Skip to content

Azimuthal Anisotropy Analysis Essay

[1]Formation anisotropy—the directional variation of physical properties—can be the result of either of the following processes:

  • Depositional processes (intrinsic)
  • Tectonic processes (stress-induced)

Formation anisotropy is evidenced through variations in:

  • Permeability
  • Rock strength
  • Fractures
  • Borehole failure

Acoustic logging can provide insight on these variations that assist in anisotropy analysis.

Formation anisotropy

In acoustic/seismic terms, intrinsic anisotropy is structural in nature and is commonly seen as transverse isotropy (TI or vertical transverse isotropy, VTI), in which properties differ in the vertical or horizontal planes, such as in shales or thinly bedded intervals.[2] Stress-induced anisotropy is known as azimuthal anisotropy (or horizontal transverse isotropy, HTI), in which acoustic parameters in a vertical borehole vary with azimuthal orientation, such as the case of fractures parallel to the borehole.

Analyses of in-situ anisotropy (primarily stress induced) are made using direct or derived shear-wave velocity and provide the magnitude and azimuth of anisotropy (i.e., direction of the maximum and minimum horizontal stresses as well as an indication of their difference). Anisotropy analysis has been widely used in solid-earth seismology, geothermal studies, and more recently, in exploration geophysics (see summaries in Crampin and Chastin[3] and Helbig and Thomsen[4]).

In the petroleum industry, these results are used in well design and well placement:

  • For optimum reservoir drainage[5]
  • To detect and characterize faults and fractures in openhole and cased hole[6]
  • To predict borehole instability and sand production,[7]
  • For optimizing the design and evaluation of well completions (perforations and hydraulic fracturing) [See section below on Crossed-dipole anisotropy analysis]

Compressional-, shear- and Stoneley-wave properties are each affected, to some degree, by the presence and type of formation anisotropy. While the shear-wave response to azimuthal anisotropy (see the section below on Crossed-dipole anisotropy analysis) and Stoneley waves to VTI anisotropy is well known, the effect on compressional-wave energy is less well characterized.[8][9][10][11] In some situations, information from additional measurements, such as borehole images, dip logs, or both, may be necessary for relating the measured anisotropy to geological features.

Borehole acoustic azimuthal and VTI anisotropy analysis is an advance made possible by the recent introduction of new inversion methods.[12][13][14][15][16][17][18][19] These methods use the crossed-dipole shear to derive azimuthal anisotropy and the Stoneley wave to derive TI anisotropy in slow formations, or a combination of these modes in deviated wells. A reasonable shear velocity can be derived using inversion techniques with low-frequency Stoneley-wave dispersion which is sensitive to the horizontal shear (in contrast to the dipole’s sensitivity to the vertical shear).[20][21][22]

Several new applications have been made possible by anisotropy analysis:

  • Identification of formation alteration using dipole-shear dispersion[23][14][24][25]
  • Stress estimation,[16]
  • Distinguishing between intrinsic and stress-induced anisotropy using dispersion crossover[26][27][28][29]

Anisotropy analysis can also be conducted using shear-wave parameters derived from Stoneley-wave dispersion.[19][30]

Crossed-dipole anisotropy analysis

In anisotropic media, shear waves (both monopole refracted and dipole flexural) split into orthogonally polarized components having different velocities. This is known as[31][32][33][34]:

  • Shear-wave splitting
  • Shear-wave birefringence
  • Shear-wave velocity anisotropy

The difference in fast and slow shear-wave slowness provides a measure of the magnitude of anisotropy. Shear-wave splitting is useful for evaluating:

  • Fractures
  • Faults
  • Bedding planes that intersect the borehole at an angle
  • Unbalanced tectonic stresses perpendicular to the borehole

Because the flexural waves induced by dipole-acoustic sources have a directional component, the use of mutually perpendicular (crossed) pairs of dipole transmitters and receivers can detect and measure dipole shear-wave splitting.[35][36][37][38] In isotropic formations, both receiver components (X and Y) will measure the same shear-wave arrival time. However, in an azimuthally anisotropic formation, the shear wave measured at the X-component of the receiver pair will be different from the one measured at the Y-component; one is called the fast shear component, the other, the slow shear component (Fig. 1). To measure the fast and slow shear-wave slowness, four components are measured (Fig. 2):

  • Two inline components, X-X and Y-Y
  • Two crossline components, XY and YX
  • Fig. 1 – Diagram illustrating the principle and configuration of crossed-dipole logging: two orthogonal dipole transmitter and receiver array systems. The tool acquires four array data sets, two in-line, XX and YY; two crossed-line, XY and YX, and uses the physics of shear-wave splitting in azimuthally anisotropic rocks to determine the fast-shear polarization azimuth (courtesy of Baker Atlas).

