A new approach to seismic interpretation improves resolution and reduces risk.

Spectral decomposition is a simple yet robust approach for generating high-resolution seismic images of stratigraphic and structural prospects or reservoirs. The images are based on amplitude response, the most powerful seismic attribute, yet these unique images provide far greater resolution of reservoir boundaries, reservoir heterogeneities and thickness than is possible with traditional broadband seismic displays. Characteristics of spectral decomposition images typically relate more directly to reservoir heterogeneity than complex trace attributes.
In terms of detail at the reservoir level, spectral decomposition images can be astounding and create a compelling new interpretation method for understanding complex stratigraphic and structural targets. Innovative spectral decomposition interpretation techniques quickly provide additional information about the target. Combining or animating a selected series of spectral decomposition images yields a more certain interpretation of real stratigraphic properties wherever amplitude can be used to infer reservoir potential. An unprecedented understanding of stratigraphic facies boundaries, structural and stratigraphic depositional controls and reservoir geometry is made possible by this technique.
Spectral decomposition leverages studies in demonstrating the effect of changing bed thickness on amplitude response. Where the top and bottom separation of an event thins to approach a quarter wavelength, amplitude response reaches its maximum or tuning thickness. For any given dominant frequency, tuning occurs at a different thickness. Spectral decomposition exploits the tuning phenomenon by generating and combining amplitude response maps from a range of discrete frequencies. It allows amplitude variations at frequencies greater and less than the tuning thickness to be tied to changing thickness, thereby removing the need to reference back to the conventional quarter wavelength approximation.
Reducing uncertainty
The quest for upstream value creation continues to be a challenge. On the one side, former strongholds of exploration and production activity, like the North Sea and the Gulf of Mexico shelf, are in decline. On the other side, new opportunities exist in complex and harsh environments like deep water. Faced with these technical issues, exploration and production companies are under increased pressure to improve upstream performance. In this type of environment, where a delayed or wrong decision or a poorly placed well can cost millions of dollars, successful companies are continuously leveraging new 3-D technologies that generate better reservoir images, reduce uncertainty, increase reserve replacement rates and lower finding and development cost.
"We are drilling in extremely challenging areas that require substantial financial investments. The cost of field development and production facilities can total well over (US) $1 billion," said Craig Cooper, BP's manager of imaging technology. "It is paramount that we optimize the location and design of every well in order to achieve the maximum production at the lowest overall cost. Ideally we are striving to drill no dry holes."
In many areas, the results of spectral decomposition imaging, integrated with other available data, are allowing companies to realistically approach this goal. Spectral decomposition images can significantly reduce drilling uncertainty for many opportunities and greatly assist the subsurface team in maximizing asset value by optimizing well placement and supporting critical decisions earlier in the project.
Imaging the details
Typically, a geoscientist will make a structure map followed by an amplitude map. Most reservoirs are thin with respect to the typical seismic wavelet. In these situations, subsequent spectral decomposition imaging adds great value to this basic workflow. By using spectral decomposition technology, asset teams are no longer restricted to interpretations from a single broadband amplitude response to the reservoir. The advantages of spectral decomposition imaging technology can be likened to the power of remote Landsat satellite imaging, where discrete bands of infrared and visible wavelengths are used to uniquely distinguish subtle features on the Earth's surface.
Explorationists can identify many parallels between spectral decomposition imaging and proven Landsat multispectral imagery (MSI) techniques. Using remote MSI techniques, a different spectral signature is recorded for each surface feature. The narrower the wavelength band, the greater the sensitivity to subtle differences between ground cover types, surface geology and mineralogy, or cultural features on the Earth's surface. By analyzing images from discrete wavelengths, it becomes possible to distinguish between features that would otherwise have no unique signature or would go undetected.
Similarly, spectral decomposition imaging extracts a suite of amplitude maps from a range of frequency slices in the reservoir zone. Each map illuminates peak amplitude response from a different bed thickness. By selectively combining images from certain frequencies, unique geologic relationships are portrayed, which provides more direct correlation with reservoir heterogeneities. Benefits for reservoir characterization and well placement are immediately apparent, particularly for thinly layered reservoir units.
Interpretation is a visual task, and spectral decomposition interpretation methods leverage the fact that humans are wired to detect movement. Motion is one of the most powerful tools an interpreter can use to extract value from a seismic survey. Interpreters are familiar with using motion, often moving through 3-D volumes to detect subtle changes difficult to detect in static seismic displays. Spectral decomposition imaging employs motion in a unique way. By animating amplitude maps generated from a range of frequency bands, interpreters can exploit motion to unravel complex variability in reservoir heterogeneity and thickness.
