A Report on the Renaissance of Gravity in the Deep Water Gulf of Mexico
A Practical View of Integration Methods

Brian S. Anderson, Mark E. Weber, and John E. Bain

In recent years there has been a dramatic increase in the addition of high-resolution gravity to seismic surveys all over the world. The reason behind this increase is improvements in gravity data resolution, increasingly difficult seismic imaging challenges, and better tools and technologies for incorporating gravity into the seismic workstation interpretation environment. Using a project workflow outline and case study from the Deep Water Gulf of Mexico, we present a generalized approach for state of the art gravity application for the Gulf of Mexico explorationist.

Introduction

Even with the best quality 3D seismic data, an interpreter can have a troublesome task in defining the salt/sediment boundary at the flanks of a salt dome, salt sheet, or other complex structure. For decades, gravity has been used in the Gulf to address this problem. The major differences in how it was done then and how it is now done are twofold: a) better acquisition technology and processed data, and b) truly integrated workstation software tools. By incorporating a co-recorded data set, with each data set independently measuring a related property of the subsurface, the interpreter can place a much higher degree of confidence in the final geologic interpretation. To quantify this observation, case studies show that incorporating 3D seismic with high resolution gravity and magnetics can alter the base of salt interpretation by several thousand feet from the 3D seismic interpretation alone. In some cases, results from gravity modeling have provided excellent insights into the geology below a salt body, enabling the seismic processors to refine their migration velocity model for the structure, and as a result, refine the seismic image through re-migrating the data using the new velocity model.

Advances in Resolving Power

When asked the question: "Why do we re-acquire gravity?" our answer to the question must be: "For the same reasons we re-acquire seismic data." Although the gravity fields mapped in prior years have not changed, our ability to accurately measure and process gravity on a ship has improved dramatically, just as we have improved our ability to shoot, record, and process seismic data.
Recent advances in gravity measurement at sea include:

                     Upgrading from analog to digital control and acquisition systems

                     Higher data sampling and recording rates (200 Hz sampling, 1 Hz recording)

                     Precise DGPS positioning for removal of ship accelerations

                     More accurate measurements of water depth

                     New data processing developments (signal to noise enhancement, micro-levelling, etc.)

With these advances, industry has seen stunning improvements over data recorded as recently as ten years ago. In many cases, there is an increase of up to ten times the data per unit area in new surveys over older data, with a correspondingly higher level of confidence in interpreted geological results. Many operators are routinely incorporating new high resolution gravity into their interpretation projects, particularly in the Deep Water Gulf of Mexico.

Team-oriented Exploration Tools

With the trend towards highly focused exploration teams, the smooth interaction and coupling of multiple geophysical disciplines is essential. Explorationists are expected to employ and be familiar with more disciplines on a continuing basis. The development of workstation applications which enable the interpreter to simultaneously refine the subsurface model using seismic, gravity, and magnetic data has been a giant step forward

Sample Project Work Flow

Using a new software tool kit which has been developed by a consortium of oil exploration companies, seismic contractors, and a gravity and magnetic contracting company, high resolution gravity is now applied to an increasing number of seismic velocity modeling projects. This technique is employed using the following procedure which has been outlined in a simplified version.

                     High resolution gravity is recorded and processed along with the 2D or 3D seismic survey. Present techniques allow for delivery                         of processed gravity data in advance of, or in parallel with processed seismic data delivery.

                     The seismic velocity data is used to create a corresponding density section (or volume, in the 3D case) by means of a flexible                         velocity - density conversion tool kit, incorporating:

                           Gardner's Equation

                           Nafe/Drake, Hilterman, and other density - velocity relationships

                           Use of available empirical data (e.g. velocity logs, check shot surveys, gamma-gamma density logs, etc.)

                           User defined conversion algorithms or formulae

                           Other approaches

                     The density model can be as simple or as elaborate as the corresponding velocity model - up to and including a discrete value                         of density for each x-y-z node within the profile or volume of data.

