uAbstract

uFigure Caps (1-11) )

uObjectives

uData Inventory

tWell spots

tSelf #82

tSeismic

uFigure Caps (12-23)

uMaps, seismic tracking, & depth conversion

uConclusions

uReference

uAbstract

uFigure Caps (1-11) )

uObjectives

uData Inventory

tWell spots

tSelf #82

tSeismic

uFigure Caps (12-23)

uMaps, seismic tracking, & depth conversion

uConclusions

uReference

uAbstract

uFigure Caps (1-11) )

uObjectives

uData Inventory

tWell spots

tSelf #82

tSeismic

uFigure Caps (12-23)

uMaps, seismic tracking, & depth conversion

uConclusions

uReference

uAbstract

uFigure Caps (1-11) )

uObjectives

uData Inventory

tWell spots

tSelf #82

tSeismic

uFigure Caps (12-23)

uMaps, seismic tracking, & depth conversion

uConclusions

uReference

uAbstract

uFigure Caps (1-11) )

uObjectives

uData Inventory

tWell spots

tSelf #82

tSeismic

uFigure Caps (12-23)

uMaps, seismic tracking, & depth conversion

uConclusions

uReference

uAbstract

uFigure Caps (1-11) )

uObjectives

uData Inventory

tWell spots

tSelf #82

tSeismic

uFigure Caps (12-23)

uMaps, seismic tracking, & depth conversion

uConclusions

uReference

uAbstract

uFigure Caps (1-11) )

uObjectives

uData Inventory

tWell spots

tSelf #82

tSeismic

uFigure Caps (12-23)

uMaps, seismic tracking, & depth conversion

uConclusions

uReference

Figure Captions (1-11)

Figure 1.  Location map of the study area with 17 wells.

Figure 2. Acoustic impedance log of Self #82.

Figure 3. Density vs. sonic values for Ordovician Wilcox Sandstone in Self #82, showing a fair level of correlation (R² = 0.6319).

Figure 4. Plot of density vs. sonic values for Pennsylvanian Glenn Sandstone in Self #82 does not show a significant correlation between them.

Figure 5. Plot of density vs. sonic values for the entire stratigraphic interval in Self #82 does not show a significant correlation between them.

Figure 6. Velocity (sonic) log of Self #82 plotted according to time, with formation tops.

Figure 7. Structural resolution, as illustrated by minimum step down (dt), is determined to be 11 ft, with the assumption of perfect removal of shallow effects.

Figure 8. Map of image area of 3-D seismic survey (A). B shows well spots. 

Figure 9. Components of data cube (time slice above vertical slices).

Figure 10. Time slice map--original version and enhanced version after time-slicing smoothing.

Figure 11. Seismic profile showing improved quality by smoothing.

Objectives

The objectives of the project are:

·        Leverage DOE project well information. 

·        Provide an analog for Ordovician Wilcox exploration.

·        Get a view of the Pennsylvanian Glenn interval adjacent to Glenn Pool oil field.

·        Test small-scale 3-D seismic survey in Northeastern Oklahoma.

Return to top.

Data inventory

Wells

There are a total of 17 wells in Sections 19/24, N17N, R12E (Figure 1). Production was discovered in June, 1985, in the Wilcox Sandstone (Ordovician Simpson Group) at a depth of approximately 2500 ft. Thickness is as much as 34 ft, and porosity is up to 18%. Cumulative production through 1996 was 950,000 barrels of oil. The productive feature is a nose, covering approximately 120 acres, on the southwest flank of Glenn Pool oil field.

Self #82

Self unit #82, in Section 21, T17N, R12E, is located approximately 2miles from the area of the 3-D survey. The suite of logs includes DIL, LDL, BHC, CNL, GR, and SP.

–   Sonic+Density => velocity X density = Impedance

–   Sonic => velocity => time/depth => event ID

Using the sonic and density logs from the Self #82, an acoustic impedance log was prepared (Figure 2). 

The sonic log can be ignored if the sonic values are predictable from the more common density log. In the case of the Ordovician Wilcox Sandstone (Figure 3), the relation of sonic to density values suggests that density values of themselves may be satisfactory. On the other hand, plots of the sonic vs. density values for the Pennsylvanian Glenn Sandstone (Figure 4) and for the entire stratigraphic interval (Figure 5) are such that the sonic values cannot be ignored in calculating acoustic impedance for synthetic seismograms. Figure 6 shows the velocity (from sonic values) in Self #82, from 300 ft to total depth, along with formation tops, plotted according to time so that the depth ticks are non-linear. 

