Spice is a seismic interpretation attribute developed in 2002 by C.-F. Li and C. L. Liner, and a patented invention owned by The University of Tulsa. It has been further developed and marketed to the petroleum seismic industry through a commercial partnership with Fairfield Industries.
Imagine a well with sonic and density logs from which we can construct an acoustic impedance (AI) curve, and from this a reflection coefficient (RC) series, and then convolve with a seismic wavelet to form a synthetic seismogram (SYN). Spice has the following remarkable property: Spice curves generated from AI, RC, and SYN have the same shape. And this is true for different seismic wavelets. In a sense, Spice looks through the wavelet and reflection coefficient series to form a shadow image of the underlying geological layering.
A distant cousin of spectral decomposition, the physical basis of Spice relates to spectral shaping during the reflection process, local transmission losses, and thin bed effects. Unlike spectral decomposition, there is no data proliferation in Spice. It operates in such a way that one trace in gives one trace out. A migrated data cube generates a Spice cube of the same dimensions.
Imagine a well with sonic and density logs from which we can construct an acoustic impedance (AI) curve, and from this a reflection coefficient (RC) series, and then convolve with a seismic wavelet to form a synthetic seismogram (SYN). Spice has the following remarkable property: Spice curves generated from AI, RC, and SYN have the same shape. And this is true for different seismic wavelets. In a sense, Spice looks through the wavelet and reflection coefficient series to form a shadow image of the underlying geological layering.
A distant cousin of spectral decomposition, the physical basis of Spice relates to spectral shaping during the reflection process, local transmission losses, and thin bed effects. Unlike spectral decomposition, there is no data proliferation in Spice. It operates in such a way that one trace in gives one trace out. A migrated data cube generates a Spice cube of the same dimensions.
(click for full size version)
Figure A shows a panel of migrated data
from a 2D seismic line. These data are from a soft, clastic basin
in Southeast Asia, and are representative of good-quality marine seismic
data world-wide.
(click for full size version)
Figure B is the Spice section computed
directly from the migration data. No velocity information is required
to generate this result. Borehole data can calibrate and validate the
Spice result, but is not required to create it.
Intriguing stratigraphic relationships can be seen throughout the Spice result. These correspond to exceedingly subtle features in the original data. As an example, consider the small region between 3-4 km and 1.2-1.3 seconds. In the migration data we see amplitude fluctuations and a bed termination (onlap) up and to the right. The Spice image localizes the terminating bed and additionally shows the unit above downcutting to form the termination. Interestingly, stratigraphic intervals are often seen on Spice images to have a bias that associate with interpreted sequence boundaries. This effect is seen as net brightening or darkening of different stratigraphic intervals.
This example is a small part of one 2D seismic line. Imagine the vast amount of stratigraphic information this attribute could pull from a 3D data set. An obvious comparison is the stratigraphic detail seen in coherence time or horizon slices. But coherence does not deliver the kind of vertical stratigraphic image seen in Spice. Spice also offers a different kind of detail complimentary to state-of-the-art impedance inversion.
Basin floor fans, turbidite deposits, fluvial systems, carbonates, and other stratigraphic targets are of increasing interest in global seismic search for petroleum. Spice will play a role in this search.
Intriguing stratigraphic relationships can be seen throughout the Spice result. These correspond to exceedingly subtle features in the original data. As an example, consider the small region between 3-4 km and 1.2-1.3 seconds. In the migration data we see amplitude fluctuations and a bed termination (onlap) up and to the right. The Spice image localizes the terminating bed and additionally shows the unit above downcutting to form the termination. Interestingly, stratigraphic intervals are often seen on Spice images to have a bias that associate with interpreted sequence boundaries. This effect is seen as net brightening or darkening of different stratigraphic intervals.
This example is a small part of one 2D seismic line. Imagine the vast amount of stratigraphic information this attribute could pull from a 3D data set. An obvious comparison is the stratigraphic detail seen in coherence time or horizon slices. But coherence does not deliver the kind of vertical stratigraphic image seen in Spice. Spice also offers a different kind of detail complimentary to state-of-the-art impedance inversion.
Basin floor fans, turbidite deposits, fluvial systems, carbonates, and other stratigraphic targets are of increasing interest in global seismic search for petroleum. Spice will play a role in this search.