Advanced Secondary Recovery Project for the Sooner ‘D’ Sand Unit, Weld County Colorado Final Report by diversified Operating Corporation, November 1996” stated that the Sooner Unit has been interpreted to be a Cretaceous D Sand stratigraphic play, located in Weld County in northeastern Colorado. The reservoir is interpreted to be funnel-mouth estuary and valley fill. By correlating log data, the D Sand reservoir appears to be a series of vertically stacked channels with a maximum gross thickness of 70 ft (21 m). Commonly, the channels have a gross thickness of 30 ft (9 m) or less.

The D Sand individual sands and valley fills have been very difficult to visualize using

Figure 1. Trace envelope. (Figures courtesy of Seismic Micro-Technology)
conventional amplitude data. The following workflow will employ a variety of seismic attributes, among them spectral decomposition (SD), to resolve this visualization problem.
Seismic resolution of the D Sand Survey spectrum analysis has shown that the dominant frequency for the D Sand is approximately 55 Hz, with a vertical bed resolution (tuning) of approximately 61 ft (19 m), whereas the tuning thickness chart for the Sooner Unit 1-21A illustrates a tuning time thickness of 8 to 9 ms. Therefore, tuning is calculated to be in the range of 55 ft to 60 ft (17 m to 18 m). The D Sand is located seismically at approximately 1.45 sec. two-way time. Most of the D Sand pay thicknesses are between 30 ft to 70 ft gross. Therefore, essentially all the net pay in the wells is at or below tuning.

Seismic attribute generation
Seismic subvolumes were created for the Sooner 3-D survey using the following attributes over a time interval of 1.0 to 2.0 seconds:
• Dip of maximum similarity;
• Instantaneous dip;
• Instantaneous phase;
• Relative acoustic impedance; and
• Trace envelope.
These seismic attributes were displayed at time slice 1.456 seconds.

The trace envelope attribute (Figure 1) best shows the overall shape of the D Sand represented by the orange color. None of the extracted seismic attributes displays the shape of the valley fill and estuary environment in detail.

SD of the Sooner seismic volume
The dominant frequency of the Sooner 3-D seismic volume is 55 Hz. However, the fact that
Figure 2. SD Envelope Sub-band of 32 Hz at 1.456 seconds.
the valley fill architecture cannot be resolved with the previously run seismic attributes may be due to the D Sandstone valley fills responding to different frequencies than the rest of the seismic. In other words, the valley fill tuning frequency may be either greater or less than the overall Sooner 3-D dominant frequency. SD was run to analyze the frequencies of the Sooner 3-D seismic volume, observe the different bands and examine each band volume.

SD was performed using Seismic Micro-Technology Inc.’s KINGDOM Software and Rock Solid Attributes on a seismic interval from 1 to 2 seconds. The SD Envelope Sub-band Attributes were selected, and the defaults were applied for banding based on a linear scale.

SD display in 3-D volumes
The best way to view the individual SD volumes is with a 3-D volume tool. For this study KINGDOM’s VuPAK was used. The Envelope Sub-Band for 56 Hz (close to the overall dominant frequency for the Sooner 3-D seismic survey) at the level of the D Sand, approximately 1.456 seconds two-way time, was examined. However, it can be noted that very little in the way of detail is visible at this bandwidth.

Upon examination of the SD Envelope Sub-band for the 32 Hz volume, it is apparent that the 32 Hz bandwidth may be construed as the sandier portions of the volume, represented by the yellows and oranges, as shown in Figure 2.

Delineating the valley fill complex
The SD Envelope Sub-band_32 volume shows the gross depositional nature of the D Sands, but the fluvial valley fill complex is still not clearly defined. To define what is thought to be the valley fill architecture, the Rock Solid Attributes Geometric Attributes were run over the SD Envelope Sub-band_32 volume. The full time range is not required as the D Sand reservoir is located between 1.4 and 1.6 seconds.

Geometrical attributes are calculated for each trace by scanning the adjacent traces and
Figure 3. Rendered valley fill complex.
computing various characteristics defining event dip, continuity, etc. Geometric attributes respond to changes in reservoir structure and stratigraphy. They are the result of areal variations of physical attributes computed over user-defined time and distance gates.
Instantaneous dip of the SD Envelope Sub-band_32 volume gave the best presentation of what is believed to be the valley fill architecture. At 32 Hz, tuning is approximately 100 ft (30.5 m), which is thicker than the individual D Sand stacked channels. This bandwidth appears to be imaging the broader and thicker overall valley fill deposits. Opacity was clipped so that the assumed valley fill complex remains opaque and the rest of the volume becomes transparent. The results are displayed in Figure 3.

Further delineation
What is believed to be the D Sand valley fill sequence has been delineated at 1.456 seconds. Data type coblending was used to further refine the valley fill.

The results of coblending of SD_Instantaneous Dip and SD_Shale Indicator can be viewed as
Figure 4. Coblended time slice 1.456.
shown in Figure 4. The bright colors are believed to be associated with the valley fill, while the darker colors are indicative of shales.

To determine if structure is influencing the imaging of the D Sand valley fill complex, one additional test was performed. The D Sand horizon was flattened. Instantaneous dip was then projected onto the flattened D Sand horizon.

Conclusions
The thickness of the D Sand reservoir is right at or below the tuning thickness of the seismic data. Conventional amplitude seismic interpretation is largely unsuccessful in delineating the reservoir. However, through the use of horizon attributes, various seismic attributes and the spectral decomposition of the seismic volume, it is possible to visualize what may be the valley fill complexes of the D Sand reservoir.