Downloaded 06/22/14 to 155.198.30.43. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.

org/

Dipole shear anisotropy logging

Cengiz Esmersoy*, Schlumberger-Doll Research; Klaas Koster, Marty Williams, Amoco
Production; Austin Boyd, GeoQuest; and Michael Kane, Schlumberger-Doll Research SL3.7
Summary
Sonic-scale shear anisotropy of formations are obtained
from dipole-shear measurements acquired by orthogonal
source and receiver pairs. These measurements resemble
miniature 4-component shear VSP surveys inside the
borehole. Data from two orthogonal sources and an array
of orthogonal receivers are processed providing three main
logs; the fast and slow shear slownesses, and the fast shear
polarization angle. Potential applications include; shale
anisotropy for better seismic models, maximum stress and
fracture/microcrack strike direction for horizontal well
drilling and fracture height and azimuth determination in
hydraulic fracturing.

Introduction
In an anisotropic medium shear waves split into two waves
with orthogonal polarizations. This property has been
utilized in seismic for azimuthal anisotropy measurements
(Crampin, 1985; Alford, 1986; Thomsen, 1988; Johnston, Figure 1. Logging azimuthal shear anisotropy in a vertical well.
1986; Winterstein, 1991). Anisotropy comes in various
forms representing the symmetry of the medium at some
scale. In this abstract only the simplest form of anisotropy;
Transversely Isotropic (TI) anisotropy is considered.
Intrinsic anisotropy of horizontal shale beds, and effective
anisotropy of horizontal fine layering represent the TIV (TI
anisotropy with a vertical axis of symmetry) anisotropy
where the shear anisotropy effects are maximum for
highly-deviated and horizontal wells (White, 1991;
Esmersoy, 1994). Oriented inclusions such as vertical
fractures and microcracks, and differences in horizontal
stresses cause azimuthal (or TIH) anisotropy. As illustrated
in Figure 1, a vertical well provides the best conditions for
azimuthal shear anisotropy measurements; largest slowness
separation between the split-shear waves and insensitivity
to effective anisotropy of horizontal layering.

Shear/Flexural wave splitting

The problem of wave propagation in a borehole surrounded
by an anisotropic medium has been studied by many
researchers and several numerical and analytical methods
have been developed to compute the wavefield (Schmitt,
1989; Ellefsen et al., 1990; Sinha et al., 1994). Here we will
consider only the form of the solutions in terms of the
shear/flexural wave Green’s functions. Figure 2 shows a
borehole in an anisotropic medium with its axis of
symmetry perpendicular to the borehole axis. A dipole
source oriented at degrees from the fast shear
direction can be replaced by two dipole sources, u(t) cos
and sin inside the fast and slow shear polarization
planes. The dipole Green’s functions representing
propagation from the source to the receiver location for fast
and slow polarization directions are denoted by g,(t) and Figure 2. Shear/flexural wave splitting in dipole logging.

1139
 2 Dipole shear anisotropy logging
Downloaded 06/22/14 to 155.198.30.43. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

g,(t). These functions are typically complicated containing

the coupling and propagation properties of the shear waves
and the dispersive waves. The exact expressions for
the and cross, components of the
waves are given in terms of these Green’s
functions by

= +
= u(t) sin cos

where denotes convolution in time. The cross component

has a very simple form; it is the difference between the slow
and the fast direction responses weighted by an angle
term. As the angle varies the cross-component waveform
shape remains the same, but its amplitude is scaled by
= sin The cross-component amplitude
becomes identically zero, for = when source
orientation is in the plane of fast or slow shear wave
polarizations. The cross-component waveforms change sign
as the angle goes through one of these planes. These
properties of cross components will be used in the following
Figure 3. Variation the cross-component waveforms and
section to interpret field data observations.
travel times, as the tool rotates during
logging. Depths in this denote the position of the
dipole source which is 12.75 ft below the center of the
Cross-component waveforms receiver array.

