Classical seismic facies approach is an exceedingly useful technique for better understanding of the geological development of the study area. Seismic facies is a descriptive analysis of the reflection patterns, such as sequence shape, reflection configuration, overall amplitude, frequency and continuity. Amplitude and frequency are used effectively to evaluate seismic attributes.
The sky is blue and water is wet – Standard attributes:
- Sequence shape
- Reflection pattern
- Reflection continuity
- Average amplitude
- Average frequency content
- Direct Hydrocarbon Indicators (DHI)
Possible Interpretation Results:
- Depositional systems
- Direction of deposition
- Potential Reservoir rock
- Potential Source rock
- Seal potential
Using the facies patterns and the seismic data reflections along with dip meter data, if available, direction of deposition can be interpreted along with depositional facies.
Such interpretation is critical not for fluvial environments only (see figures), but for other types of depositions as well – both clastic and carbonate.
Once the depositional facies is interpreted then migration paths and possible trapped hydrocarbons can be evaluated.
In the example here the depositional environment is a fluvial flood plain with channels, in a non-marine fluvial environment, where the seismic facies are primarily characterized by semi-parallel to parallel reflections, occasionally hummocky or chaotic reflection patterns can be observed.
Seismic facies analysis is a part of amplitude interpretation of seismic data to study sedimentation conditions of targeted intervals and lithofacies prediction based on seismic exploration data. The procedure is implemented in Stratimagic package with NNT (SISMAGETM Neural Network Technology). NNT is based on application of self-organizing neural network to recognize and evaluate a change of seismic signal form in the interval under study. The result of this work is seismic facies maps, which are created in three stages.
At the first stage, actual traces in the targeted interval are systematized in form. The form is defined as a function of trace change rate from one to another count, i.e. tilts of trace segments between adjacent counts.
In the second stage, the model traces forms are determined. Their quantity is fixed by the user and is considered as a number of different traces classes. The system modifies model traces at each iteration to select the final forms that would most fully describe the entire range of wave changes across the area.
At the third stage, model traces are successively compared with actual seismic data (trace by trace). Each actual trace is numbered by seismic facies model class, which is closest to the trace on identity criterion. The final result is a seismic map and a set of model traces describing the variety of waveforms over the entire area in a given interval. The classification process takes into account the location of each trace relative to adjacent trace. Therefore, seismic facies, identified as a result of classification, are logically distributed over the area (i.e., data classified as facies 1 is located near data classified as facies 2).