Seismic

Features of 3-D seismic data used in this study are:

–   Vibroseis, with bin size of 55 x 55 ft 

–   141 E-W lines x 145 N-S lines

–   1420 acres, 2.2 square miles

–   1 sec, 2 ms

–   Frequency band--15-120 Hz

 Based on the calculation in equation (1), the 55-ft bin size is a little too large because fault imaging is degraded and dips greater than 34^o are also degraded.

 Bin <V_int / (4 f_max) = 15000 / (4 X 120) = 31 ft                                         (1)

 The vertical resolution is shown by equation (2) to be 62 ft. Correspondingly, the Wilcox, with thickness of 34 ft or less, is a “thin bed.” Lateral resolution, given in equation (3), is 155 ft, or approximately 2 bins. Structural resolution is 11 ft, as derived in equation (4) and illustrated in Figure 7.

 VR = V_int / (4f_dom) = 15000 / (4 X 60) = 62 ft                                          (2)

LR = 2 X VR = 144 ft (~ 2 bins)                                                                 (3)

 DZ = (V_avgdT)/2 = 1000 X .002 / 2 = 11 ft                                               (4)

 Data footprint is shown by the images of the survey area in Figure 8, with outline of live traces (Figure 8A) and outline of the area with well spots (Figure 8B). 

 Data cube is illustrated in Figure 9, with a time (or horizontal) slice and two vertical slices. The data in the study area are noisy, reflecting a rough terrain and near-surface issues, but there is good frequency. The challenge is how to improve the data. 

Improved quality of the data is illustrated in Figure 10, with the original time-slice map and the resulting enhanced map after smoothing, and in Figure 11, which shows a vertical profile, also with enhancement by smoothing. 

Return to top.

Figure Captions (12-23)

Figure 12. Structure map, along with index maps, on top Ordovician Wilcox Sandstone. Well depths are feet below 700 ft datum. 

Figure 13. Enlargement of structure map in Figure 12, showing detail of Ordovician Wilcox structure. 

Figure 14. Ordovician Wilcox structure map from well depths compared to two time-slice maps from times comparable to the top of the Wilcox; the last two are substantially different from the well-depth map.

Figure 15. Detailed time-slice map from 434 ms--gray-scale vs. color--is substantially different from the well-depth map. 

Figure 16. Detail well-depth structure map (left) and time-slice map from 434 ms, illustrating the difference between them.

Figure 17. Seismic tracking, utilizing time-plot of sonic log and seismic profile.

Figure 18. Seismic tracking, illustrated by 3-D auto-tracking of events.

Figure 19. Seismic tracking, illustrated by seismic profile with tracked events and overlay of time-plot of sonic log (left); Ordovician Wilcox amplitude and time structure maps on right (from Liner, 1999).

Figure 20. Time structure of Ordovician Wilcox structure map (left) and map of average velocity.

Figure 21. Structure map of Ordovician Wilcox structure map, from well and seismic data.

Figure 22. Comparison of structure maps—from well data only (left; Figure 13) and from well and seismic data (right; Figure 21).

Figure 23. 3-D representation of Ordovician Wilcox depth structure.

Maps, Seismic Tracking, and Depth Conversion

 With the data described above, a well-depth structure map was made of Ordovician Wilcox sandstone (Figures 12, 13, 14, and 16) and comparable time-slice maps were prepared from seismic data (Figures 14, 15, and 16). Tracking geologic events seismically, specifically the Wilcox, was achieved by utilizing the time-plot of the sonic log and overlaying it on seismic data (Figures 17, 18, and 19). Because of a strong lateral velocity gradient (Figure 20), the maps from well data and from seismic data show significant differences (Figures 14 and 16). When both types of data are used, the result is an enhanced structure map of the Wilcox (Figures 21 and 22) and a 3-D representation (Figure 23).

Conclusions

The following are conclusions from this 3-D study, some 1420 acres in size, of an oil-productive area, near Tulsa in northeastern Oklahoma:

·        Data improvement through smoothing by time slice (or FXY deconvolution).

·        Time structure is not depth structure, reflecting a strong lateral velocity gradient.

·        Postage-stamp sized 3-D seismic surveys can add detail (and detail adds value).

Reference

Liner, C.L., 1999, Elements of 3-D seismology: Tulsa, PennWell, 438 p.

Return to top.