IPTC 10546

A Novel Ultrasonic Cased-Hole Imager for Enhanced Cement Evaluation

R. van Kuijk, SPE, S. Zeroug, B. Froelich, SPE, M. Allouche, SPE, S. Bose, D. Miller, J.-L. le Calvez, V. Schoepf,
SPE, and A. Pagnin, Schlumberger

Copyright 2005, International Petroleum Technology Conference

energy mainly though shear coupling to the surrounding cement,
This paper was prepared for presentation at the International Petroleum Technology Conference so that well-bonded solid cement attenuates more quickly than a
held in Doha, Qatar, 21–23 November 2005.
fluid. Due to the low frequency, the CBL logs made with these
This paper was selected for presentation by an IPTC Programme Committee following review of
information contained in an proposal submitted by the author(s). Contents of the paper, as
tools lack azimuthal resolution, which makes it difficult to
presented, have not been reviewed by the International Petroleum Technology Conference and are distinguish channeling from poor cement properties.
subject to correction by the author(s). The material, as presented, does not necessarily reflect any
position of the International Petroleum Technology Conference, its officers, or members. Papers
presented at IPTC are subject to publication review by Sponsor Society Committees of IPTC. The ultrasonic tools, such as the UltraSonic Imager (USI*),
Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes
without the written consent of the International Petroleum Technology Conference is prohibited. operate on a pulse-echo mode and use a single transducer acting
Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations
may not be copied. The abstract must contain conspicuous acknowledgment of where and by whom
as a transmitter and a receiver. The transducer sends a short
the paper was presented. Write Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, ultrasonic pulse with a frequency content ranging from 200 to 700
U.S.A., fax 01-972-952-9435.
kHz at normal incidence to the casing surface. This pulse excites
the casing into its thickness resonance mode, which is detected as
Abstract an exponential decay in the reflected echo. The decay rate is
Current acoustic (sonic and ultrasonic) techniques for cement controlled by the acoustic impedances of the inner mud and the
evaluation have proved to be limited in providing unambiguous outer cement. Knowing the mud impedance, obtained from a
answers to the zonal isolation issue. This is especially true in separate measurement, enables the extraction of the cement
lightweight cements where they often fail to differentiate cement impedance. The transducer geometry and the high frequency
from mud. Also, as far as imaging of the cement sheath is signal combine to give a high lateral resolution of about 30 mm.
concerned, ultrasonic pulse-echo tools fail to image beyond the Transducer rotation provides full azimuthal coverage of the
cemented region adjacent to the casing, thus providing limited casing circumference, and the measured acoustic impedance can
diagnosis of the annulus. A new ultrasonic imaging tool has been be displayed as a map, which allows local defects such as
developed to address these limitations. channels to be identified. The accuracy of the acoustic impedance
measurement is 0.5 MRayls, which means there must be at least
The new imager combines the classical pulse-echo technique 1.0 MRayls difference between fluids and cement impedances to
with a new ultrasonic technique that provides temporally compact distinguish one from the other.
echoes arising from propagation along the casing and also
reflections at the cement-formation interface. Processing these Recent developments in cement formulations2,3,4 led to high-
signals yields unprecedented characterization of the cased hole performance, light-weight cements with densities less than 1
environment in terms of the nature and acoustic velocity of the g/cm3. Because the solid volume fraction of such high-
material filling the annulus between casing and formation, the performance cement is comparable to that of conventional neat
material immediately behind casing, the position of the casing cement, the compressional ultrasonic velocity is not significantly
within the hole, and the geometrical shape of the hole. affected by density. The acoustic impedance, a product of density
and acoustic velocity, can be lowered to about 2.5 MRayls for
Different wells cemented with conventional and light cements uncontaminated fully set cement. With a mud impedance of 1.5
were logged with the new experimental tool. The results MRayls or more the present ultrasonic tools face a problem of
demonstrate enhanced cement evaluation for both cement types discriminating between light-weight cements and mud. Another
and significant reduction in the uncertainty in making a squeeze problem faced by both sonic and ultrasonic tools is the shallow
or no-squeeze decision. depth of investigation in the radial direction. The measurements
are sensitive to the cement region immediately behind casing and
Introduction do not probe the annulus width for defects within the cement
sheath or at the cement-rock interface.
Cement evaluation logging tools have been used successfully for
many years to evaluate casing and cement conditions. These These limitations prompted the development of a new
tools, which use sonic or ultrasonic1 techniques, are designed for ultrasonic cement evaluation tool, based on the combination of
conventional steel casing and cements. The sonic tools, two casing modes that operate at a sufficiently high frequency to
commonly known as Cement Bond Log or CBL, operate at keep the required azimuthal and vertical resolutions. The first
frequencies of about 20 kHz and measure the amplitude or the mode is the thickness resonance mode, described previously, with
attenuation of a wave traveling along the casing. The wave loses
*
Mark of Schlumberger
2 IPTC 10546