Asset team collaboration
The visual impact and detail of spectral decomposition images naturally attracts the attention of subsurface asset teams. When viewing images of a complex reservoir, geologists, geophysicists and reservoir and production engineers can see immediate impact on their respective disciplines, but more importantly, the team can quickly and collaboratively realize their individual contributions and their impact on the project as a whole. Collaboration can occur earlier in the project cycle, leading to improved communication and ultimately to better decisions.
"At BP, spectral decomposition is more than an attribute; it is an interpretation approach for the asset team" said Cooper. "Many 3-D technologies have the potential to provide a great deal of valuable information about the reservoir target, but few offer so much information in so little time." Finer-scale reservoir detail is possible only with much more complex modeling processes and the expertise of specialists. Spectral decomposition images are related to physical reservoir properties more directly than many attributes, so the results can be quickly understood by the entire subsurface team. The new high-resolution information can be integrated into a project quickly.
Exploration and development
Although spectral decomposition is a relatively new technology, it has a successful track record in early exploration and appraisal scenarios as well as in complex development programs. Using these robust subsurface images and animated interpretation techniques, geoscientists are reducing uncertainties inherent to complex reservoir settings, making more confident decisions, compressing development schedules and affecting the bottom line positively.
Recently, an Apache geoscientist mapped a promising 3-D seismic data amplitude anomaly early in the exploration phase. Depositional controls and trapping mechanisms were unclear from static views of the conventional data, leaving the team to question the potential of the prospect and how best to appraise it. Spectral decomposition images not only confirmed the amplitude anomaly but also illuminated something much subtler that had not been seen on the traditional amplitude map.
By animating through discrete spectral decomposition images, a crucial piece of the puzzle was exposed. A previously unseen stratigraphic feature was clearly associated with the amplitude anomaly. Spectral decomposition images had illuminated the thin stratigraphic trapping sequence where a conventional amplitude map had not. Armed with a new understanding of this potential reservoir, the team was better prepared to exploit this area.
Apache asset teams also are applying the technology in development scenarios. "If you are spending $3 million on a well and trying to hit the sweet spot, then having more detail of thick and thin stratigraphic features is extremely valuable," said Craig Jarchow, a geoscientist and Apache's director of information technology and e-business. "Spectral decomposition imaging is highly beneficial for delineating the thickest pay, which immediately impacts the bottom line."
Although the conventional 3-D amplitude data in an onshore dataset showed reasonable detail of the shape of this onshore reservoir, spectral decomposition images clearly illuminated the thickest and thinnest sequences in the reservoir. Amplitude maps from certain frequencies even showed the thinning of levies, helping interpreters map detailed reservoir geometry. Well control confirmed the interpretation.
BP has leveraged spectral decomposition to address reservoir connectivity in the appraisal phase of a deepwater turbidite oil discovery offshore West Africa. The asset team, evaluating this expensive development, identified flow unit architecture and reservoir connectivity as important uncertainty factors. The objective of this work was to guide the placement of appraisal wells and build simulation models that would predict reservoir performance under different development scenarios.
Spectral decomposition animations highlighted systematic changes in stratigraphic architecture vertically through the reservoir, confirmed by draping images of discrete frequency bands over the appropriate model layers. Channel edges were verified and mapped in 3-D. In addition to stratigraphy, spectral decomposition also was used to verify previously seen faults and map several small faults that could be seen for the first time in the lower part of the reservoir. "Integration of spectral decomposition into our geologic and reservoir models has greatly increased confidence in our development plans for this important deepwater reservoir," said Jim Lantz, BP geoscientist (Figure 1).
Clearly, spectral decomposition imaging and animation interpretation techniques hold great promise for unraveling complex reservoirs if amplitude can be used to distinguish reservoir presence. The keys to its success are the high-resolution images, animation to detect subtle features that are not obvious in a static view and integration of the results with subsurface team workflows.

Bibliography
Kallweit, R.S. and Wood, L.C.: "The Limits of Resolution of Zero-phase Wavelets," Geophysics, pp. 1035-1046, v. 47.
Partyka, Greg, Gridley, James and Lopez, John: "Interpretational Applications of Spectral Decomposition in Reservoir Characterization," The Leading Edge, pp. 353-360, March 1999.
Widess, M.B.: "How thin is a thin bed?" Geophysics, pp. 1176-1180, v. 38, 1973.

Comments