SPE/IADC-194135-MS

A Novel Mechanical Tool for Annular Cement Verification

Jesus De Andrade, Sondre Fagerås, and Sigbjørn Sangesland, Norwegian University of Science and Technology

Copyright 2019, SPE/IADC International Drilling Conference and Exhibition

This paper was prepared for presentation at the SPE/IADC Drilling International Conference and Exhibition held in The Hague, The Netherlands, 5-7 March 2019.

This paper was selected for presentation by an SPE/IADC program committee following review of information contained in an abstract submitted by the author(s).
Contents of the paper have not been reviewed by the Society of Petroleum Engineers or the International Association of Drilling Contractors and are subject to correction
by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers or the International Association of Drilling Contractors,
its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers or
the International Association of Drilling Contractors is prohibited. 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 SPE/IADC copyright.

Abstract
The annular casing cement is an important part of the well barrier throughout the life cycle of a well. With
the increasing number of subsea plug and abandonment (P&A) operations, increased attention is now given
to annular cement evaluation and the ability to prove adequate zonal isolation. Today, cement evaluation by
logging is performed almost exclusively using acoustic logging tools. One of the concerns when it comes
to cement integrity is the frequently occurring micro-annulus at the casing-cement interface. Yet, a typical
cement evaluation tool may lack of accuracy on its evaluation. Hence, a novel concept for the evaluation
of casing-cement micro-annulus has been proposed.
This paper describes a mechanical-based approach for cement evaluation – the Annulus Verification Tool
(AVT). The AVT applies a radial force on the casing inner wall that yields an ovalization of the cross-section,
while recording the radial displacement of the casing. A prototype of the AVT has been constructed along
with an experimental setup to allow for initial testing of the tool. This comprises the construction of full-
scale diameter samples representing a typical 9 5/8-in. production casing cement job, with the possibility
to generate a uniform micro-annulus of a known size at the casing-cement interface.
The tests performed have shown that the AVT is able to differentiate a casing supported by an annular
cement sheath from a free pipe, due to the stiffness contrast. By measuring the casing radial displacement
with high resolution, the results have shown that a microannulus can be detected and its size quantified
with good accuracy. Experimental tests performed with tool eccentricity and tilting has shown that the AVT
should be kept centralized to achieve accurate quantification of the microannulus size.
The AVT module is meant to complement the acoustic tool sting used today, and to improve evaluation
of the cement sheath's sealing capability, especially in cases where a micro-annulus is detected or suspected.
If an existing microannulus is suspected, the AVT logging response may confirm its occurrence, quantify
its size and aid the planning of remedial operations to restore the annular barrier.

Introduction
The annular casing cement is an important part of the well barrier throughout the life cycle of a well. The
main purpose is to provide mechanical stability to the wellbore and to ensure zonal isolation through the
entire well service lifetime. Failures to achieve a proper primary cementing, and to ensure long-term sealing
2 SPE/IADC-194135-MS

capabilities of the cement sheath, may severely limit the ability of a well to reach its planned life cycle.
Often, in P&A operations, the annular cement sheath is also evaluated for serving as a barrier in an eternal
perspective (Oil and Gas UK OP-071, 2012). In these cases, it is especially important to assess the status
of the cement sheath and evaluate potential cement degradation over time.
As several of the major offshore oil and gas fields are now maturing, an increase is seen in offshore
P&A operations. It has been estimated that, on the Norwegian Continental Shelf (NCS) alone, over 3,000
wells need to be plugged and abandoned (Myrseth et al., 2017). When also considering the environmental
responsibility associated with P&A of oil and gas wells, it is clear why reducing the operational expenses
as well as ensuring proper integrity of the plugged wells is a high priority for the stakeholders involved.
According to the widely referred Norsok D-010 (2013), downhole verification of the cement sheath
barrier is mandatory in most cases when the casing approaches the source of hydrocarbons inflow.
Concerning P&A, the barrier needs to be placed adjacent to an impermeable formation with sufficient
formation integrity, and extend across the full cross section of the well including the annulus. This means
that, if an annular barrier cannot be verified, it has to be restored by performing operations behind the casing.
Such operations will contribute to an increased time and cost spent on the P&A. Contrary, if an annular
barrier can be verified, significant time can be saved by only performing plugging operations within the
casing. In this case, placing a plug inside of the casing is sufficient to establish a cross-sectional barrier.
This can also be an enabler for performing more of the P&A without a drilling rig, further reducing the
costs and the operational footprint.
Therefore, increased attention is now given to cement sheath evaluation and the ability to prove adequate
zonal isolation. Hence, the objective of this work is to contribute to an improved evaluation of the annular
cement sheath integrity, in particular for cases where an inner microannulus is suspected. This is done by
introducing, testing and evaluating a novel tool for cement evaluation.