a normal incidence single transducer. The second mode is the

flexural mode, never before used in well logging and described in
the next section. The combination of the two modes enables a
better evaluation of low-impedance cements, and the flexural
mode has the potential to probe the full annulus between the
casing and formation.

Flexural Mode

At sufficiently high frequencies (80 kHz and more), the

ultrasonic transducer pulse interacts with an azimuthally localised
area of the casing. It then becomes appropriate and convenient to
approximate this casing area as part of an infinitely unbounded
steel plate. In the frequency range of interest, two modes
dominate the wave physics of a fluid-loaded plate. The first mode
has a particle displacement symmetric with respect to the middle
plane and mainly parallel to the plate; it is referred to as the
extensional mode (at very low frequencies, it becomes the basis
for classical CBL tools). The second mode has a particle
displacement antisymmetric with respect to the middle plane and
mainly perpendicular to the plate; it is referred to as the flexural
mode. Its particle displacement and spectral characteristics make
the flexural mode a prime candidate for probing the cement
sheath. The extensional and flexural modes of a plate are also Fig. 1—Phase and group velocity of the flexural mode in an 8-mm
referred to more formally as the two lowest symmetric (S0) and thick steel plate.
antisymmetric (A0) Lamb modes, respectively.5
The dispersion curves shown in Fig. 1 also dictate how to excite
a flexural wave with a plane wave radiated by the fluid-immersed
Flexural Mode Velocity transducer and impinging on the casing. The phase-matching
The flexural mode is dispersive, meaning that its velocities (phase angle, measured with respect to casing normal, is given by Snell’s
and group velocities) are frequency dependent. The phase law,
velocity is the velocity of a phase front; i.e., the velocity of the
peaks and troughs of a purely sinusoidal wave. The group v
sin(θ ) = f , (1)
velocity is the velocity of the wave packet centered on a given v flex
frequency.

The calculated phase and group velocities for the flexural mode, where v f and v flex are the fluid velocity and the plate flexural
as a function of frequency, are shown in Fig. 1 for a water-loaded wave phase velocity, respectively. This angle is slightly
8-mm-thick steel plate. The phase velocity is far from constant, dependent on the fluid velocity and ranges from 28− 38 degs
meaning that the flexural wave shape (before taking the envelope) when the fluid velocity covers a typical range of 1,250 −1,650
changes rapidly during propagation along the casing. On the other m/s. The higher the slowness value, the larger is the phase-
hand, the flexural group velocity exhibits a peculiar feature: It matching angle. In a reciprocal way, a flexural wave propagating
goes through a maximum around 200 kHz and remains along the casing reradiates or leaks a wave into the materials
approximately constant over a fairly wide interval of 100 to 400 present on both sides. In particular, this allows the receiving
kHz. This means that all the frequency components of a transducer to pick up the signals. The radiated (or leaky) wave in
broadband wave packet within this frequency range will the fluid propagates at an angle again given by Snell’s law such
propagate at nearly the same velocity; hence, the wave packet that maximizing its reception by a receiving transducer requires
remains temporally compact. This fact has an important the latter to be aligned at the same angle that maximizes the
implication: Using the envelope of the flexural wave packet, it is excitation of this wave by the transmitting transducer.
possible to make accurate time of arrival and amplitude
measurements.
Flexural Mode Attenuation

The radiation into the inner and outer media, with respect to the
casing, is accompanied by attenuation of the flexural wave
amplitude along its path. This attenuation is sensitive to the
mechanical properties of both materials. To a good
approximation, the overall attenuation is the sum of the
attenuations due to the inner fluid and the attenuation due to the
IPTC 10546 3

outer material. The calculated attenuation due to water on one this effect decreases the flexural attenuation. The different cases
side of the casing (with vacuum on the other side to prevent any are depicted on Fig. 3 for fluid, slow cement, and fast cement.
leakage) is plotted in Fig. 2 as a function of frequency. For an 8-
mm-thick plate, this curve goes through a broad minimum of 0.31
dB/cm over the frequency range of flat group velocity (100−400
kHz). The fact that the attenuation is approximately constant with
frequency makes estimating it possible by measuring the decay
rate of the envelope of the received broadband pulse.

