The bandwidth of the P-wave migrated data is approximately 10-72 hz. while the S-wave bandwidth is much lower, approximately 10-24 Hz. Bandwidth, in this context, is the portion of the signal spectrum that is repeatable between the initial and repeat surveys using the processing flow outlined in Section 6. Please note that more aggressive processing techniques could be applied to the data to increase the apparent frequency content, but results in a greater amount of noise in the repeatability measurements such as difference stacks.

The overall reflectivity can be correlated between the P-wave and S-wave data but differences can be seen, particularly in the shallow section above the Rustler. In the shallow well control there is an interface with an abrupt increase in density. This coincides with the strong reflectivity at 0.80 seconds on the S-wave data. This reflection is the dominate energy on the S-wave data but not on the P-wave data. For both the P-wave and S-wave volumes, the reflection strength within the reservoir zone is weak due to low reflection impedance. The low reflectivity was verified using dipole sonic log information from two wells in the survey area (Roche, RCP Sponsors Meeting Notes, September 25, 1995; DeVault, RCP Sponsors Meeting Notes, April 15, 1996). Reflection continuity is good on both the P-wave and S-wave data for the section above the San Andres horizon (Queen to Rustler). Continuity of the S1 and S2 data is comparable. The interface of high reflectivity at 0.800 seconds on the S-wave data corresponds to the base of an interval of high Vp/Vs ratio identified in the VSP data (RCP Sponsors Meeting Notes, October 14, 1996).

Over the reservoir interval on the S-wave data, approximately 1600 to 1700 ms, the reflection continuity is less continuous. I ascribe the discontinuous nature of the reflected energy to be caused by fracturing and local stress variations in the more rigid dolomite as opposed to the less rigid lithology above the San Andres. The anhydrite and evaporite content of the section above the San Andres increases (see Figure 2.5) as the basin desiccates. The Salado is primarily salt and anhydrite. Lower S/N ratio could also be a factor, with the high S-wave reflectivity at 0.80 seconds acting as an elastic wave barrier to energy transmission.

Analysis of reflection continuity by Talley, 1997, associates faulting and fracturing with discontinuities in the reflection character. His work identifies faults and fracture systems propagating upward from basement faults such as the north - south trending regional fault system identified by Hill, 1984, and Shumaker, 1990 (Figures 2.2 and 2.3). Fracture systems in the San Andres are influenced by the deeper faulting and modified by the east - west trending Permian shelf edge margin (Figures 2.8 and 2.9). Talley, 1997, notes the intensity of the reflection discontinuity, related to faults and fractures, diminishes above the San Andres. The increased salt and anhydrite content may reduce the rigidity of the shallower lithology, lessening the fracture intensity and spatial variations in the local stress field.

Within the reservoir interval, the S2 reflection character is more continuous than that of the S1 data. Core analysis (Scuta, in RCP Sponsors Meeting Notes, April 15, 1996) show fractures are present, both open and healed with anhydrite. Fractures are more numerous along the shelf edge, CVU 100, as opposed to north of the shelf edge, CVU 345. Analysis of fracture orientation from a single FMI log in the Warn State 26-1 well, approximately 2000 feet south - southwest of CVU 97, measures high-angle fracture sets, parallel and perpendicular to the southwest to northeast trending shelf margin (Scuta, in RCP Sponsors Meeting Notes, April 10, 1997). The more discontinuous nature of the S1 reflections in the reservoir zone may be related to more numerous fractures trending along the shelf edge. Fault and fracture trends delineated by seismic continuity analysis also parallel the San Andres shelf edge (Talley, 1997).

Evidence for southwest to northeast trends can be seen in the P-wave data and reservoir production statistics. Interpretation of the P-wave data provided a structural image of reservoir geometry. Ties to well control are excellent. Time to depth conversion was accomplished using the time : depth relationships (P-wave and S-wave) obtained from the VSP survey at CVU 200. Data were datumed at the Queen horizon, then a single spatially invariant velocity function was applied to convert the deeper horizons to depth. Ties to well control were in good agreement through the Grayburg to the top of the upper San Andres. The seismic response beneath the top of San Andres is primarily a "tuning" phenomena, thus a horizon pick on a waveform associated with porosity changes within the reservoir did not produce an accurate depth tie to a specific marker picked from well control. Figure 7.6 is a perspective view of the time structure of the top of San Andres from the P-wave data volume. Perspective view is from the southwest toward the northeast. The shelf crest is visible, trending away from the viewer. The blacked-out contour highlights a southwest to northeast lineation through the center of the survey area. This feature coincides with reflection character discontinuities (Talley, 1997) and with the general demarcation of the reservoir production data dividing the reservoir into two general regions. The reservoir produces more fluid at a higher water cut with less gas to the north and northwest of this lineation (Figures 3.2 to 3.7).

These data provide several fundamental insights concerning time-lapse, multicomponent seismology. The first insight is not related to time-lapse issues, rather the different images of the subsurface reflectivity provided by the S1 and S2 volumes. The S1 and S2 reflectivity are providing different and complementary information of the subsurface bulk properties. Note that reflectivity is more similar in the shallow section, Rustler to Yates interval, which is dominated by thick salt and anhydrite formations. Below the Yates marker the section is predominately carbonates. With respect to stress, salt is a more isotropic material that may absorb local stress variations while the more rigid carbonate rock would preserve local stress variations. It is possible that the variations in S1 and S2 reflectivity are delineating variations in lithology, porosity, stress conditions and fluid characteristics (pressure and properties). An alternate explanation is a decrease in signal / noise ratio beneath the high shallow reflectivity. Note that examination of the deeper S-wave data (figures 6.39, 6.40, 6.41 and 6.42) shows good S/N but poorer repeatability at the deeper Paleozoic reflections.