  • Fig. 2 – Example of four-component, crossed-dipole waveform log data showing shear-wave splitting caused by formation anisotropy (courtesy of Baker Atlas).

The dipole flexural shear mode is affected by a variety of factors including formation anisotropy. Significant borehole ellipticity (the result of borehole failure or breakouts)[39] and high relative inclination between the borehole and formation[1][40] may result in erroneous interpretation of dipole-derived anisotropy and must be accounted for during data processing.[41] Additionally, the presence of shale anisotropy in high-angle and horizontal wells, can significantly influence compressional velocity, which must be corrected for this effect.[42]

Initially, mathematical rotation methods originally developed for use in processing surface seismic data[43] were used to determine the fast-wave direction together with the slowness values for the fast and slow shear wave. More recently, inversion methods[44] are being applied to the four data sets to simultaneously determine the azimuth and magnitude of the anisotropy (Fig. 3).

  • Fig. 3 – Post-processing results from crossed-dipole analysis indicating the magnitude and azimuth of anisotropy (courtesy of Baker Atlas).

An anisotropy map (Fig. 4) combines the derived average anisotropy and its azimuth to generate an azimuthal image. This display facilitates interpretation by allowing the analyst to quickly assess depth intervals of interest by looking at the brightness, direction, and continuation of the features on the map. The map also facilitates comparison with borehole-image logs. Rose diagrams (Track 3) provide an accurate indication of the fast shear azimuth over each labeled depth interval. The integration of monopole and dipole measurements yields improved estimates of[22][45]:

  • Anisotropy
  • Magnitude of anisotropy
  • Permeability
  • Fig. 4 – The results of anisotropy analysis displayed as an anisotropy map (courtesy of Baker Atlas).

Horizontal stress and hydraulic fracturing

Fractures, both natural and hydraulically induced, develop in relation to regional or localized stress patterns and play a major role in optimizing production and reservoir drainage. An estimate of the magnitude and azimuth of the horizontal stresses surrounding a borehole is needed for accurately placing wells to take advantage of existing fracture patterns and for artificially inducing fracture patterns during well completion through hydraulic stimulation.[5][46]

The stresses operating on a rock formation are described by a triaxial coordinate system that consists of:

  • Two principal horizontal stresses, σx and σy
  • A vertical stress component, σz, which is the overburden.

In a borehole, these downhole stresses are expressed as radial components at the borehole wall:

  • The vertical component σz
  • The radial component σr
  • The tangential component σθ
  • The (azimuthal) shear component, σ

Unbalanced-formation-stress components produce distortion around the borehole (stress-induced anisotropy) that results in shear-wave splitting. The azimuth of the fast-shear wave parallels the direction of maximum horizontal stress and the azimuth of the slow-shear wave parallels the direction of the minimum horizontal stress. Hydraulic-fracture azimuth is parallel to the direction of maximum horizontal stress.[47]

Hydraulic well stimulation consists of perforating an interval, then packing and pressuring these perforations to create fractures behind casing to allow increased production. Crossed-dipole anisotropy logging can estimate the vertical extent of the formation-stress fracture along the borehole and its azimuth in the formation (Fig. 5).[48][16][17][18][38][49][50][51][52][53][54]

  • Fig. 5 – Crossed-dipole logging determines a fast shear azimuth of 51° before fracture treatment. The maximum stress direction of 46°, determined after the fracture job, is in good agreement with the stress direction from the crossed-dipole log results (courtesy of Baker Atlas).


