Figure Captions

	Figure 1. A HorizonCube consists of a set of auto-tracked seismic horizons using a pre-computed dip volume. This is an exponent of an emerging group of global interpretation techniques.
	Figure 2. A deepwater channel system of the Australian Northwest Shelf is sub-divided into several channel stages attribute volume. Top: Color-coded channel stages are overlain on the seismic data. Bottom: Time-slicing through the channel stages in a flattened domain (channels top).
	Figure 3. Same channel systems are visualized using HorizonCube density attribute. Note that the locations of higher densities values mostly correspond to channel base, levees or over-bank regions.

New Family of Stratigraphic Attributes

For the continuous set of seismic horizons, a new family of attributes can be computed that visualize geologic features previously hidden. Attributes include:

Isochron thickness – This attribute highlights not only sedimentary bodies but also picks up local pinchouts, condensed intervals and local unconformities.

Curvature – This curvature attribute tends to be smoother than the conventional volume curvature attribute, computed using the seismic dip volume without mapping.

HorizonCube density – Events can be counted within user-defined time windows, with high-density values corresponding to horizon convergence and low density values corresponding to the horizon divergence. The “HorizonCube density attribute” helps define the zones of pinch-outs, condensed sections and unconformities.

Arbitrary layers – A dense set of seismic horizons can be separated into a set of arbitrary (stratigraphically uninterpreted) layers that divide the mapped seismic horizons into fix layers with a unique ID for each layer. This plays an important role in quality control where visualizing hundreds of horizons can often be slow and laborious.

Derivatives – Finally, first or second derivatives on a set of horizon events can be extended to perform a derivative on an isochron between a given ranges of events. These attributes measure subtle geometrical changes and discontinuities.

From a stratigraphic interpretation standpoint, the mapped seismic horizons in truncated form can be sub-divided into sequence stratigraphic units through the co-visualization of a structural domain and a Wheeler domain in conjunction with well data. In both domains various data, such as wells and seismic, can be combined to interpret a set of sequence stratigraphic units, such as systems tracts and sequences.

Typical attributes that are extracted using this sequence stratigraphic framework include stratigraphic unit IDs (identification number) and isochrons. The IDs of interpreted stratigraphic units can either be unique (each systems tract unit is assigned a unique identification number) or common (each identical systems tract is assigned a common number).

An isochron attribute calculated the thickness of a stratigraphic unit. The unit of this attribute depends on the seismic survey type (TWT or in depth). This is a key attribute to understand how sedimentation filled a sedimentary basin as a function of geologic time.

The relative rate of preservation – the ratio between an isochron volume and a known geologic time-span for a particular unit – also can be calculated with the results being relative measurements of the rate of preservation per geological time unit. Finally, an attribute can be generated that defines the difference between two isochron grids as a volume at a certain trace location.

Example: Extracting Deepwater Channel Stages

During the Neogene Period, the Carnarvon Basin in the Australian Northwest Shelf comprised several deepwater meandering channelized systems. One of those systems is studied using the new stratigraphic unit ID attribute. To our knowledge, this is the first study in which this attribute was used to define an unconventional set of stratigraphic units.

Convention nomenclatures are not adopted, as the interval may fall in one systems tract. Therefore, a further subdivision of channel system requires a new sequence model. In this case, a “channel stages” sequence model is defined. It contains a set of N stratigraphic units defined as channel stages. Each unit is defined by its own unique ID.

Upper and lower boundaries of the channel system were manually mapped to form an initial framework to process the densely mapped set of seismic horizons. Within such a framework, a data-driven collection of densely mapped seismic horizons was created. These were then sub-divided into a channel stages model that contained a user-defined set of color-coded stages. The criteria of sub-dividing a channel system into its stages are based on observations of features such as crosscutting relationships, timing and geomorphologic patterns.

Five channel stages can be seen in Figure 2. The same interval also can be studied using the density attribute (Figure 3), which clearly shows channel base, levees and banks of channel systems.

Conclusion

Workflows such as these support building better stratigraphic models to predict reservoir quality sands and to perform reservoir characterization. Dense sets of seismic horizons help in defining new stratigraphic attributes – and these attributes help in understanding stratigraphy and geomorphology, and can be used as a means of performing stratigraphic interpretation on seismic data.

Return to top.