Cement evaluation by logging

Cement evaluation is undertaken to determine whether the casing cement meets the structural and zonal
isolation requirements. Verification by logging is one of these evaluation processes, regularly required for
critical cement jobs (Norsok D-010, 2013). An example is presented in Fig. 1, where the same annular
casing cement is part of both the primary and the secondary permanent barrier for a plugged well. Hence,
the annular cement shall be verified by logging.
Today, cement evaluation by logging is performed almost exclusively using acoustic logging tools
conveyed by wireline. The acoustic logging techniques are acknowledged for both locating the top-of-
cement (TOC), as well as for evaluating the hydraulic seal of the annular cement. There are several acoustic
logging techniques to evaluate the cement job quality, including the most commonly used sonic cement
bond logs (CBL), ultrasonic pulse-echo logs and flexural ultrasonic logs (Pilkington, 1992; Ellis and Singer,
2007). These techniques have been developed and improved since the 1960's to quantify the bonding at the
cement-to-casing interface, and also the cement-to-formation to some extent. The underlying principle of
cement bond logging is that the strength of the casing signal recorded at the receiver is a function of the
material surrounding the casing. The tool measures the CBL amplitude and is therefore affected by both
the coupling- and propagation attenuation.
SPE/IADC-194135-MS 3

Figure 1—An example of a critical cement, where the production casing annular cement is
part of both the primary and secondary permanent barrier envelope (Norsok D-010, 2013).

Recent versions of ultrasonic tools can now detect the presence of channels and voids within the cement
sheath (Bellabarba et al., 2008). According to Morris et al. (2007) the main advantage the CBL tools have
over the ultrasonic tools is the minor effect of the acoustic properties of the borehole fluids. The CBL tools
may be run in heavy-mud – e.g. density above 15 lbs/gal – in which the use ultrasonic logs is limited.
The CBL tools provide qualitative information about the bond to the formation, where the ultrasonic logs
are limited. The ultrasonic tools have two main advantages. First, they are capable of better vertical and
radial resolution due to a rotating transducer of the tool. This provides the interpreter an improved picture
of what the cement sheath looks like. Second, the flexural ultrasonic logs do not depend on specific acoustic
properties of the cement, which makes them more suitable for light-weight cements.
Therefore, to reduce uncertainty in the evaluation of the cement barrier, a modern cement evaluation
logging string includes an ultrasonic flexural tool, an ultrasonic pulse-echo tool, and a sonic CBL/VDL. This
is reflected in Fig. 2, where the advantageous characteristics (green) and the disadvantageous characteristics
(red) of the acoustic cement evaluation tools are illustrated. This categorization in the figure is based upon
the features and discussions found in the references presented along this section, as well as from Hayden
et al. (2011) and Catala et al. (1991).
4 SPE/IADC-194135-MS

Figure 2—Summary and categorization of the characteristics of the acoustic cement evaluation
methods. The combination of the three techniques is shown in the rightmost column.

Nevertheless, it has been consistently stated by Operators that the lack of accuracy in logging tools may
be a challenge. There is a need to increase the reliability of cement bond logs to quantify the bond to the
casing and to the formation (Vignes, 2011). Moreover, there are issues related to the standardization of the
logging tests, of which the outcomes can vary among logging service companies, engineers performing the
test and interpreting the acquired data. An example is the interpretation of zone isolation with ultrasonic
tools that requires the input of the acoustic impedance of the fluid between the cement and casing (Morris
et al., 2007), given the case that an inner microannulus exists. As a result, incorrect input settings to the tool
due to uncertainties on the presence of gas-contaminated cement, gas- or liquid-filled microannulus, may
lead to erroneous interpretations of the cement quality.

Response to casing-cement microannulus

Traditionally, a microannulus at the casing-cement interface has been considered challenging for the
industry-standard logging tools because it hinders evaluation of the cement sheath behind the microannulus.
For instance, a high-quality annular cement sheath of high acoustic impedance may be present, but with a
separation of a few microns between the casing and cement, a CBL may classify the cement job as poor,
SPE/IADC-194135-MS 5

or even as free pipe. Moreover, the response of acoustic cement tools may be dependent on the fluid in
the microannulus, either liquid- or gas-filled. In that sense, some relevant considerations are summarized
in Table 1.

