When it comes to placing, drilling, completing and stimulating new wells, the ability to discern anisotropy as small as 1% could be crucial to pinpoint the stress direction. But measuring the anisotropy magnitude and direction is not enough. To apply this knowledge properly, it is essential to understand the dominant mechanism generating the anisotropy, such as intrinsic anisotropy due to layering or parallel fractures, or stress-induced anisotropy generated by stress imbalance.

This article describes the use of sonic and seismic techniques to measure weak anisotropy, how these data can be interpreted to understand the earth stress state, and the use of this information to improve well treatments and field development.

Tight gas sand formations in Mexico’s Burgos Basin, being reactivated by Pemex with many new wells being drilled, provided opportunity to confirm the performance of a new Sonic Scanner modular sonic tool and the use of borehole seismic methods in quantifying small degrees of anisotropy. Sonic data from five Burgos basin wells showed that local stress direction could vary significantly ─ as much as 20° from the strike of nearby faults. This came as a surprise because maximum horizontal stress had been previously determined to be parallel to the strike of the nearest faults.

The Cuitlahuac field near Reynosa, Mexico, has been producing gas from Oligocene sands since 1972 and has been re-activated with many new wells being drilled. The field is composed of about 20 sand packages. The tight target gas sands, which also frequently present thin laminations, have very low permeability and must be hydraulically fractured to produce hydrocarbons in commercial quantities. As a consequence, the wells exhibit elliptical drainage patterns. Optimum reservoir drainage depends on correct well placement to avoid interference between wells (drainage areas overlapping) or not leaving areas untouched (drainage gaps).

The knowledge of elastic anisotropy orientation and magnitude significantly aids geomechanical applications such as determination of new well locations, identification of infield drilling opportunities and application of new perforation techniques to improve hydraulic fracturing stimulation in addition to geophysical applications like amplitude variation with offset (AVO), seismic inversion analysis, seismic processing and multicomponent seismic acquisition design processing.

Sonic and seismic imaging

Until recently, the reliable measurement of anisotropy magnitude and orientation was possible only when the acoustic anisotropy was greater than 5%. Within the sonic realm, quantification of less than 5% anisotropy can now be achieved by means of improved transmitters and additional receivers.

Figure 1. The Sonic Scanner tool, with 13 axial stations in a 6-ft (1.8-m) receiver array. Two orthogonal dipole transmitters generate flexural waves for characterization of shear-wave slowness. (All figures courtesy of Schlumberger)

The data for this study was obtained using the new acoustic tool that incorporates improved monopole and cross-dipole transmitter technology plus an extensive receiver array incorporating 13 axial levels with 8 azimuthal sensors at each level (Figure 1). Each receiver is individually digitized, resulting in 104 waveforms for each transmitter firing for improved slowness and anisotropy estimation. There is enhanced wavenumber resolution at all frequencies. Because of the quality of the data, the differentiation between isotropic and anisotropic zones is extremely clear, as determined by the minimum and maximum crossline energy.

The dipole sonic measurements were made every 0.5 ft (.15 m) at frequencies of about 3 kHz, and the seismic measurements were made at 50-ft (15-m) intervals at frequencies of about 40 Hz.
Compared with the seismic measurements, the sonic measurements are localized. On the other hand, although seismic measurements are made over total intervals from about 600 ft (183 m) to several thousand feet, with the effect of anisotropy being cumulative, they can measure a very small effective anisotropy.

The seismic data were collected with two orthogonally oriented shear vibrators located near the well, inducing particle motion in the east–west direction and the north–south direction, respectively. In addition to the shear vibrators, two vertically polarized vibrators were used to generate compressional waves, one located next to the two shear vibrators, and one located about 5,450 ft (1,600 m) to the north.

Both the sonic and seismic methods revealed evidence of weak acoustic anisotropy in a well representative of the area. The anisotropies identified by the sonic data were confirmed by the borehole shear seismic measurements.

2. Anisotropy log from Cuitlahuac field showing a zone of isotropy (A), a zone of strong anisotropy (B) and a zone of weak anisotropy (C).

With two orthogonal dipole sources and two orthogonal dipole receivers, one can reconstruct the measurements that would have been made with a dipole source oriented in any azimuthal direction. The Alford rotation process (Alford, 1986) finds the rotation angle for which the energy in the wavefield component polarized orthogonal to the rotated source (the “crossline direction”) is at a minimum. Figure 2 demonstrates the clarity of differentiation between isotropic and anisotropic zones derived from use of the new sonic tool. Zone A shows an isotropic zone, Zone B shows strong anisotropy and Zone C shows weak anisotropy. In the first track, the minimum and maximum crossline energy indicates that the Alford rotation process worked very well throughout the interval because the minimum crossline energy is very close to zero. That the maximum crossline energy is different from the minimum crossline energy indicates whether or not there is a preferential direction in the data.

