle (ATV) mounted elastic wave generator (EWG), which is an accelerated weight drop surface seismic source. We also surveyed two test lines using one-third pound charges buried five feet. Shot and receiver spacing varied somewhat amongst the tests however the most important lines were shot with ten foot shot spacings and five foot receiver spacings with 48 channels typically yielding 12 fold data. The field data typically showed strong refraction arrivals but strong reflection arrivals only in some areas. Refraction solutions were achieved for two widely separated lines, one being near the headframe of the salt works and the other being further down over the flank of the dome in the area of main interest. Both refraction solutions indicated what we came to designate as the "7000 ft/s refractor" 50 to 100 feet below the surface. The data from only one EWG line over the area of greatest interest were processed to yield a reflection image. In the areas along this line where strong reflection arrivals were achieved these signals stacked to a strong continuous reflector at about 100 ms two-way travel time. In other areas the reflections were much weaker and discontinuous. The field data from the one-third pound charges did not contain significantly better reflection returns from any subsurface horizons than the EWG data and were not processed further.
NS21A-05 09:30h
Nature of the Salt-Clastic Sediment Interface Over Weeks Island Salt Dome, Louisiana Using Shallow Imaging 2-D Seismic Data and Well Log Data
Shallow-imaging surface-sourced 2-D seismic data on the flank of Weeks Island, Louisiana yielded what has been termed a 7000 ft/s refraction layer 50 to 100 ft below the surface and a strong 100 ms (too shallow to be top of salt) locally continuous reflection as well as areas with no strong reflector. A geophysical well-log near the line of the seismic data indicates that there is another density discontinuity about 130 ft above the salt and the driller's log reported increased chlorides at this level. Near the apex of the dome a shallow-imaging survey performed by a crew from the Kansas Geological Survey and the University of Kansas located a well-defined top of salt reflector. Model calculations and geological arguments indicate that the 7000 ft/s refractor and the strong reflector result from a layer of salt cemented clastic sediments above the top of salt. Dissolution of the salt by groundwater flowing down the flank of Weeks Island disrupts the salt sediment interface creating rugosity that destroys coherent reflections, leads to karstic collapse on the island which breaks up the reflector, has led to the topography of the island and has at times and in places led to deposition of salt within the clastic sediments which has created the reflector. At the top of the dome these processes are not as active, because of the lesser flow of water, and the salt clastic sediment interface remains much better defined. We were not able to image the top of salt in our area of interest. Therefore, Morton Salt Co., for whom we performed this study, remained uncertain of the safety of mining a zone shallower than their existing workings within the dome and did not develop the proposed level.
NS21A-06 09:45h
Velocity Modeling of Land 3-D Surface Seismic Data for Prestack Depth Migration
Since the early 1990's prestack depth migrations (PSDM) have been routinely applied to 3-D seismic data. Using current methods building an accurate velocity model is dependent on the quality of the seismic data. Data quality is a reason why PSDM have been applied nearly exclusively to marine data. To apply the power of PSDM to land 3-D surface seismic data deriving an initial sediment velocity model that overcomes limitations of data quality associated with land 3-D seismic data is essential. In this study I use a non-seismic approach to generate an initial sediment velocity model for PSDM of land 3-D surface seismic data. Differences in data quality between marine and land 3-D seismic data provided the motivation to investigate innovative velocity modeling techniques for land 3-D surface seismic data. In this study current practices primarily for marine data were reviewed and problems with land 3-D surface seismic data were identified. An approach that integrates the efficiency and accuracy of current methods and overcomes the limitations imposed by data quality was implemented. Data were acquired over Vinton Dome in southwest Louisiana and used in this case study to derive an initial sediment velocity model using sonic logs. Comparing the results with a prestack time migration (PSTM) and a PSDM that had been migrated using a seismically derived velocity model demonstrated that this approach enhanced the fidelity of the Vinton Dome 3-D surface seismic data.