The center panel in Figure 3 shows one of the The anisotropy logs
cross-component waveforms of the first dipole source across
a ft section. The right panel shows the azimuthal The discussion in the previous section showed that if the
orientation of the first dipole, and the left panel shows the tool rotates rapidly, we can determine the fast-shear
variation of the travel times of the first (light curve) and polarization direction in a section from the
second (dark curve) dipole components at cross-component energies and from inline-component travel
equal-offset receivers. These times are obtained by picking times. To measure the fast-shear polarization direction, and
an early zero crossing in the waveforms, slow and fast shear slownesses at every log point (e.g., every
therefore their values are somewhat larger than the the 0.5 ft), one could stop the tool at that depth and record
actual shear travel times. and cross components as it rotates azimuthally. But
this would be impractical and, in fact, it is not necessary.
Below ft, the tool orientation does not change and both 4-component data refer to data acquired by two orthogonal
the cross component amplitudes and the relative positions dipole sources located at some depth and to two orthogonal
of the travel times remain the same. Above this depth, dipole receivers located at some other depth. Theoretically,
however, the tool starts rotating rapidly causing significant 4-component data contain all azimuthal information at that
variations in the cross-component amplitudes and the travel
depth position. In other words, assuming perfectly matched
times. The cross-component amplitudes go through minima
source and receiver pairs, given 4-component data one can
at around 490 and 484.5 ft. The tool rotation between reconstruct dipole waveforms at any azimuthal orientation
these two depth points is about 90 degrees, as seen in the
of the tool. This “tool” rotation is obtained by using
right panel, consistent with an azimuthal anisotropy model.
(1986) technique of rotating shear
Moreover, as predicted by the theoretical model in the
seismic data. As indicated in Figure 2 and equation as
previous section, the cross-component waveforms change
the tool is rotated synthetically, when the dipole source
sign on two sides of these minima. The travel times indicate
direction is aligned with the fast or slow shear polarization
that the faster wave is polarized along the
directions of the medium, the cross components become
second dipole direction at 490 ft, and along the first dipole
zero. Therefore, one way to determine the fast or slow
direction at 484.5 ft. Since the azimuth log on the right
shear directions, at each depth location, is to rotate the
panel shows the first dipole orientation, in this section the
data as explained above and find the angle that
fast shear direction is about N E (which is the same as
minimizes the energy in the cross components.
N E). This observation will be compared later with
the results obtained by inverting all four components, at Figure 4 shows the 4-component tool rotations applied to
each depth point, to find the fast-shear polarization field data. The four light traces at 0 and 90 degrees, are the
direction logs. The inversion will show that all through measured and cross components. The black traces
section the fast shear polarization is around N E. are computed from the measured traces for rotation angles
 Dipole shear anisotropy logging 3

from 0 to 180 degrees in 2 degree increments. These

Downloaded 06/22/14 to 155.198.30.43. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

rotated waveforms agree with the theoretical predictions

assuming azimuthal anisotropy. The cross components
become minimum at around 60 and 150 degrees (separated
by 90 degrees), and the cross component waveforms change
sign across these angles. The inline components mainly
change in amplitude and in travel time. The inline
component at 150 degrees of rotation arrive earlier than the
one at 60 degrees of rotation, indicating that it is the
fast-shear polarization direction. The dipole source
orientation at this depth is N -40° E, meaning that the 150
degrees of rotation points at N 110° E, consistent with the
observations in Figure 3.
The procedure described above; rotating 4-component data
by various angles, picking the one that minimizes the cross
energy and giving the earlier inline arrival, can be used
with dipole data at each depth to generate angle logs. This
is shown in the center panel of Figure 5. Independent
estimates of the fast-shear azimuth at every 0.5 ft are
consistent across this whole section giving an overall Figure 4. Rotation of component dipole shear waveforms.
azimuth of about N 110° E. The left panel shows the the
two slownesses along the fast and slow polarization
directions. These logs are obtained by first rotating the Figure 5. Dipole shear anisotrop y logs; fast and slow shear
4-component data at each depth using the angles in the slownesses (left), fast-shear polarization azimuth (center)
center panel, and then by computing the slownesses from and minimum and maximum energies in the cross
the inline receiver arrays. The slowness anisotropy components.

