Seismic Attributes - Rock Solid Images



solutions_services_seismic attributes

Seismic Attribute Generation and Classification
Complex seismic trace attributes were introduced around 1970 as useful displays to help interpret the seismic data in a qualitative way. Walsh of Marathon published the first article in the 1971 issue of Geophysics under the title of " Color Sonograms". At the same time Nigel Anstey of Seiscom-Delta had published �Seiscom 1971� and introduced reflection strength and mean frequency. Realizing the potential for extracting useful instantaneous information, Taner, Koehler and Anstey turned their attention to wave propagation and simple harmonic motion. This led to a landmark paper in the June 1979 issue of Geophysics by Taner, Koehler and Sheriff. At that time, amplitude, phase and frequency plus combinations of amplitude and phase, namely weighted average frequency and apparent polarity were first introduced. Today, however, the explorationist is faced with a bewildering choice of hundreds of attributes and the almost insurmountable task of determining which of these attributes may offer greater insights into the petrophysical composition and geometry of the reservoir.

It is our goal at Rock Solid Images to assist today's explorationist by computing and assembling those attributes that are the most significant diagnostic indicators of rock and fluid properties in the reservoir. It is also important that the seismic data input for an attribute generation study has been suitably processed. Without the correct data conditioning in the

Reservoir characterization is increasingly dependent upon high quality seismic data and the attributes derived from it. Attributes may be utilized as a qualitative interpretive adjunct, or, through the adaptation of various statistical procedures integrated with log and core data to provide a powerful tool for the identification of various reservoir quality indices in seismic data.

Physical, Geometric and Hybrid Attributes
Rock Solid Images offers the computation of an extensive suite of prestack and poststack, instantaneous or wavelet-based attributes that fall into three broad classes: physical, geometric and hybrid attributes.

Physical attributes are those that are computed for single traces and which represent measurements made from the seismically-propagated wavefield, e.g. derived amplitudes (RMS, average, etc.), time-thickness, frequency content, attenuation and absorption, complex trace attributes, AVO indicators (intercept, gradient, fluid factor, Poisson's reflectivity, etc.).
Geometric attributes are those that consider multiple or "windowed" traces. This class provides a morphological expression of the subsurface structure and stratigraphy. Examples of this second class include semblance, parallel/chaotic/thin bedding, event termination/edge detection, etc.
Hybrid attributes are computed by implementing a knowledge-based approach. The attributes in this class are those that are derived from both physical and geometric attributes. The motivation for developing this class of attributes is to assemble those attributes that, in certain circumstances, duplicate the human discrimination process. For example, diagnostic characteristics of shales might be considered to be lateral continuity, non-chaotic and thin bedding (therefore of higher spectral content) etc. The converse may be applied to sands. The Sand-Shale indicator attribute is one such hybrid that, in certain settings, has been shown to effectively discriminate between sands and shales.

For further discussion on Seismic Attributes, we advise reading Attributes Revisited. Also a complete list of attributes and depiction of their differences can be explored in the Attrib3D section.

Attribute Generation and Classification Deliverables

Individual attribute volumes
In certain cases, a client may be interested in augmenting the interpretation of a standard amplitude (reflectivity) volume by considering various of the qualitative and quantitative attributes that Rock Solid Images has developed. These are delivered in SEG-Y or .vol formats for straightforward import to interpretation and volume visualization systems.
Clustered "hybrid" attribute volumes
These volumes are generated by utilizing the pattern recognition and nonlinear computational capabilities of artificial neural network (specifically feed-forward and self-organizing feature map implementations). Although un-calibrated to reservoir properties, these clustered "hybrid" attribute volumes can be of significant use to the interpreter particularly in "frontier exploration" scenarios where well control is absent.
Reservoir attribute volumes
Finally, when combined with artificial neural networks and petrophysical data derived from log and core data, Rock Solid Images is able to deliver a fully calibrated reservoir attribute volume. Details of our approach can found in the next services section - Seismic Reservoir Characterization.