The compressional and shear wave reflectivity are similar but different in the Phase VI data volume. This reaffirms the observation that p-wave and shear-wave data provide more information about the subsurface reflectivity when used together. Compressional data provide a measure of the bulk rock compressibility, rigidity and density while shear-wave data is sensitive to rigidity and density. Compressibility and rigidity are bulk properties and are likely to be anisotropic. In porous media, pore fluid pressure and fluid properties affect the bulk properties. The combined use of p-wave, S1 shear and S2 shear seismic data allow different views of the bulk properties in the subsurface. Interpretation strives to use these volumes to delineate spatial variations in the subsurface related to lithology, porosity, variations in pore structure related to preferred permeability directions, and variations in pore fluid pressure and properties.

Figures 7.7 and 7.8 show the S1 and S2 polarizations for the S-wave migrated data for Inline 68 from the repeat survey. Reflectivity above the reservoir zone is very similar between the initial and repeat surveys for both the S1 and S2 data. Subtle waveform variations between the Rustler and Yates interval compare very closely ( for example 0.93 seconds or 1.08 seconds). Difference ensembles for the S1 and S2 data for the Inline 68 traverse are shown in Figures 7.9 and 7.10. The high amplitude reflectivity at 0.80 is virtually absent on the difference stacks indicating the acquisition and processing have produced the same image, in amplitude and phase, of the shallow reflectivity that is assumed to be constant between the initial and repeat surveys. There are differences between the surveys in the reservoir zone and even above the reservoir zone that will be discussed in the dynamic interpretation section.

The north - south traverse, inline 68, extracted from the migrated data volume through wells CVU 97 and 200 was selected since significant reservoir processes occurred at these locations. There are other traverses which could be shown where the reflectivity between the initial and repeat surveys is virtually identical and the difference stacks show very little energy. This forms a second significant insight, the repeatability of the S-wave data volumes, especially above the San Andres formation. Although the S1 and S2 volumes exhibit less continuity than the P-wave volumes, shear-wave data provide very similar images for the initial and repeat surveys. I interpret this observation to indicate shear-wave data provide real and valid information about subsurface bulk properties. The lack of continuity of S-wave data when compared to P-wave data should not be construed as "poorer" data quality. It is likely to be actual information regarding the spatial distribution of rigidity, density and velocity anisotropy in the subsurface.

Another indication of a general permeability trend of northwest to southeast through well CVU 97 can be inferred from the water injection volumes and surface tubing pressures. The area of higher Vp/Vs ratios greater than 1.9 in figures 7.11 and 7.12 delineates a "fairway" approximately 2500 feet wide in the southwest to northeast dimension, trending west - northwest to east - southeast through the center of the survey area. Figures 7.19 and 7.20 are Vp/Vs ratios with contours of surface tubing pressure overlayed. This trend of Vp/Vs higher than 1.9 generally coincides with the S1 direction. The preferential permeability direction will be along the S1 direction due to open fractures systems, microfractures and low-aspect-ratio pore structure. Although there is an operational overprint to the water injection pressures and volumes, this "fairway" allows the injection of a greater volume of water at a lower surface tubing pressure than outside the fairway. Note the anomalous region of a possible change in the local S1 direction near CVU 244 is in agreement with the surface tubing pressure data.

My interpretation of the higher Vp/Vs ratios in the fairway is indicative of more fractures and crack-like or low-aspect-ratio pore structure with better permeability; i.e., the reservoir is more compliant with less filling of porosity by anhydrite. S-wave velocity is reduced relative to P-wave velocity, increasing the Vp/Vs ratio. Arestad, 1995, presented similar results from the Joffre Field study in RCP Phase V. At Joffre, the reservoir is also a dolomite with the reservoir production characteristics governed by the anhydrite plugging of porosity and fractures. The RCP Phase V report included the observation that better reservoir production coincided with areas of higher Vp/Vs ratio.

There are different spatial patterns to the Vp/Vs1 and Vp/Vs2 ratios shown in Figures 7.11 through 7.20 that can be explained in terms of local changes in anisotropy. The data volumes were rotated to a single S1 and S2 azimuth; local variations in anisotropy will cause the fast shear-wave polarization to change. Analysis of anisotropy by Talley, 1997, showed evidence for spatial variation of shear-wave splitting in the San Andres reservoir interval and that areas of high anisotropy measures coincided with the better producing wells. The same spatial variation of S1 and S2 velocity associated with shear-wave birefringence is incorporated in the Vp/Vs1 and Vp/Vs2 measurements. For example, the narrow region of high Vp/Vs1 between wells CVU 196 and CVU 244 may indicate an area with open fractures or pore structure aligned in the northeast direction. A second example, the area near CVU 97 shows a higher Vp/Vs2 ratio, indicating open fractures or pore structure aligned to the northwest. Areas with higher Vp/Vs ratios for both the S1 and S2 data may suggest multiple azimuths of open fractures.

Note that the area of higher Vp/Vs2 ratio appears to cover more area within the fairway as opposed to the area of higher Vp/Vs1 ratio. The regional horizontal stress or S1 direction is dominate, opening more pores and fractures aligned in this direction thus increasing the area of higher Vp/Vs2 ratios. Areas with higher Vp/Vs1 ratios are smaller, more localized features, indicating changes in open fracture direction, micropore structure or stress conditions. Remember that these anomalous areas on "S1" maps (regional S1) may represent local S2 regions which are affected by fracture and pore structure.

During the infill drilling down to 10 acre spacing, two wells produced approximately four times the expected rate (Wehner, 1997). These wells are CVU 195, located 3 locations east of CVU 97 at the eastern edge of the migrated data volume, and CVU 203, located south - southeast of CVU 97. Both of these wells are located near regions of higher Vp/Vs ratios.