1141
 Dipole shear anisotropy logging
Downloaded 06/22/14 to 155.198.30.43. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

(slowness difference divided by average slowness) is mostly

Acknowledgments
in the range of 3 6%. The right panel shows the
minimum and maximum percent cross energy. These curves We would like to thank Amoco Production Company for
are the minimum and maximum possible values (as a their collaboration in this study and for their permission to
function of of the energy in the cross component data publish the data set presented here.
divided by the total energy. A small minimum cross energy
means that the assumed anisotropic model fits the data
References
well at the solution. This is a necessary but not a sufficient
condition for the angle estimate to be reliable. For example R. M., 1986, Shear data in the presence of azimuthal
in an isotropic formation any value of 8 gives identically anisotropy: 56th Ann. Internat. Mtg., Explor.
zero cross components. A large maximum cross energy Geophys., Expanded Abstracts, 476-479.
indicates that the cross energy is highly dependent on
and that the environment is not isotropic. Finally, a small S., 1985, Evaluation of anisotropy by shear-wave
minimum cross energy and a large maximum cross energy is splitting: Geophysics, 50, 142-152.
a reasonable indicator of anisotropy that is consistent with
Ellefsen, K. J., Cheng, C. H. and M. N., 1990, Elastic
the model. In Figure 5, there is a correlation between the wave propagation along a in an anisotropic
anisotropy measured in the slowness logs and the behavior medium: 60th Annual Internat. Mtg., Expl.
of the cross-energy plots. For example, in the energy plot, Geophys., Expanded Abstracts, Expl. Geophys.,
the best separation between the minimum and maximum 14-17.
curves is around 478 ft, referenced to the center of the
receiver array. Since the source is 13 ft lower than this Esmersoy, C., 1994, Dipole shear anisotropy logging: 10th
reference, for this position of the tool, waves are Petroleum Congress of Turkey, Expanded Abstracts.
propagating through the highly anisotropic zone between
Johnston, D. H., 1986, VSP detection of fracture-induced
478 and 488 ft consistent with the large maximum energy.
velocity anisotropy: 56th Ann. Internat. Mtg.,
The small minimum energy, then, indicates that the
Explor. Geophys., Expanded Abstracts, 464-466.
anisotropy is consistent with the TIH model, therefore the
angle estimate is likely to be accurate. Koster, K., Williams, M., Esmersoy, C. and Walsh, J., 1994,
Applied production geophysics using shear wave
anisotropy: production applications for the dipole shear
Conclusions imager and the multicomponent VSP: 64th Ann.
Dipole shear acquisition, for the first time, provides an Internat. Mtg., Explor. Geophys., Expanded
opportunity for measuring sonic-scale in-situ anisotropy of Abstracts.
formations surrounding the borehole. It is expected that in
Mueller, M. C., Boyd, A. J. and Esmersoy, C., 1994, Case
most highly deviated wells, particularly in shales, shear
studies of the dipole shear anisotropy log: 64th Ann.
anisotropy will have significant effect on the dipole shear
Internat. Mtg., Explor. Geophys., Expanded
logs. It is also expected that, differences in horizontal
Abstracts.
stresses and/or oriented will have a
measurable effect on the dipole measurements in vertical Schmitt, D. P., 1989, Acoustic multipole logging in
wells. The field example, indicates that, at least in this transversely isotropic poroelastic formations: J. Acoust.
normal-slow formation, relatively small amount of shear Am., 86,
anisotropy can be measured. This conclusion is supported
B. K., Norris, A. N. and Chang, S. K., 1994,
by the fact that, different and somewhat independent ways
modes in anisotropic formations: to appear in
of obtaining the fast-shear azimuth; the cross-component
Geophysics.
waveform minima and the 4-component processing gave
consistent results. The fast-shear azimuth results are, also, Thornsen, L., 1988, Reflection seismology over azimuthally
supported by a shear VSP conducted at this well. Koster et anisotropic media: Geophysics, 53, 304-313.
al. (1994) s h ow that the azimuth obtained from a
4-component shear VSP survey is in very good agreement White, J., 1991, Recent North Sea experience in formation
with the dipole shear results. Mueller et al. (1994) present evaluation of horizontal wells: Presented at the SPE
the results of two case studies where the dipole shear Offshore Europe conference, Aberdeen, Scotland.
anisotropy logging is used to characterize fractures in fast Winterstein, D. F. and Meadows, M. A., 1991, Shear-wave
format ions. polarizations and subsurface stress directions at Lost Hills
field: Geophysics, 56, 1331-1348.