                     Input of digital horizon data (again, 2D or 3D) as interpreted on the Landmark, Geoquest, or other seismic workstation. The                         system incorporates a "universal translator" for the conversion of one type of horizon to another to accommodate company                         partner teams, etc.

                     Computation of the gravity field of the model, input of gravity data as recorded on the survey, and a direct comparison between                         the two fields.

                     Manipulation of the model using both forward modeling and inversion processes based on minimizing the misfits between                         model and measured gravity fields.

                     On completion of the modeling and/or inversion process, the revised earth model is converted into the velocity domain, providing                         an improved starting point velocity model for depth migration.

                     This iterative process and feedback loop continues throughout the seismic migration and interpretation process.

Gulf of Mexico Example

Typically, a full three dimensional model of a salt feature is used in the Gulf of Mexico. The density cube is derived from available well control. The top of salt is typically obtained from a simple initial stretch to depth from the time interpretation. Later in the interpretive processing sequence, this is updated with the Post-Stack or Pre-Stack Depth Migration results. The base of salt is input from an initial time interpretation. In many cases the initial base of salt interpretation is provided with confidence factors, e.g. a 10 might be assigned to high seismic confidence areas, a 0 being assigned to seismic blind zones, and grades in between. The gravity modeling can then be constrained by the high seismic confidence areas, and the low (seismic) confidence areas are then of most interest in the search for a better interpretation using gravity modeling results.

The density and velocity data are analyzed, typically using cross plots, and a function is derived to convert between the density and the velocity volumes. The gravity effect of the density volume is computed and compared with the observed gravity data, and the differences are resolved through a series of automated structural and density inversion techniques. The final model should contain as much seismic - gravity constraint as possible for optimal results, often involving close interaction between the gravity interpreter and seismic interpreter at the same workstation.

Once the final density model is constructed, the density - velocity function is used to translate the alterations into an apparent velocity cube for incorporation into the seismic depth migration process.

This process, in addition to providing important and independent corroboration and improvement to the seismic interpretation of the base of salt, also provides an important source of long wavelength velocity information beneath the salt masses. This information, when injected back into the velocity model used for producing the final base salt and sub- salt images, can have a dramatic impact on the enhanced quality of the seismic processing results. Figures 1 & 2 illustrate, in part, the impact of gravity constraints on seismic velocity models used in depth migration.

Figure 1: Pre-Gravity Modeling Pre-Stack Depth Migration, Deep Water Gulf of Mexico (Data and migration courtesy of Geco-Prakla and TGS-NOPEC).	Figure 2: Post-Gravity Modeling Pre-Stack Depth Migration, Deep Water Gulf of Mexico (Data and migration courtesy of Geco-Prakla and TGS-NOPEC).

Economic Impact

The approach above breaks with the traditional approach to the use of gravity and magnetic data in oil exploration. In years past, the in-house gravity expert or consultant would disappear with all the required data for the modeling process and return to the client with his or her interpretation. In today's team-oriented exploration environment, the availability and use of real-time interpretation software tools allow for the integration of gravity and magnetic data at the same workstation. This approach is now embraced by a growing number of oil companies for increasing confidence in their geologic interpretations, and decreasing risk. To be most effective, the integration of gravity and magnetics must take place at the earliest stage of prospect development, and can continue throughout the exploration process.

Conclusions

Technical and commercial considerations now indicate that all the available geophysical data be incorporated into the depth migration process in order to minimize both time and risk. The application of gravity to seismic velocity and seismic imaging ambiguities in the Gulf of Mexico has now become a proven and widely accepted technique, in use by a growing number of exploration companies.

Acknowledgements

We would like to acknowledge the work of Jeff Rutledge, Greg Johnson, and the Geco-Prakla Depth Imaging Group for the pre-stack depth migrated seismic data, and their direction on seismic imaging considerations. We would like to also thank TGS-NOPEC for access to the Phase 45 seismic data for this presentation.