Fig. 3—Radiation of the flexural wave into water, slow cement ( v p <

2,650 m/s), and fast cement ( v p > 2,650 m/s). Compressional

wavefronts are shown in blue and shear wavefronts are in green.

Fig. 4 shows the theoretical attenuation versus acoustic

impedance of the material in contact with the casing. The blue
curve is relative to a fluid ( vs = 0) with a constant velocity of
1,500 m/s, the red curve is relative to a solid material with a fixed
density (1,500 kg/m3) and a fixed Poisson ratio (0.30, typical of
cements). Below the critical impedance, Zc (Zc ~ 3.9 MRayls),
Fig. 2—Attenuation of the flexural mode due to radiation in water on the attenuation increases linearly with the material impedance
one side of the casing.
(liquid or solid). Above Zc, the attenuation drops rapidly to fairly
small values. A given attenuation (for example 0.31 dB/cm)
This attenuation is inversely proportional to the casing corresponds to two values of the acoustic impedance, typical
thickness, implying, as anticipated, a lower sensitivity for thicker either of liquid or good cement. As a consequence, the flexural
casings. For a fluid, we find that its attenuation is also attenuation alone cannot discriminate between liquid and solid
approximately proportional to the acoustic impedance; however, and needs to be combined with another measurement such as the
for a solid material such as cement bonded to the steel plate, the casing resonance technique.
situation is more complex. Apart from the density, acoustic wave
propagation in a solid material is defined by two velocities, the The effect on attenuation caused by casing-to-cement
compressional and the shear wave velocities. The effect of a solid debonding and a water-filled microannulus has also been
material on the flexural attenuation depends on the value of its investigated. For cement impedance lower than the critical
velocity (compressional v p and shear vs ), compared to flexural impedance, the attenuation is reduced by about 15% as soon as
the casing and cement are debonded. This effect is similar to that
phase velocity v flex at the central frequency (200 kHz for an 8- observed with the pulse-echo technique where the reduction in
mm-thick casing). For the flexural wave to radiate into the cement apparent impedance is of the order of 30%. The attenuation
as a compressional or shear (bulk) wave, the cement reading is not affected by the microannulus width up to ~ 250
compressional or shear velocity must be less than the flexural µm, which is on the same order of magnitude as for the pulse-
phase velocity. At the pulse center frequency being considered, echo technique. Above the critical impedance, a large increase in
this figure is about 2,650 m/s (derived from Snell’s law). For attenuation is observed as soon as the cement is debonded. The
cements, the shear velocity is always less than 2,650 m/s. As a attenuation increases approximately with the cement impedance,
consequence, the shear wave is always radiated into the cement again without significant effect from the microannulus thickness
annulus. However, the cement compressional velocity can either up to ~ 250 µm. This behavior is notably different from the pulse-
be larger or smaller than this critical velocity, depending on the echo technique and reduces the possible confusion between good
cement type and contamination. In the first case, the cement with microannulus and mud.
compressional wave cannot be radiated into the cement annulus
(it is said to be supersonic compared to the flexural wave), and
4 IPTC 10546

comparison, the pulse-echo technique waveform in the same

geometry (also shown in Fig. 5) barely shows the very small
effect of the formation reflection. The strong TIE from the
flexural wave makes it possible to accurately measure the time of
arrival and amplitude. It should be noted that the transit time
difference δτ between the casing arrival and the TIE does not
depend on the transmitter-to-receiver spacing, transducers
standoff with respect to the casing, or properties of the fluid
inside the casing. Instead, the transit time difference is a sole
function of the annulus thickness and wave velocity. Knowing
one quantity allows for computing the other. For example, if the
borehole size is known, the annulus material wave velocity can be
computed: Either the compressional wave velocity for fluids, or
both the compressional and shear wave velocities for a slow
cement, or the shear wave velocity only for a fast cement.
Similarly, if the annulus wave velocity is known from a nearby
in-gauge section, then the size of moderate borehole enlargement
can be estimated.

Fig. 4—Flexural attenuation at 200 kHz versus acoustic impedance

for gas (red), liquid (blue), and solid (brown) materials.