Table 1—Summary of the microannulus response considerations for

acoustic cement logging, mostly based upon Jutten and Hayman (1993).

Liquid-filled microannulus Gas-filled microannulus

CBL • Experiments have shown that for a liquid-filled
microannulus, the propagation attenuation is similar to
that of free pipe for common sizes of microannulus.
• CBL amplitude does not change much with the change
in microannulus size due to a fairly constant shear
coupling attenuation. However, it is significantly higher
than the well-cemented value due to the lost propagation
attenuation.
• The microannulus effect is even more dramatic for a gas-
filled microannulus.
• Coupling attenuation and the propagation attenuation
rapidly approach that of free pipe as the size of the
microannulus increases (greater than 5-10 μm).

Ultrasonic pulse- • The ultrasonic tools do not rely on a shear coupling
echo between the casing and the cement, as the wave motion
is, in theory, normal to the casing wall.
• However, experimental studies have shown that
this technique is fairly insensitive to a liquid-filled
microannulus, up to a microannulus size in the order
of 100 μm. For greater microannulus sizes, it becomes
difficult to distinguish the cement from liquids.
• The tools are far more affected by a microannulus if it is
gas-filled. Gas has an acoustic impedance (AI) close to
zero and acts like a barrier towards ultrasound.
• For microannulus sizes approaching 10 μm, the AI
reading looks as for a gas-filled annulus. Even a small
gas-filled microannulus (1 μm) makes the measurement
unable to detect an annular cement sheath.

Flexural ultrasonic • The response of the flexural ultrasonic logging tools to a • The response of the flexural ultrasonic tools to a gas-
liquid-filled microannulus has been shown to be similar filled microannulus is similar to that of the ultrasonic
to that of the pulse-echo tools (Kuijk et al., 2005) pulse-echo tools (Hayden et al., 2011).
• The flexural attenuation measurement is not significantly
affected by a liquid-filled microannulus up to about 250
μm.

Hence, detecting a microannulus is typically not a challenge with the existing technology. However, none
of the tools is able to quantify with accuracy the size of the inner microannulus, and neither the detrimental
effect it may have on the zonal isolation of a well section.

The Annulus Verification Tool (AVT)

In this work, a mechanical concept for cement evaluation hereby called Annulus Verification Tool (AVT),
has been considered to complement the acoustic tools used today. It is meant to provide additional insights
into the cement sheath verification process, especially in cases where an inner microannulus is detected or
suspected.
The AVT concept comprises the assessment of the mechanical response in a cased wellbore section
towards an imposed mechanical load. Detecting and quantifying the size of a potential microannulus will
increase the understanding of the situation, giving a foundation to take better decisions regarding the cement
sheath sealing capability and potential remedial solutions.

Description
The principle behind the AVT is illustrated in Fig. 3a. A mechanical force is applied radially to the inner
casing wall, while the displacement of the casing wall is recorded. This gives a measurement of the stiffness
of the casing and the surrounding material as the load is applied. A linear response is expected as long as
the deformations are within the linear elastic region of the materials. In that sense, it is anticipated that a
casing supported by an annular cement sheath has a greater stiffness than an uncemented casing.
In the case of a well-cemented casing, the combined stiffness of the casing, cement, and the surrounding
formation should be reflected in the dataset, as illustrated by the blue line in Fig. 3a. For an uncemented
casing, the stiffness should be related to the casing string alone since it is not backed by the cement sheath,
6 SPE/IADC-194135-MS

as illustrated by the green line Fig. 3a. In the case of a casing-cement microannulus, the casing will be
unsupported as displacement starts, such that only the casing stiffness will be experienced for the initial
displacement. However, as the force and displacement increase, the casing is pushed towards the cement
and contact will be obtained as the microannulus is closed. At this point, the stiffness of the system will
increase from only the casing stiffness to the combined stiffness of the casing, cement, and formation (as
for a well-cemented casing). This is illustrated by the red line. The zone where contact between casing and
cement is obtained and an increase in stiffness is observed yield an estimate of the microannulus size.
A schematic of the AVT concept is shown in Fig. 3b. The tool consists of two radial pads with a curvature
matching that of the inner casing wall. The two pads are spaced 180° apart and each pad is mounted on a
piston. As the tool barrel is internally pressurized, the pressure will act on the two pistons, making the pads
move towards the casing. When the pads contact the inner casing wall, an increasing pad force is applied
as the pressure is further increased.