  1. 1.01.1De, G.S. and Schmitt, D.P. 2005. Issues With Shear-wave Azimuthal Anisotropy in Highly Deviated Wells. Presented at the Offshore Technology Conference, Houston, Texas, 2-5 May. OTC-17647-MS.
  2. ↑Wang, Z. 2002. Seismic Anisotropy in Sedimentary Rocks, Part 1—A Single-Plug Laboratory Method; Part 2—Laboratory Data. Geophysics 67 (5): 1415–1440.
  3. ↑Crampin, S. and Chastin, S. 2003. A Review of Shear Wave Splitting in the Crack-Critical Crust. Geophysical J. Intl. 155: 221–240.
  4. ↑Helbig, K. and Thomsen, L. 2005. 75-plus Years of Anisotropy in Exploration and Reservoir Seismics: A Historical Review of Concepts and Methods. Geophysics 70 (6): 9ND–23ND.
  5. 5.05.1Franco, J.L.A., de la Torre, H.G., Ortiz, M.A.M. et al. 2005. Using Shear-Wave Anisotropy To Optimize Reservoir Drainage And Improve Production In Low-Permeability Formations in the North of Mexico. Presented at the SPE Annual Conference and Technical Exhibition, Dallas, Texas, 9–12 October. SPE-96808-MS.
  6. ↑Klimentos, T. 2003. Shear-Wave Anisotropy Applications for Perforation Strategy and Production Optimization in Oil Bearing Porous Sands, paper LL. Trans., 2003 Annual Logging Symposium, SPWLA, 1–12.
  7. ↑Klimentos, T., Farghaly, A., and Qleibo, M. 2003. Finding Faults with Shear-Wave Anisotropy, paper F. Trans., 2003 Annual Logging Symposium, SPWLA, 1–12.
  8. ↑Furre, A.-K. and Brevik, I. 1998. Characterization of Angle Dependency in Sonic Logs, paper BH 2.1. Expanded Abstracts, 1998 Annual Meeting Technical Program, SEG, 292–295.
  9. ↑Hornby, B., Howie, J., and Ince, D. 1999. Anisotropy Correction for Deviated Well Sonic Logs—Application to Seismic Well Tie, paper BH/RP 4.7. Expanded Abstracts, 1999 Annual Meeting Technical Program, SEG, 112–115.
  10. ↑Wang, Z. 2001. Seismic Anisotropy in Sedimentary Rocks, paper RP 2.4. Expanded Abstracts, 2001 Annual Meeting Technical Program, SEG, 1740–1743.
  11. ↑Sun, Y. et al. 2003. Effects of Stress-Induced Anisotropy on Monopole and Dipole Logging, paper RBG P1.2. Expanded Abstracts, 2003 Annual Meeting Technical Program, SEG, 1282–1285.
  12. ↑Koster, K. et al. 1994. Applied Production Geophysics Using Shear-Wave Anisotropy—Production Applications for the Dipole Shear Imager and the Multicomponent VSP, paper DP1.1. Expanded Abstracts, 1994 Annual Meeting Technical Program, SEG, 233–235.
  13. ↑Esmersoy, C. et al. 1995. Fracture and Stress Evaluation Using Dipole-Shear Anisotropy Logs, paper J. Trans., 1995 Annual Logging Symposium, SPWLA, 1–12.
  14. 14.014.1Tang, X. 1996. Processing Dipole Waveform Logs for Formation Alteration Identification, paper BG 3.4. Expanded Abstracts, 1996 Annual Meeting Technical Program, SEG, 1, 162–165.
  15. ↑Huang, X. et al. 1999. Formation Stress Estimation Using Standard Acoustic Logging, paper BH/RP 2.6. Expanded Abstracts, 1999 Annual Meeting Technical Program, SEG, 53–56.
  16., X., Cheng, N., and Cheng, A. 1999. Identifying and Estimating Formation Stress from Borehole Monopole and Cross-Dipole Acoustic Measurements, paper QQ. Trans., 1999 Annual Logging Symposium, SPWLA, 1–14.
  17. 17.017.1Badri, M., Sousa, S., and Klimentos, T. 2000. Shear Anisotropy Applications in Production Optimization, Western Desert, Egypt, paper RPB 1.5. Expanded Abstracts, 2000 Annual Meeting Technical Program, SEG, 1695–1698.
  18. 18.018.1Tang, X., Patterson, D., and Markovic, M. 2000. Fracture Measurements in Open/Cased Holes Using Cross-Dipole Logging—Theory and Field Results, paper RPB 1.6. Expanded Abstracts, 2000 Annual Meeting Technical Program, SEG, 1699–1702.
