Application of Bandlimited Attributes to Characterization of Fluvial /Deltaic Reservoirs

Adams, Donald; Markus, David
Saddleback Geosolutions, Houston, TX.

The Middle Oligocene Frio sandstone reservoirs of the Texas Gulf Coast are dominated by amalgamated thin fluvial channel and crevasse-splay reservoirs as characterized by those found in Stratton Field, but analogous to reservoir systems that are found throughout Southeast Asia. At Stratton, the field is draped over a broad dome shaped structure in the Vicksburg Flexure Trend above the Vicksburg fault system. The channel sands are 3 to 10 m thick and channel splay deposits are 1.5 to 7 m thick with porosities as high as 25 percent. Based on the log data the individual amalgamated channel fill deposits are up to 850 m wide. Production is challenging given fluvially dominated setting where thin heterogeneous reservoirs cause significant compartmentalization and stranded reserves. Such reservoirs tend to have low ultimate recoveries.

These reservoirs have highly tuned seismic responses, for example, a 10 m thick reservoir at 3080 m/s is 6 ms thick and thus is still tuned at or below a dominant frequency of 83 Hz. The seismic data used in this study consists of a full stack dataset with an effective bandwidth of 10 - 60 Hz and a dominant frequency of 35 Hz which is well below the level needed to resolve individual channel systems through conventional interpretation methods.

To mitigate some of the resolution issues we apply advanced seismic analysis techniques consisting of post-stack seismic processing, attribute analysis and detailed interpretation. Post stack processing includes steps to reduce noise, broaden and balance spectra resulting in a 10 - 80 Hz bandwidth. Attribute analysis of the post-stack processed data includes calculation of bandlimited instantaneous attributes from a high-fidelity spectral decomposition for both amplitude and phase, a unique tuned similarity, volumetric curvature and other attributes. The results are calibrated to well data using rock physics relations to select key attributes for statistical analysis and time-varying waveform characterization. The results are interpreted in detail using stratal slices and well data for channel related stratigraphic features.

The results show decreased tuning thickness and enhancement of visible channel geometries compared to the original data.

AAPG Search and Discovery Article #90155©2012 AAPG International Conference & Exhibition, Singapore, 16-19 September 2012