Figure 3—(a) Measurement principle of the AVT (b) Schematic of the AVT concept.

To be able to quantify the size of a microannulus, the casing displacement has to be measured with
high accuracy and resolution. A linear variable differential transformer (LVDT) has been found suitable
for this purpose. It is environmentally robust, friction-free and inherently has infinite resolution (National
Instruments, 2018). However, the resolution is limited by noise, in addition to the signal conditioning
equipment used. The two moving pistons are coupled mechanically to the LVDT through a core-coil
assembly. One of the pistons can be coupled to the LVDT core, while the other can be coupled to the LVDT
coil. The motion of the core relative to the coil assembly gives rise to a changing differential voltage, which
is a linear function of the motion. Since the two pistons coupled to the LVDT are also coupled to the tool's
pads, the LVDT can be calibrated to give the distance between the two pads, corresponding to the outer
diameter of the tool. The diameter can then be recorded as the force is applied to the casing wall, enabling
a measurement of the diametral displacement of the casing.

Experimental Testing
A prototype of the AVT has been constructed and a laboratory setup designed to allow for testing of the
tool. This includes construction of full-scale diameter samples with the aim of representing a typical 9 5/8-
in. production casing cement job, with the possibility generate a uniform microannulus of a known size at
the casing-to-cement interface.
SPE/IADC-194135-MS 7

Preparation of test samples

The samples consisted of two concentric pipes representing the formation and the casing, while the annular
volume between the two was filled-up with a cement slurry. In Fig. 4, images concerning the construction
of the full-scale diameter samples are presented.
The inner pipe corresponded to a 9 5/8-in. 53.5 lb/ft P-110 casing, with an inner diameter of 8.535 in.
For representing the formation, an outer carbon steel pipe with 12.75 in. OD, thickness 5/16 in. and grade
S355J2H was used. This resulted in an annulus between the casing and outer pipe with a thickness of 1.25 in.
Portland API Class G cement was mixed with a water-cement weight ratio of 0.44, according to API Spec
10A (2009). The cement slurry was poured into the annulus through a funnel and distributor with several
hoses to minimize air entrapped and ensure proper azimuthal distribution of the cement slurry.
To be able to construct samples with a controllable microannulus at the casing-cement interface, some
modifications were implemented to the casing. A schematic of the setup used for microannulus generation
is shown in Fig. 4a. The outer casing wall was machined with a chosen taper, from the original OD at the
top to a smaller OD at the bottom. A flange was welded onto the top of the casing, as shown. Each bolt fits
in a pit on the top surface of the outer pipe to ensure casing centralization during cementing and to minimize
rotation and tilting of the casing, as it is lifted/lowered.
The microannulus generation procedure comprised, first, the use of a hydraulic puller to break the casing-
to-cement bond by pulling the casing apart from the cement sheath and outer pipe. An oil-based form release
agent was applied on the casing outer wall prior cementing, to reduce the adhesion between the casing and
the cement sheath. Next, the bolts were used to hoist and lower the casing as preferred, and the relative
position of the casing compared to the initial "cemented-in" position can be measured to further calculate
the microannulus size (Fig. 4b).
A heating cable was wound onto the outer pipe and a thermistor was mounted at the top surface of the pipe
to monitor and control the temperature. The outer pipe was pre-heated to the chosen curing temperature,
before cementing. The temperature was kept constant, at 40°C, during cement hydration and subsequent
logging tests to avoid any contraction/expansion effects due to temperature variations.

Figure 4—(a) Setup for generation of a casing-cement microannulus of a known size, (b) controllable microannulus
by hoisting the tapered casing, (c) visualization of annular cement sheath by pulling the casing after tests.

AVT prototype and testing details

For proof-of-concept, a prototype was designed and constructed to log a 9 5/8-in. 53.5 lb/ft casing, as the
tool pads match the curvature of an 8.535-in casing ID. The tool body is hydraulically sealed by the use of
8 SPE/IADC-194135-MS

different sealing elements and is rated to an internal pressure of 3,500 psig, limited by electrical feedthrough
connectors. Static O-ring seals are used between flanges, while dynamic rod seals are installed to provide
a seal between the piston housings and the moving pistons.
Fig. 5 depicts the AVT prototype mounted on a cradle-shaped holder. The holder was designed such that
the tool center is positioned at the midpoint of the test samples height, to minimize end effects during tests.