  19. 19.019.1Tang, X.M. 2003. Determining Shear-Wave Transverse Isotropy from Borehole Stoneley Waves. Geophysics 68 (1): 118–126.
  20. ↑Cheng, C.H. and Toksoz, M.N. 1983. Determination of Shear Wave Velocities in ‘Slow’ Formation, paper V, Trans., 1983 Annual Logging Symposium, Soc. of Professional Well Log Analysts (SPWLA) 1–18.
  21. ↑Stevens, J.L. and Day, S.M. 1986. Shear Velocity Logging in Slow Formations Using the Stoneley Wave. Geophysics 51 (1): 137–147.
  22. 22.022.1Cheng, N. and Cheng, C.H. 1996. Estimations of Formation Velocity, Permeability, and Shear-Wave Velocity Using Acoustic Logs. Geophysics 61 (2): 437–443.
  23. ↑Plona, T. et al. 2002. Mechanical Damage Detection and Anisotropy Evaluation Using Dipole Sonic, paper F. Trans., 2002 Annual Logging Symposium, SPWLA, 1–14.
  24. ↑Murray, D., Plona, T., and Valero, H.P. 2004. Case Study of Borehole Sonic Dispersion Curve Analysis, paper BB. Trans., 2004 Annual Logging Symposium, SPWLA, 1–14.
  25. ↑Sinha, B.K., Bratton, T.R., Cryer, J.V. et al. 2005. Near-Wellbore Alteration and Formation Stress Parameters Using Borehole Sonic Data. Presented at the SPE Annual Technical Conference and Exhibition, Dallas, Texas, 9-12 October 2005. SPE-95841-MS.
  26. ↑Plona, T.J., Winkler, K.W., Sinha, B.K. et al. 1998. Measurement of Stress Direction and Mechanical Damage Around Stressed Boreholes Using Dipole and Microsonic Techniques. Presented at the SPE/ISRM Rock Mechanics in Petroleum Engineering, Trondheim, Norway, 8-10 July 1998. SPE-47234-MS.
  27. ↑Plona, T. et al. 1999. Stress-Induced Dipole Anisotropy—Theory, Experiment and Field Data, paper RR. Trans., presented at the 1999 Annual Logging Symposium, SPWLA, 1–14.
  28. ↑Plona, T.J. et al. 2000. Using Acoustic Anisotropy, paper H. Trans., 2000 Annual Logging Symposium, SPWLA, 1–12.
  29. ↑Sinha, B.K., Kane, M.R., and Frignet, B. 2000. Dipole Dispersion Crossover and Sonic Logs in Limestone Reservoir. Geophysics 65 (2): 390–407.
  30. ↑Xu, S., Wu, X., Huang, X. et al. 2005. Evaluation of Anisotropic Rock Properties in Sedimentary Rocks From Well Logs. Presented at the Offshore Technology Conference, Houston, Texas, 2-5 May. OTC-17251-MS.
  31. ↑Crampin, S. 1985. Evaluation of Anisotropy by Shear-Wave Splitting. Geophysics 50 (1): 142–152.
  32. ↑Rai, C.S. and Hanson, K.E. 1988. Shear-Wave Velocity Anisotropy in Sedimentary Rocks—A Laboratory Study. Geophysics 53 (6): 800–806.
  33. ↑Yale, D.P. and Sprunt, E.S. 1989. Prediction of Fracture Direction Using Shear Acoustic Anisotropy. The Log Analyst 30 (2): 65–70.
  34. ↑Winkler, K.W. 1996. Azimuthal Velocity Variations Caused by Borehole Stress Concentrations. J. of Geophysical Research 101 (B4): 8,615–8,621.
  35. ↑Esmersoy, C. et al. 1994. Dipole Shear Anisotropy Logging, paper SL 3.7. Expanded Abstracts, 1994 Annual Meeting Technical Program, SEG, 1,139–1,142.
  36. ↑Hatchell, P.J. et al. 1995. Quantitative Comparison Between a Dipole Log and VSP in Anisotropic Rocks from Cymric Oil Field, California, paper BG1.4. Expanded Abstracts, 1995 Annual Meeting Technical Program, SEG, 13–16.
  37. ↑Beckham, W.E. 1996. Seismic Anisotropy and Natural Fractures from VSP and Borehole Sonic Tools—A Field Study. Geophysics 61 (2): 456–466.