Third-Interface Reflections

The information provided by the flexural attenuation is related to

the state of the material in contact with the casing and does not
probe deeper into the cement sheath. However, the pulse radiated
by the flexural wave packet into the annulus may be reflected by
the third interface, the cement-to-formation interface (with the
casing inner and outer walls being the first and second interfaces
seen by the ultrasonic pulse). The casing is actually fairly
transparent to this reflected pulse and it can be detected by the
receivers with sufficient signal level. Since the third-interface
echo (TIE) propagates through the annulus, it carries information
about the annulus geometry and material, as well as formation
wall geometry and acoustic contrast with the annulus material

Depending on the type of material in the annulus, different

echoes will result: For a liquid annulus, only compressional (P)
waves are supported and there is one type of TIE (PP). For fast
cement, only the shear (S) wave is excited in the cement and
detected as a shear-shear (SS) echo, after conversion into a
pressure wave in the logging fluid. For a slow cement, both
compressional and shear waves are launched and reflected at the
formation wall in addition to partial conversion from P to S (and
vice versa) that occurs upon reflection on the same interface. In Fig. 5—Comparison of pulse-echo and flexural synthetic waveforms
total, three types of echoes are received, PP, SS, or PS/SP (which and their envelope for a 25-mm water-filled annulus between casing
and formation.
arrive at the same time). In all cases, multiple reflections between
the casing and the formation may also occur.
However, TIE may not always be detectable. Factors adversely
A simulation6 of the waveforms computed at a receiver is shown affecting the TIE amplitude are, in order of decreasing
in Fig. 5 for a water-filled annulus with reflection on a hard occurrence:
formation (the large fluid-formation contrast provides for a large
TIE amplitude). As shown, the TIE has a large amplitude • Casing eccentering within the borehole. Along the
compared to the casing arrival and is temporally compact. For direction of eccentering, the casing wall and the
IPTC 10546 5

formation wall are parallel. However, in other directions, (30 mm). The vertical resolution of the attenuation measurement
they are not parallel, and the TIE is not reflected in the is controlled by the inter-receiver spacing (102 mm), while the
direction of incidence, leading to an amplitude drop (Fig. resolution of TIE-related measurements is controlled by the
6). transmitter-to-receiver spacing, at about 150 mm. The azimuthal
sampling is fixed at 10 degs and the vertical sampling can be
• Attenuation in the annulus material. Although water or adjusted, typically between 75 and 150 mm.
uncontaminated set cements have low attenuation, heavy
muds or contaminated/unset/foamed cement may have Because fluid properties can be estimated directly from the two
high attenuation. sets of measurements, no dedicated hardware for separate fluid
properties measurement is required.
• Acoustic contrast between annulus material and
formation. If the contrast is low, then the TIE amplitude
is low. This is true for example between lightweight
cement and shales. On the contrary, in double strings, the
contrast is very high, leading to rather strong TIE.

• Roughness of the borehole, at the scale of the acoustic

wavelength or larger, will reduce TIE amplitude.

• Large distance (over ~ 75 mm) between the casing and

formation, due to large hole or washouts, will move the
TIE outside of the recorded window and outside the
optimal transmitter-to-receiver spacing.

Fig. 7— Sketch of the rotating subassembly supporting the four

ultrasonic transducers.

Data Processing

The first objective of the data processing is to provide a robust

interpreted image of the material in contact with casing. To obtain
this image, the cement impedance data from the normal incidence
transducer and the flexural attenuation computed from the
amplitude of the casing arrivals on the near and far flexural
receivers are utilized. These data inputs are independent
measurements linked to the properties of both the inside fluid and
Fig. 6—Geometry of an eccentered casing in the borehole the outside medium through an invertible relation. They are first
combined to eliminate the effect of the fluid inside the casing,
thus eliminating the need for specific hardware for fluid
Tool Description properties measurements. The output is produced in the form of a
solid-liquid-gas (SLG) map displaying the most likely material
The architecture of the new Imaging Behind Casing* tool, is state behind the casing. The material state is obtained for each
highly similar to the USI tool. The most visible difference is a azimuth by locating the two measurements (the acoustic
new rotating sub-assembly supporting four transducers. To
impedance and the flexural attenuation, corrected for the effect of
optimize the distance from the transducers to the casing and
the inside fluid) on the SLG map giving the area covered by each
minimize mud attenuation, three subassembly sizes are provided
state. This map is computed in an initialization step before the log
to cover the casing range from 4.5 to 9.675 in. The normal
and uses a priori knowledge of the possible materials:
incidence transducer is oriented at 180 degs from the three
flexural transducers. The latter are disposed such that one • Gas is defined as a very low-impedance material,
transmits and two receive the returned signals, providing in independently of any input.
particular an amplitude attenuation measurement.
• Liquid is defined as a medium with the expected
The flexural transducers are operating at about 200 kHz, and
acoustic impedance of the mud displaced by the cement,
their azimuthal resolution is similar to the pulse-echo transducer
with some provision for possible deviations from this
* value.
Mark of Schlumberger
6 IPTC 10546