Figure 5—AVT prototype (left) setup and (right) centrally positioned for logging of a cemented sample.

A high-pressure positive displacement pump was used to pressurize the AVT. The pump had two
cylinders, providing a pulse-free, controllable rate. The differential pressure transducer was mounted on the
tool's barrel to provide a direct measurement differential pressure exerted on the pistons, as the tests were run
at ambient surroundings and the gauge pressure is known. A LabVIEW-program was built to both display
and record the pressure and diametral displacement, using the output current from the pressure transducer
and the output voltage from the LVDT signal conditioner, respectively.
Several test cases were studied to assess the response of the AVT prototype towards different
characteristics of the casing and annular cement sheath, as well as the relative position between the prototype
and casing, indicated as follows:

• Uncemented casing: Before cementing a sample, the casing was logged to establish the response
for the uncemented case (free pipe). This sample is then cemented and used to characterize a well-
cemented casing.
• Well-cemented casing: This sample did not undertake any actions for microannulus generation, to
be as representative as possible of a properly cemented casing. Thus, after pouring the cement and
waiting for the hydration process to occur, an AVT test was performed at the same location as for
the uncememented casing case.
• Casing-cement microannulus: Another sample was built using the required setup for microannulus
generation, i.e. machined casing taper and form release agent on the casing outer wall. Here,
the casing was also logged before cementing to establish the stiffness of free pipe. After cement
hydration, an AVT test was performed to establish the stiffness of the well-cemented condition.
To ensure minimum damage to the cement sheath for later tests, the pressurization was stopped
when a stiffness representing that of a good cement sheath was clearly identified. After breaking
the casing-cement bond using a hydraulic puller, the microannulus size was increased in steps by
alternately tightening the bolts. For each microannulus size, logging was performed with the AVT
prototype.
• Eccentricity and tilting: To investigate the effect of poor centralization on the AVT measurements,
representative tests of improper alignment between the tool's pads and the casing wall were
performed with the well-cemented casing sample. Fig. 6 depicts the four scenarios investigated.
SPE/IADC-194135-MS 9

Figure 6—Experimental testing procedure to evaluate the effect of tool eccentricity and
tilting. The red arrows represent the direction in which the tool was decentered/tilted.

Results
Well-cemented vs uncemented casing
Fig. 7 shows the results from logging the same casing, in the same direction, before and after it was
cemented. Except for the initial displacement region, in which the radial pistons travel counter to the internal
springs and the pads accommodate against the casing inner wall, a clear linear trend is seen in both cases,
as expected. This confirms that the deformation is mostly elastic. Linear least squares regression was used
to determine the overall stiffness of the system, represented by the slope of the best-fit straight line. The
stiffness of the well-cemented casing and the uncemented casing were 59.3 kN/mm and 37.4 kN/mm,
respectively.

Figure 7—Results from logging the same casing before and after cementing. Sample
constructed according to the improved set-up using steel to represent the formation.

For the initial displacement region, after pad contact is first obtained with the casing wall, a stiffness less
than the one represented by the straight line is experienced. For both tests, significant pad contact appears to
occur at a diameter of about 219.07 mm, while the full stiffness (straight line) is not seen before reaching a
diameter of about 219.20 mm. This might be caused by a slight mismatch in the alignment between the pads
and the casing. This would result in partial contact between the pads and the casing for the initial contact.
However, as the pressure is increased, the pads and casing would begin to deform due to the developing the
10 SPE/IADC-194135-MS

stresses and result in greater contact areas. This would then give an apparent gradual increase in stiffness
before the full stiffness (i.e. the straight line) is seen when full pad-to-casing contact is obtained.