  38. 38.038.1Walls, J.D., Dvorkin, J., and Mavko, G. 1996. In-Situ Stress Magnitude and Azimuth from Well Log Measurements, Final Report, Report GRI-95/0356, 1-147. Chicago, Illinois: Gas Research Inst.
  39. ↑Grandi, S.K. et al. 2003. In Situ Stress Modeling at a Borehole—A Case Study, paper BH 1.4. Expanded Abstracts, 2003 Annual Meeting Technical Program, SEG, 297–300.
  40. ↑Tang, X.M. and Patterson, D. 2005. Characterizing Seismic Anisotropy Using Cross-Dipole Measurement in Deviated Wells, paper BG 2.5. Expanded Abstracts, 2005 Annual Meeting Technical Program, SEG, 372–375.
  41. ↑Patterson, D.J. and Tang, X. 2005. Pit Falls In Dipole Logging - Anisotropy: Cause Of Discrepancy In Borehole Acoustic Measurements. Presented at the Offshore Technology Conference, Houston, Texas, 2-5 May. OTC-17644-MS.
  42. ↑Hornby, B.E., Howie, J.M., and Ince, D.W. 2003. Anisotropy Correction for Deviated-Well Sonic Logs—Application to Seismic Well Tie. Geophysics 68 (2): 464–471.
  43. ↑Alford, R.M. 1986. Shear Data in the Presence of Azimuthal Anisotropy—Dilley, Texas, paper S9.6. Expanded Abstracts, 1986 Annual Technical Meeting Program, SEG.
  44. ↑Tang, X.M. and Chunduru, R.K. 1999. Simultaneous Inversion of Formation Shear-Wave Anisotropy parameters from Cross-Dipole Acoustic Array Waveform Data. Geophysics 64 (5): 1,502–1,511.
  45. ↑Gelinsky, S. et al. 1998. Anisotropic Permeability in Fractured Reservoirs from Integrated Acoustic Measurements, paper RC 3.5. Expanded Abstracts, 1998 Annual Meeting Technical Program, SEG, 956–959.
  46. ↑Smith, M.B. 1979. Effect of Fracture Azimuth on Production With Application to the Wattenburg Gas Field. Presented at the SPE Annual Technical Conference and Exhibition, Las Vegas, Nevada, 23-26 September 1979. SPE-8298-MS.
  47. ↑Hubbert, M.K. and Willis, D.G. 1957. Mechanics of Hydraulic Fracturing. Trans., AIME 210: 153–166.
  48. ↑Sinha, B.K., Vissapragada, B., Kisra, S. et al. 2005. Optimal Well Completions Using Radial Profiling of Formation Shear Slownesses. Presented at the SPE Annual Technical Conference and Exhibition, Dallas, Texas, 9-12 October 2005. SPE-95837-MS.
  49. ↑Cipolla, C.L., Liu, D., and Kyte, D.G. 1994. Practical Application of In-Situ Stress Profiles. Presented at the SPE Annual Technical Conference and Exhibition, New Orleans, Louisiana, 25-28 September 1994. SPE-28607-MS.
  50. ↑Cipolla, C.L. 1996. Hydraulic Fracture Technology in the Ozona Canyon and Penn Sands. Presented at the Permian Basin Oil and Gas Recovery Conference, Midland, Texas, 27-29 March 1996. SPE-35196-MS.
  51. ↑De, G.S., Winterstein, D.F., Johnson, S.J. et al. 1998. Predicting Natural or Induced Fracture Azimuths From Shear-Wave Anisotropy. SPE Res Eval & Eng 1 (4): 311-318. SPE-50993-PA.
  52. ↑Garg, A., Desai, A.M., and Towler, B.F. 1997. Horizontal Stresses in Anisotropic Formation and Prediction of Hydraulic Fracture Direction. Presented at the SPE Western Regional Meeting, Long Beach, California, 25-27 June 1997. SPE-38342-MS.
  53. ↑Aquila, F.J., Barajas, J.S., Mesa, H. et al. 2003. Using Cross Dipole Sonic Anistropy Data to Improve Reservoir Understanding in the Southern/Marine Areas of Mexico. Presented at the SPE Annual Technical Conference and Exhibition, Denver, Colorado, 5-8 October 2003. SPE-84204-MS.
  54. ↑Barajas, J.S., Patino, A.H., Garcia, E.R. et al. 2004. Case History - Cased Hole Dipole Sonic Applications in Mexico. Presented at the SPE Annual Technical Conference and Exhibition, Houston, Texas, 26-29 September 2004. SPE-90703-MS.