• Solid is defined through the expected type of cement. 2 and 10 m. The total width of these areas is also evaluated and
Using a laboratory-measured database, this material displayed as a single curve, which indicates the channeling
selection is converted into acoustic properties according severity.
to Table 1, and provisions are made for some
contamination or cement not fully set. Because the SLG map pertains to material immediately behind
casing, the second objective of the processing is to extract
relevant information from the TIE and quantify the full annulus
between the casing and the formation. First, the TIEs following
Cement Density P Velocity Z (MRayls) casing arrival are identified on the waveform envelope, and their
(kg/m3) (m/s)
arrival time and amplitude are determined. From the time
Class G 1,800 3,000 5.4 difference between TIE and casing arrival, and provided enough
azimuthal coverage is available, it is possible to extract the casing
Light Cement 1,200 2,800 3.4 centering in a dimensionless unit, where 100% represents perfect
Ultralight Cement 900 2,800 2.5
centering, and 0% represents fully eccentered casing (contacting
the formation). If the borehole diameter is known, then the time
Table 1—Acoustic properties of uncontaminated set cements. difference between casing arrival and TIE arrival can be further
converted into material velocity and displayed in map form.
The next step of the processing is to predict the measurements
from the expected acoustic material properties, which is trivial for Log examples
the acoustic impedance, but requires using a model for the
The first example (Fig. 9) is from a test well with 7-in., 23-lbm/ft
flexural attenuation. Then, multiple realizations of the
casing inside of a 9.625-in. casing and Class G cement. The
measurements noise are added to generate three clouds of points
section shown straddles the cement top. The left three tracks
(solid, liquid, and gas) in the two-dimensional measurements
display the SLG map and the two data sets used to compute the
plane. From these clouds of points, it is possible to define an
map, the flexural attenuation, and the pulse-echo cement
occurrence probability for each state (solid, liquid, gas), and the
impedance maps. These two maps are strikingly different; i.e., the
measurement plane can be mapped into different regions (Fig. 8)
flexural attenuation identifies two distinct, well-defined materials
by color. The white area corresponds to inconsistencies between
with high and low impedance. The impedance map, somewhat
the measurements, and may occur for example at collar locations.
affected by formation reflections from the outer casing, is barely
able to distinguish such regions. The interpretation is that higher
impedance cement partly displaced lower impedance cement
between the 330- and 290-m depths. The higher impedance
material is above the critical impedance and has a low
attenuation, actually lower than water, as seen from above the
cement top. The lower impedance material, either by design or
due to contamination, is below critical impedance and its high
attenuation leads to the high contrast. Both measurements are
combined into the SLG map, and a solid is shown where it
belongs, for both cement types, thus simplifying the display and
the low-level interpretation. Collars appear as white streaks
because the two measurements are disturbed enough to be
inconsistent. It can be noted that the flexural attenuation is much
less affected by formation reflection than the impedance map, as
long as the two casings are not touching. Contact actually occurs
at 255 m, and the attenuation shows some effect.

The four tracks on the right are related to the TIE information.
Track 4 is the map of TIE transit time difference with the casing
arrival. The white zones correspond to difficulty in the automatic
TIE picking due to the effects of the annulus material and or the
casing centering. This track is used to identify the sections with
high eccentering and the location of the near side, which is
consistent with the “galaxy” patterns seen on the impedance map.
Fig. 8—SLG mapping of the measurement plane for a Class G The transit time data are used to compute the casing centering
cement. Z is the impedance estimated from the pulse-echo technique, shown in track 7. This centering varies from ~100% (perfect
whereas the attenuation pertains to the flexural wave.
centering) to 0% (contact) within a short distance. Track 5 is the
TIE amplitude map and is affected by casing centering (which
The channel map is a further simplification of the measured affects the reflection geometry at the annulus-to-formation
SLG map focusing on the potential cementing problems. It shows interface and the propagation length inside the annulus) such as in
only the continuous uncemented areas with significant vertical the free-pipe section but also by the material type as seen in the
extent. This vertical extent can be adjusted, for example, between
IPTC 10546 7