In the second track it can be observed that the tool was constantly rotating in this vertical well, while in the third track a fast shear azimuth is remarkably constant due to formation anisotropy, as confirmed by the processed time and slowness anisotropy displayed in the fourth track. There are magnitude variations in the anisotropy (note that the scale is 0 to 20%, whereas the normal display is from 0 to 100%), from which a stable fast shear azimuth in zones with amounts of anisotropy varying from 2% to 8% is shown.

Where the azimuth is not stable in Zone A, it can be inferred from the maximum crossline energy and from the time and slowness anisotropy that this is an isotropic formation. This was confirmed through slowness frequency dispersion analysis. These overlays also fit the isotropic model perfectly (Figure 3a).

Figure 3. a) A zone of isotropy with both flexural dispersion curves (fast and slow) lying directly on top of the homogeneous isotropic model. b) A zone of stress-induced anisotropy indicated by crossover in the dipole dispersion curves.
These dispersion curves are steep, which is partly due to a very slow mud, which creates large dispersion. Slowness frequency dispersion analysis is used to identify the mechanism of the anisotropy (Plona et al., 2000). In one section, differential horizontal stresses form the mechanism of azimuthal anisotropy, as determined from the crossing dipole dispersion curves (Figure 3b). This stress-induced anisotropy characterization was confirmed by image log analysis in which the observed tensile wall fractures are in the same fast shear direction, and both of these are indicative of the maximum stress direction.

Intervals were also recorded where bedding could be identified as the dominant mechanism for anisotropy with a fast shear azimuth of 20 northeast. The image logs also highlight the layered nature of the formations, with the strike of the bedding in agreement with the fast shear azimuth.

Within the seismic realm, equations governing propagation of elastic waves are quite complex. However, for a shear wave traveling along an axis of symmetry in a medium that is only weakly anisotropic, the shear wave can be described in a 2 by 2 matrix and diagonalizing the data by using a similarity transform. The transformed data are equivalent to what one would record with the multicomponent source and receivers rotated in the horizontal plane by the same angle (Figure 4).

Figure 4. Energy of off-diagonal data elements as a function of rotation angle.

Figure 5 shows the difference in traveltime between the rotated data components Vyy and Vzz, indicating that the rotated yy direction is faster and corresponds to the direction of maximum stress. That the traveltime difference appears to increase almost linearly with depth indicates that on the average there is a persistent anisotropy over the entire depth interval between 5,250 ft and 9,840 ft (3,000 m).

Data from the new sonic tool and the shear vertical seismic profile (S-VSP) showed similar shear anisotropy orientation (within 5°) and magnitude (about 1%), despite the different volumes investigated and frequency range: kHz for the sonic tool, tens of Hz for the S-VSP. The S-VSP provided a way to upscale the new sonic technology results and validated the use of measurements made at the well level to be extrapolated at the reservoir scale, enabling conclusions to be drawn about drainage areas or to use the sonic anisotropy analysis for oriented perforating applications.

The work continues in this area, and there are plans to perform hydraulic fracture monitoring (HFM) surveys in the near future. With HFM, by placing downhole triaxial geophones in a nearby monitoring well, it is possible to map the actual geometry of a hydraulically induced fracture in a treated well, providing frac azimuth, wing length, frac height and estimates of frac complexity and frac symmetry. With that it will be possible to confirm that the hydraulic fracture propagates in the fast shear azimuth and finally “close the loop” between sonic-seismic acoustic measurements and the geomechanical stress earth model.

Formula for success

Figure 5. Difference in travel time between rotate data components.
As exemplified in the Burgos Basin, sonic and seismic methods were used to evaluate anisotropies ranging from 1% to 5%. In the surveyed formations, both sonic and seismic methods provided evidence of weak acoustic anisotropy, and the results agreed in magnitude and orientation. Also, using sonic waves dispersion analysis, two distinct mechanisms of anisotropy, differing in magnitude and direction, were identified.

The logging data from five wells not only indicated that the stress direction can vary significantly in this area but underscored the importance of localized anisotropic measurement prior to design of perforation, stimulation, or drilling operations.