Datapages, Inc.Print this page

Quantifying the Temporal and Spatial Extent of Depositional and Structural Elements in 3-D Seismic Data Using Spectral Decomposition and Multi-Attribute Rgb Blending

Leppard, Chris 1; Purves, Stephen 2; Henderson, Jonathan 2; Eckersley, Adam 2
1 StatoilHydro, Bergen, Norway.
2 ffA, Newcastle Upon Tyne, United Kingdom.

Single and multi-trace seismic attributes are widely used to help to visualize and delineate the spatial extent of depositional bodies on interpreted horizons. Commonly, the internal complexities of these interpreted depositional bodies are not as well-resolved, due to spatial resolution constraints and low contrast. This increases the uncertainty of quantitative analyses or inferences that utilise the resulting map, e.g. reservoir presence. This also holds true in the delineatation of structural elements, particularly when attempting to quantify the potential for internal barriers to act as baffles to fluid flow within individual reservoir units. Therefore, there is a need for further development and understanding of seismic attribute workflows used to evaluate geological elements, in order to gain an enhanced understanding of their detailed morphology, spatial extent and temporal location to fully assess their impact on reservoir fluid-flow. The recorded reflected seismic wavelet is the primary source of information used to interpret in the subsurface. The spectral content of the recorded seismic wavelet is dependent on the acoustic properties of the media along its propagation path, and decoding the subtleties in the variation of its spectra can provide a exceptional high-resolution insight into complex geological variations. Advanced spectral decomposition techniques allow the comparison of the 3D variation in the wavelet response at different frequencies through generating and comparing different spectral attributes. The workflow presented here compares and contrasts 3 techniques for extracting the spectral content; constant bandwidth, constant Q and matching pursuit decomposition. In order to visually extract the subtle details and complexity contained within the individual frequency bands, each needs to be examined and compared in a coeval visualization environment. Blending 3 unique frequency volumes mapped individually to an RGB channel allows the interpreter to combine the information provided by each volumes into one single ‘full colour’ 3D display. The resulting RGB blend provides a dramatically enhanced seismic image of both depositional bodies and structural elements than previously obtained through traditional attribute techniques. Quantitative measurements can be extracted either directly from the RGB volumes through the use of innovative opacity based geobody extraction techniques or from more traditional horizon extraction and sculpting.


AAPG Search and Discovery Article #90090©2009 AAPG Annual Convention and Exhibition, Denver, Colorado, June 7-10, 2009