Noteworthy papers in OnePetro

Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read

External links

Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro

See also

Fracture identification with acoustic logging

Hydraulic fracturing


Measuring P-wave azimuthal anisotropy has been in the recent past an elusive task; therefore, the interpreter ignored this attribute of the seismic data and left the subject to the research and technology group geophysicists. Conversely, the interpreting geophysicist knew that when measured, the anisotropy could yield important reservoir properties related to fractures and stress fields. However, little did we suspect that P-wave azimuthal anisotropy both in velocity and AVO would change our perception of data interpretation as is happening now.

Two very important changes recently occurred in P-wave anisotropy research. First, very accurate methods of measuring both azimuthally varying velocity and AVO have been developed (Jenner, 2001 and Williams and Jenner, 2001 see for example the paper by Jenner in the August 2002 Leading Edge). Second, the very large anisotropy values published in the literature were found to be fairly normal and ubiquitous for basins that are not under primary burial. (See the suggested readings for P-wave velocity anisotropy magnitudes.) Therefore, understanding the anisotropy becomes a prerequisite to both obtaining the correct imaging from processing and determining the effects of the anisotropy on the conventional seismic interpretation. An interpreter will also find there is no escaping an anisotropic world even when not working with true 3-D data. The truth is, the earth does not care which tools we use, it still knows that it is anisotropic.

True 3-D data, wide azimuth, are data shot with a complete compliment sampling of source-receiver azimuths between the source and receiver with the additional requirement that offsets are fully populated in each azimuth class. These are the data that are appropriate for very accurate measurements of azimuthal velocity and azimuthal AVO. However, most data are not acquired in this way especially in the marine environment. Therefore, little can be done to directly view the anisotropy of the earth with these data. One might then say, no problem exists, concluding that there is no need to think about the anisotropy. However, opportunity exists in 2-D and narrow azimuth data to see the effects of azimuthal anisotropy and, when the interpreter realizes the effects, more insightful interpretations are available.

Azimuthally varying velocity

Azimuthally varying velocity is probably the most significant property of the seismic data that affects the seismic interpretation. A stacking velocity that varies with azimuth creates a list of issues for the interpreter:

  • Loss of frequency content in stacking
  • Character degradation of events
  • Laterally varying non-geologic amplitude changes in the stacked data
  • Poor fault resolution
  • AVO signatures that degrade near faults
  • Amplitude and phase footprints in 3-D data and with 4-D comparisons
  • Poor merge seams between data sets
  • Incorrect azimuthal AVO analysis
  • Poor interval velocity resolution
  • Velocity function misties at the intersections of 2-D lines.

And questions for the interpreter:

  • Were the data migrated with the correct velocity field?
  • Do the RMS velocities represent bulk rock velocities? (Can these velocities be used for acoustic inversion and depth conversion)??)
  • Can workable land AVO measurements be correct? Can land AVO be sufficiently accurate to delineate reservoirs, fluids and fractures?

To take just one of the effects of azimuthally varying velocity, let’s look at the mis-stacking issue. Mis-stacking is the cause of the first seven points listed above. As an example, Figure 1 shows a common midpoint (CMP) sorted into offset. The far offsets are then sorted by azimuth. In the offset sort (Figure 1a), the CMP shows apparent reflector degradation with offset. When the far offsets are sorted by source-receiver azimuth from north, the apparent reflector degradation is seen as a set of time delays associated with changing velocity as a function of azimuth. It is the time shifts that create all the processing and interpretation issues, especially if the anisotropy changes quickly in a spatial sense. Figure 2 shows another azimuth sorted CMP where peaks and troughs are being stacked together.

