A hybrid inversion approach to determine base of salt from gravity data
April 14, 2015
Fairmont Singapore & Swissôtel The Stamford, Hullet Room, Level 4
Singapore | Website
Presenter: Ash Johnson, Geosoft Inc.
Co-authors: Ash Johnson, Robert Ellis and Gaud Pouliquen, Geosoft Inc.
This talk is part of the EAGE Worshop on Non-Seismic Methods. To attend this event, please register through the EAGE event registration page.
Sub-salt – stratigraphically younger rock beneath allocthonous salt layers – has become an important exploration and production play in oil and gas. Accurate representation of base of salt and sub-salt environments has a major impact on selection of drilling locations in these environments. Interpreting top-of-salt structures is a relatively simple process due to the low density and high seismic wave velocity of salt, which generates a strong seismic response. The challenge is to define the base of salt and sub-salt, which are poorly imaged by seismic methods, where significant hydrocarbon traps are formed. Gravity inversion and modelling techniques can improve base of salt/sub-salt imaging where interpretation based on seismic data alone is challenging or impossible. These techniques help to eliminate uncertainty during the interpretation and target selection process, reducing risk and cost when advancing oil and gas exploration programs in sub-salt regions.
Existing modelling and inversion techniques typically use one of the three most widely used data structures used for representing 3D models: grids, triangulated surfaces, and 3D voxels. Grids have obvious advantages for representing single-valued surfaces but are difficult to use for representing multi-valued surfaces such as the boundary of a complicated salt body. Triangulated surfaces work well for representing complex bodies of constant density but are clearly inefficient for representing large complicated volumes that vary in all 3 dimensions. Voxels represent this latter situation well but suffer from large size, increased calculation expense, and quantization errors (i.e., limitations on the vertical size), particularly near the observation location.
The base of salt problem can be improved by taking the best of layered-earth modeling and voxel-based inversion in a hybrid model to allow for the full range of geometries associated with salt/sub-salt structures. The use of cloud-based inversion in this process also allows for effective hybrid calculation techniques to do joint inversion on layers and voxels by combining the advantages of space-domain and frequency-domain techniques. We will illustrate these advances using gravity and gravity gradient data over complex salt structures found in the Gulf of Mexico using the SEAM synthetic model.
Our team will be attending the Seapex 2015 Exhibition and Conference that follows the EAGE Workshop on Non-Seismic Methods. To arrange a meeting time with us during the show, please contact Chris Bishop.