two cements region. Casing collars from the outer casing string (tracks 5 to 8) answer the question and illustrate the potential
can also be located, for example at 315 and 325 m. Track 6 is the benefits in complex situations. The annulus velocity data show
computed annulus velocity map that is obtained from the TIE that the section xx68–xx73 m has a velocity ranging from 1,800
time map and the knowledge of the outer casing inner diameter. m/s to 1,900 m/s, clearly outside the range of plausible liquid, and
From the top, a value of 1,500 m/s, characteristic of water, is corroborating the SLG map. Furthermore, this velocity map and
shown, gradually increasing until 300 m where it reaches 2,000 to the amplitude map indicate that this solid material is not
2,500 m/s, consistent with the compressional velocity of a slow homogeneous but is three layers of slightly different materials. At
cement. Below 300 m, the TIEs are mostly propagated as shear, the top (xx30−xx34 m) of the liquid section, the velocity again
and the measured velocity drops back to ~1,500 m/s, consistent reaches 1,900 m/s, indicating the presence of a solid, although
with the shear velocity of a fast cement. even the flexural attenuation barely shows a slight increase above
its fluid values. This solid explains the nonflat upper interface,
The free-pipe section immediately above the cemented interval and since it floats on top of water, its density is probably less than
of the same well is shown in Fig. 10. The white line in the middle 1.0 g/cm3 and the targeted cement density. A small amount of
of the SLG map track is due to a casing groove caused by cable segregation within the cement slurry is a likely explanation for
motion, while the red patches along the same azimuths are due to such solid. The casing centering (track 8) is perfect near the
contact between the two casings. These data can be verified on casing centralizer, but 10 m below that depth quickly drops to
the right tracks that display the flexural waveforms along two about 50 %.
orthogonal diameters. The first echoes from the center are the 7-
in. inner casing arrivals, while the following ones are TIE Fig. 14 is a section from a well with 9.625-in., 47-lbm/ft casing
reflections from the outer 9.625-in. casing. The faint echoes after inside a 12.25-in. borehole. The mud density is 1.30 g/cm3, and
the first TIE are multiple reflections (up to five) within the the cement is Class G with a slurry density of 1.90 g/cm3. The
annulus. Such a display provides a geometrical sketch of the SLG map (track 1) indicates a uniform cement across the zone.
casing strings where contact points or potential geometrical The acoustic impedance ranges from 5 to 7 MRayls with some
defects can be identified. indication of formation reflections at about x810 m. The flexural
attenuation (track 2) exhibits both high (0.9 dB/cm) and low
Fig. 11 is an example of a 9.625-in., 47-lbm/ft single-string values (0.5 dB/cm) with the low values matching the high-
casing in a vertical well cemented with a high-performance, impedance area. The cement impedance value is close to the
lightweight cement (1.08 g/cm3). The mud density is 1.3 g/cm3. critical impedance value where the attenuation peaks. The TIE
The impedance image (track 5) is affected by the low impedance time map (track 4) confirms that the “galaxy” pattern shown at
of the cement and by casing roughness such that locating a depth X810 m is indeed due to the closer proximity of the
channel is made virtually impossible. This situation can be borehole wall. The higher impedance at x725 m and 180 degs
contrasted with the flexural attenuation (track 4) map that shows azimuth on track 3 can also be related to a formation reflection
a channel between x465 and x480 m with excellent contrast. The effect. The casing centering curve (track 8) is 100 % near the
processed SLG map (track 3) maintains the same quality with a centralizers, but eventually drops to below 50 % in the areas
clear liquid channel embedded within the cement. The SLG map between them. The small oscillations on the time map and the
has been further processed by the hydraulic communication centering curve are artifacts created by a corkscrew shape of the
algorithm to produce the channel map (track 6) and the channel borehole with a period of about 2 m. The annulus velocity map
width curve (track 7). This processing cleans the SLG map by (track 7) displays a rather uniform value around 1,750 m/s,
discarding small liquid patches and only keeping the liquid- characteristic of the shear velocity of class G cement. One
connected channels with a significant vertical extend, 2 m in this possible exception is the blue stripe at x775 m. Rather than a
case. A polar plot of the flexural waveforms in a variable density lower cement velocity, this is more likely due to a slight borehole
fashion displays the geometry of the casing within the borehole. enlargement. Assuming a constant cement velocity of 1,750 m/s
An inspection of the curvature of the TIE detected within the would give the possibility to present a cement sheath thickness
channel (Fig. 12) reveals that despite a casing centralizer at x474 over the interval. The TIE amplitude map (track 6) exhibits a
m the casing is slightly eccentered in the borehole and the channel striking correlation with the gamma ray (GR) (track 5) low
is located on the narrow side, contrary to first intuition. The amplitudes in shaly zones, as can be expected from a lower
absence of TIE across the cement azimuth may be due to a low impedance contrast between the cement and shales. Apart from
acoustic contrast between cement and formation. The CBL curve casing collars and centralizers, this map is also affected by casing
(Fig. 11, track 1) shows a fairly high reading due to the low centering that creates the two dark vertical stripes seen at about
cement impedance and a water-filled microanulus. x810 m and above x730 m.