The azimuthal variation in NMO velocity observed in Figures 1 and 2 can be described by an ellipse in the horizontal plane (Figure 3). For a single set of vertical fractures, the fast velocity will be equal to the bulk rock velocity and will be orientated parallel to the fractures. When near a fault, however, multiple fracture directions are often observed. With these multiple directions of anisotropy, multiple ellipses are superimposed, often significantly reducing the seismically observed velocity. However, in this case the fast velocity will no longer represent the bulk rock velocity and may not be aligned to any of the fracture directions. In addition, Rapid change in fracture intensity, particularly near faults, can cause very rapid changes in the velocity field (Figure 4). Therefore, unless one is willing and able to pick very detailed velocity analyses, there are going to be areas in the data that mis-stack and these areas will more than likely be in proximity to the faults.

Now the interpreter has to deal with a frustrating problem. That is, the faults are not well imaged because of mis-stacking (Figure 5), and the AVO or direct hydrocarbon indications dim as the horizons near the faults. Note that Figure 4 indicates that the velocity measured in a particular direction can change by 5% very rapidly. This change is not due to changes in the bulk rock velocity, but fracture intensity. Therefore, fault or image degradation can occur regardless of whether the data are wide azimuth, narrow azimuth or 2-D. To properly stack data, the velocity field must be characterized with dense enough velocity analysis to capture the variability of the earth’s anisotropy.

The idea of mis-stacking is also germane to the seismic footprint. (We refer readers to the June 1999 article in Leading Edge by Hill et. al.). Any source of change in amplitude with offset creates a footprint, and incorrect stacking velocities are one of the most effective ways of leaving an amplitude, time and phase footprint in the data. If the rapidly changing velocity field from anisotropy is not characterized accurately, a footprint results. This is especially noticeable where offsets and azimuths change from CMP to CMP or where anisotropy is high.

We should also remember that 2-D lines cross the earth’s azimuthal velocity field. Therefore, 2-D data are not immune to the effect of the anisotropy. Take the example in Figure 6. The azimuthal velocity field shown as the background colour comes from a wide-azimuth 3-D survey. Suppose 2-D lines are acquired over this same patch of earth. In that case, the individual stacking functions at line intersections may, or may not, tie depending on the local velocity anisotropy. Note also that as with Figure 4, the variations in anisotropy are spatially rapid. Thus, even for 2-D lines, the stacking velocities should be densely sampled, particularly near fault/fracture zones.

An anisotropic earth also affects 4-D and merge-zones. When the data are acquired with differing azimuths (land or marine), combining the data to obtain one velocity function will not work. In an anisotropic earth, these data need differing stacking velocity fields before comparing or merging. If anisotropy is not accounted for in the processing, one can expect apparent phase drift at seams and apparent phase and amplitude differences between 4-D surveys.

To conclude this section and reiterate a point, strong P-wave velocity anisotropy is being observed geographically everywhere and in every geologic environment except possibly in basins under primary deposition and burial. The data may not have to be corrected; however, the interpreter will do better to understand the influence of anisotropy on his/her 3-D and 2-D data.

Azimuthally varying AVO

Although not as serious an issue as the velocity anisotropy to basic interpretation, azimuthally varying AVO, resulting from the earth’s anisotropy, still has an impact on the interpreter. Opportunities exist to better evaluate AVO when anisotropy resulting from fracturing, is incorporated into the interpretation.

What is azimuthally varying AVO (amplitude variation with offset and azimuth; AVOA)? It is the change in amplitude gradient (i.e. amplitude variation with offset response) with azimuth from a reference direction, usually north. Figure 7 shows the amplitude response vs. incidence angle for a fractured carbonate. Parallel to the fractures the amplitude response is the same as if no fractures were present (the red curve). Perpendicular to the fractures (the green curve) the amplitude response is significantly different. Note the intercept or zero offset amplitude does not vary with azimuth. So, what does AVOA mean to an interpreter and how can the interpreter possibly use these properties of the pre-stack amplitudes?

