Contributors: Supratik Sarkar
While seismic processing shops usually are the “professionals” when it comes to velocity analysis, the interpreter may have greater insight into the cause of the poor data quality.
In such cases, re-picking the velocities on a tight grid with a better knowledge of the geology can provide significant improvement.
The following example (figure 1) is from a tight turbidite formation in Mexico’s Chicontepec Basin, where the data are often handicapped by interbed multiples from overlying shallow volcanic sills.
While the deeper, non-targeted Cretaceous horizons were well resolved – appearing as coherent, broad-band reflectors – the shallower Eocene/Paleocene Chicontepec Formation reflectors of interest were “wormy” and narrow-band.
The original data were prestack time-migrated using a Kirchhoff algorithm, resulting in 50-meter offset bins ranging between 50 and 3,000 meters. The preliminary interpretation of the stacked data showed significant reduction of data quality below the volcanics.
Maps of the volcanics were made to highlight the problem areas.
The original migration velocities were then removed (figure 1b) using simple reverse normal moveout, followed by a velocity analysis on a dense 375-meter x 375-meter grid.
Seismic processors will recognize this workflow as comprising the key steps within a typical processing sequence – using simple software tools to generate a residual velocity analysis of the previously migrated data.
The key step was to take care beneath the volcanics to pick the slower Chicontepec reflectors rather than the strong, slightly faster interbed multiples (figure 1a), resulting in the flattened gathers like the one shown in figure 1c.
After the residual velocity analysis, the frequency content and the signal-to-noise ratio of the dataset improved in many places within the Chicontepec interval compared to the original prestack time-migrated data.
This residual velocity analysis resulted in better vertical and lateral definition of stratal units as seen by seismic amplitude as well as computed attributes (figure 2).
This velocity analysis also helped enormously to delineate turbidite channels, internal stratal geometries and the distribution of potential reservoir element.
Several volcanic activities in east-central Mexico from the Late Cretaceous to Miocene have been published. The burial history chart from the adjacent Veracruz Basin – where petroleum generation occurred from Upper Jurassic source rocks similar to those in the Chicontepec play – shows that oil generation and migration started around 20 Ma and continued through the Miocene.
Since most of the oil migrated after the major volcanic activities, improving the imaging below the volcanics also provided us ideas how the volcanic bodies enhanced fracture porosities in reservoirs in some areas – specially through enhanced seismic attribute images.
I wish to thank my adviser, AAPG member Kurt J. Marfurt from the University of Oklahoma, for his constant persistence and encouragement (to a geologist) to complete the work, part of which is reflected in this article.
I also would like to thank Pemex, as well as Sergio Chavez Perez from IMP, who provided the data for working this project. I am thankful to AASPI consortium at OU for all the support.