Fig. 13 shows the top portion of the same well in a double

string section with 13.325-in. outer casing. The acoustic Conclusions
impedance map (track 4) indicates free pipe up to xx30 m, with
air above. Closer examination suggests a slightly higher reading The new cement evaluation tool implements both the traditional
between xx68 and xx73 m. The flexural attenuation map (track 3) pulse-echo technique and the new flexural wave concept. This
and the subsequent SLG map clearly identify this zone as being combination reduces the ambiguity in evaluating the material
solid, which is further confirmed by the dip of the CBL curve in immediately behind casing, further than any of the two techniques
this zone. The presence of a nonflat interface at xx30 m, between taken alone. The processed measurements are displayed in easy-
the alleged liquid and air, raises a question; however, the TIE data to-read formats such as a SLG map or a channel map. In an
8 IPTC 10546

unprecedented way, the flexural mode enables deep imaging of High-performance Lightweight Cement Slurries for
the cement sheath up to the cement-formation interface. Improved Zonal Isolation in Challenging Situations,”
The potential this tool holds has been demonstrated in various paper 47830 presented at the 1998 SPE/IADC meeting,
field tests. Enhanced contrast between light-weight cement and September.
4. Al-Suwaidi, A. et al.: "Light as a Feather, Hard as a
the displaced mud significantly improves data evaluation. When Rock," Oilfield Review (2001), 13, No. 2, 2-15.
reflections from the formation are detected, completely new 5. S. Zeroug and B. Froelich: “Ultrasonic Leaky-Lamb
measurements are possible, including the casing position within Wave Imaging through a Highly Contrasting Layer,”
the borehole, estimation of the cement velocity or the borehole Proc. 2003 IEEE Ultrason. Symp., November, 794-798.
shape, or imaging the formation behind casing. 6. S. Zeroug: “Forward Modeling for Ultrasonic Leaky-
Lamb Wave based Imaging through a Highly Contrasting
References Steel Cylindrical Layer,” Proc. 2004 IEEE Ultrason.
1. Hayman, A.J., Hutin, R., and Wright, P.V.: ”High Symp., August.
Resolution Cementation and Corrosion Imaging By
Ultrasound,” paper KK presented at the 1991 SPWLA
32nd Annual Logging Symposium, June 16-19.
2. Moulin, E., Revil, P., and Jain, B.: “Using Concrete
Technology to Improve the Performance of Lightweight
Cements,” paper 39276 presented at the 1997 SPE/IADC
meeting, November.
3. Revil, P. and Jain, B.: “A New Approach to Designing
IPTC 10546 9

Fig. 9—Top section of the cement in a test well with 7-in. casing inside of 9.625-in.
10 IPTC 10546

Fig. 10—Free-pipe section in a test well with a 7-in. casing inside a 9.625-in. casing
IPTC 10546 11

Fig. 11—High-performance light weight cement section in a 9.625-in. single-string casing

12 IPTC 10546

Fig. 12—Polar plot across channel from the far receiver at a depth of 477 m.
IPTC 10546 13

Fig. 13—A 9.625-in double casing section above the high-performance, light-weight cement
14 IPTC 10546

Fig.14 9.625-in. casing within a 12.25-in. borehole with Class G cement with indications of centralizer positions.