First, AVOA does not just influence the pre-stack data. AVOA will carry over into the stacked amplitudes as well. When the AVO gradient is different in one azimuth than the perpendicular azimuth, the stack of these amplitudes over a wide offset range yields differing stack amplitudes (Figure 7). Therefore, in the simplest case, where 2-D lines intersect over a fractured reservoir section, the interpreter should not expect to have the same stack amplitudes. In the case of narrow azimuth 4-D acquired at differing azimuths, spatial amplitude differences between the two surveys may be due to spatial variations in the fracture density, rather than dynamic reservoir changes. Therefore, when evaluating the AVO in an anisotropic environment, the interpreter needs to be cognizant of the azimuth at which the AVO is measured. In one azimuth the AVO may exhibit a class II AVO; whereas, in the perpendicular azimuth, the reflecting horizon may exhibit no AVO or class I AVO.

In the interpretation of AVO from wide azimuth data, isotropic AVO analysis may be completely invalid because the AVO gradients are a mix of the differing azimuths. For wide azimuth data, the interpreter has two choices. Either solve for the azimuthally varying AVO solution, or only observe the AVO along one azimuth and realize that the AVO differences observed are only relative to that azimuth.

One other aspect of AVO from a fracture source is that the AVO gradient can cause intra-lithology reflections that may not be evident from acoustic impedance. Take for example a brittle shale interval within a thick interval of acoustically similar shale. The seismic synthetic from the sonic and density log may show no reflecting interface; whereas, a strong event may be present on the seismic data. In this example, a gamma ray or SP log is more likely to correlate with the seismic reflectivity than impedance from the sonic and density


The earth is anisotropic and P-wave anisotropy is often significant. Both the P-wave velocity field and amplitudes (pre-stack and post-stack) are influenced by the anisotropy. In addition, azimuthal anisotropy is an issue in narrow azimuth and 2-D data as well as wide azimuth 3-D. The interpreter, when aware of how the anisotropy manifests itself in seismic data, can incorporate these effects into the interpretation to evaluate risk. In addition, they may be able to extend the interpretation to address the source of the anisotropy.

Mostly this azimuthally varying anisotropy is attributed to fractures, which may not necessarily be macro fracturing related to reservoir permeability. Crampin has suggested micro-fracturing in the earth, which may not increase reservoir permeability, is prolific. This view is supported by our observations of P-wave azimuthal velocity anisotropy over large survey areas in various parts of the continental US. It should also be noted any alignment, whether depositional fabric or crystallographicly aligned cements filling fractures, may also cause the anisotropy. Therefore, we should not jump to the conclusion that what we observe in the anisotropy is always the open fractures.


The authors would like to thank Victor Vega (BP), Jon Huggins (Devon Energy), Chris Besler (Stone Energy) and Heloise Lynn (Lynn Inc.) for many useful discussions and comments on the issues relating to P-wave azimuthal anisotropy.


Suggested reading

Corrigan, D., Withers, R., Darnall, J. and Skopinski, T., 1996, Fracture mapping from azimuthal velocity analysis using 3-D surface seismic data, 66th Ann. Internat. Mtg: Soc. of Expl. Geophys., 1834-1837.

Craft, K. L., Mallick, S., Meister, L. J. and van Dok, R., 1997, Azimuthal anisotropy analysis from P-wave seismic traveltime data, 67th Ann. Internat. Mtg: Soc. of Expl. Geophys., 1214-1217.

Grimm, R. E., Lynn, H. B., Bates, C. R., Phillips, D. R., Simon, K. M. and Beckham, W. E., 1999, Detection and analysis of naturally fractured gas reservoirs: Multiazimuth seismic surveys in the Wind River Basin, Wyoming: Geophysics, Soc. of Expl. Geophys., 64, 1277-1292.

Jenner,E.,2001, Azimuthal anisotropy of compressional wave seismic data, Weyburn Field, Saskatchewan, Canada, Ph.D. Thesis, Colorado School of Mines.

Lynn, H. B., Campagna, D., Simon, K. M. and Beckham, W. E., 1999, Relationship of P-wave seismic attributes, azimuthal anisotropy, and commercial gas pay in 3-D P-wave multiazimuth data, Rulison Field, Piceance Basin, Colorado: Geophysics, Soc. of Expl. Geophys., 64, 1312-1346.

Williams, M. and Jenner, E., 2001, How important is the effect of azimuthal anisotropy in 3D seismic data? Enhancing quality and extending the potential of the 3D interpretation, 71st Ann. Internat. Mtg: Soc. of Expl. Geophys., 126-128.