Refining of Geobody Amplitude Anomalies Using Seismic Attributes
By Satinder Chopra, Kurt. J. Marfurt, Rongfeng Zhang and Renjun Wen
T he September 2021 Geophysical Corner was an article entitled ‘3-D Visualization and Geobody Picking of Amplitude Anomalies in Deepwater Seismic Data,’ in which we reviewed the use of geobody tools to rapidly visualize the extent of geologic features that give rise to a strong seismic amplitude response. In that example, the highlighted strong amplitude anomalies corresponded to turbidite channels and fans, mass transport deposits and current-generated bars. Figure 1 shows a steeply dipping high amplitude anomaly against a salt diapir in a survey acquired in the deepwater East Breaks Alaminos Canyon area of the western U.S. Gulf of Mexico. For seismically thin reservoirs, the response of a gas-saturated sandstone gives rise to a trough at the top and a peak at the base. In this case, we would need to select two separate geobodies to map the reservoir. In contrast, we showed in our September 2021 article that the instantaneous envelope attribute allows the interpreter to pick a single geobody to represent the anomalous feature.
Using the Instantaneous Envelope Attribute
In figure 2a we show the geobody tracked using the instantaneous envelope attribute. If we carefully examine the seismic amplitude variation shown in figure 1, we notice the amplitude anomaly corresponding to the upper part (delimited by the two green block arrows) has higher amplitude values than the lower part, even though the whole anomaly shows strong red/blue colors. In contrast, the instantaneous envelope attribute shows a uniform amplitude throughout, with a break at the location of the lower green arrow, and does not exhibit the clean horizontal base that we interpret to be the gas-water or gas-oil contact. Consequently, a geobody tracked on the envelope attribute (as shown in figure 2a) would be somewhat larger than when tracked on the seismic amplitude, or relative acoustic impedance attribute. Once the geobody tracking has been completed, the computation of the original oil in place based on estimates of porosity, water-saturation and formation volume factors for oil and gas would provide overly optimistic values.
Acoustic Impedance and Iso-Frequency
In order to generate more realistic figures for the OOIP computations, the geobody tracking volumes can be refined. One way to achieve this is to use additional amplitude-sensitive seismic attributes. In figure 2b, we show the equivalent geobody derived with the use of instantaneous envelope, relative acoustic impedance and iso-frequency (20 hertz) attribute volumes, all of which define the amplitude anomaly well. The iso-frequency volumes were generated using the matching pursuit algorithm in which the dominant frequency of the data in the time-window being shown is close to 20 hertz. By restricting the range of each attribute corresponding to the anomaly reasonably, a more refined and a realistic geobody volume can be tracked, as shown in figure 2b. Figure 3 illustrates the basic details of a multiattribute geobody extraction workflow. The interpreter defines a range of acceptable values for each attribute. If all voxels honoring these criteria are displayed in the original (t,x,y) volume, we have simple crossplot volume. If we also add the constraint that an acceptable crossplotted voxel is also connected in an appropriate way to the original seed point, we obtain a geobody. In the present case, while the low impedance defines the anomaly well (figure 1), the envelope and iso-frequency attributes give a composite signature for the trough defining the anomaly and the peak above. The original envelope geobody shown in figure 2a is now constrained to contain only those voxels that exhibit a significantly negative impedance and significantly large spectral magnitude, resulting in the smaller geobody shown in figure 2b.
Using Voxel Depth
The depth (two-way travel-time value) of a voxel is also an “attribute” that can be used to define a threshold. For a given reservoir, it is more likely that there is gas in the shallower than in the deeper part of the reservoir. Examining figure 1a, we use the distinct horizontal cutoff and the petrophysical model of a gas-water or gas-oil contact to define the lower limit of our desired geobody.
In figure 4a we see a different high amplitude anomaly from the eastern part of the survey that also follows the bedding. However, in this figure the amplitude increases updip until it meets a fault, which we interpret to be a hydrocarbon seal. Figure 4b shows the geobody volume tracked using instantaneous envelope, and figure 4c shows the equivalent geobody tracked with the use of instantaneous envelope, relative acoustic impedance and isofrequency (20 hertz) attribute volumes. Again, notice a smaller size for the overall tracked geobody, which honors all three thresholds, with fewer outlying geobodies. It may be appropriately mentioned here that there might be other ways of computing a geobody. In this article we describe a simple threshold technique which allows voxels of interest to be picked up as a geobody. Another geobody technique could use a threshold as well as a seed point, with the latter used to pick adjacent voxels that satisfy the threshold and connectivity constraints.
Thus, picking geobodies is an important step toward volume interpretation of seismic data, and if one or more geobodies are picked with the use of more seismic attributes and reasonable restrictive range of values for display, such an exercise will yield a more accurate estimation of OOIP, should that be the ultimate objective.