Casing-cement microannulus
Fig. 8a shows the results from the experiments conducted to establish the free pipe and well-cemented
casing (no microannulus) response with the setup used for microannulus generation. As expected due to
the machining of the casing taper, the stiffness for both cases is somewhat lower than what was found in
the previous test case. The well-cemented case shows a greater stiffness than the uncemented case; 45.6
kN/mm compared to 27.4 kN/mm. Again, some differences on stiffness may be noticed during the build-
up of pad-to-casing contact.
Fig. 8b shows the result from logging a case with 175 μm microannulus. Initially, a more or less constant
slope representing an uncemented casing is seen. This stiffness is somewhat lower than the free pipe stiffness
shown in Fig. 8a, possible due to the casing taper giving a reduction in wall thickness when the casing is
hoisted –while the tool is kept in the same position relative to the outer casing. After some displacement,
an increase in slope is seen; indicating that contact between the casing and the cement sheath is obtained.
The stiffness is then increasing towards the stiffness of a well-cemented casing.
The changes in stiffness, from zero to the free pipe stiffness and then to the well-cemented stiffness, occur
through a gradual increase and not as a sudden change in slope. Thus, picking the diameters d1 and d2 to
characterize the free casing stiffness and then calculate the microannulus size, becomes less straightforward.
To avoid random noise affecting the selection of the diameters, they were picked as the points where the
pad force data differs from the straight line with more than 0.15 kN. Thus, for the dataset in Fig. 8b, the
resulting microannulus size is given by the calculation (d2-d1)/2, equivalent to 170 μm.
For consistency, the same procedure was used for all microannulus sizes logged. Fig. 9 shows the resulting
microannulus sizes found from the AVT measurements are plotted against the "known" microannulus sizes
calculated from the casing taper. From the dataset, it is noticeable that the AVT quantifies the size of a full-
microannulus with an uncertainty of approximately ±10 μm.

Figure 8—Results from experiments conducted to establish (a) the free pipe and well-cemented casing
(no microannulus) response for the microannulus setup, (b) from logging a 175 μm microannulus.
SPE/IADC-194135-MS 11

Figure 9—Resulting full-microannulus sizes from the AVT measurements plotted versus the "known" size
calculated from the casing taper. The black straight line represents the ideal case of the two sizes being identical.

Tool eccentricity and tilting

Fig. 10 shows the results from logging the same well-cemented casing sample with the tool properly
centralized, eccentric in the direction of the pad movement, eccentric in the direction perpendicular to the
pad movement and with the tool tilted. It is noticeable that eccentricity in the direction of the pad movement
does not significantly affect or disturb the AVT measurement.
The case of eccentricity in the direction perpendicular to the pad movement, however, proved to have a
significant impact on the measurement. In this case, both pads obtained casing contact simultaneously, but
at a diameter far less than the casing inner diameter (ID). This also made the tool move itself towards a more
centralized position after pad contact was first obtained, explaining the early bumps in logged pad force. It
is also manifest that the recorded stiffness is significantly lower than for the centralized case. This may be
explained by the partial pad-to-casing contact. A clear linear trend is seen, indicating that no significant pad
or casing deformation is occurring to improve the pad-casing contact. Similar effects may also be observed
for the result from logging with a tilted tool, however, to a lesser extent.

Figure 10—Results from experimental testing of a well-cemented

casing with tool eccentricity in two different directions and tool tilting.
12 SPE/IADC-194135-MS

Discussion
Experimental testing
The experimental testing has shown that an uncemented casing gives a characteristic and clearly linear
response, represented by the mechanical properties of the casing itself. This makes a free pipe relatively
easy to detect. A cemented casing, however, proved to give a more complex response: It is also linear, but
seems to depend on other factors than solely the physical properties of the casing, cement, and formation.
However, a greater stiffness is easily noticeable in the cases where the casing is supported by an annular
cement sheath, compared to the case of a free pipe. As the free pipe response is easily recognized, it is
possible to detect when it deviates from this mechanical response. This contrast in stiffness makes the AVT
able to identify the presence of a casing-cement microannulus through the change in slope on the logged
response.
The results from the experimental testing indicate that AVT prototype was able to quantify the size of
a uniform microannulus at the casing-cement interface, with an associated and rather small uncertainty
of ±10 μm. To ensure consistency in the microannulus size calculations, a semi-automated procedure was
formulated, using least squares linear regression to fit the straight line. A constant threshold was used to
detect the diameters where the logging response departs from the straight line. However, the source data
used for the linear regression had to be picked manually. For large microannuli, this is straight forward, as
the linear region is predominant. For small microannuli, the free pipe region is rather small, making it more
challenging to select this dataset.
Furthermore, it was confirmed that the AVT is susceptible to poor centralization within the casing, as it
causes a mismatch in the alignment between the tool pads and the casing wall. However, the tool has proven
to be self-centralizing in the direction of pad movement. A minor pad force was needed to move the tool
into a centralized position. This self-centralizing characteristic could possibly be exploited by including
additional pairs of pads perpendicular to each other, but separated some distance in the longitudinal direction
of the casing string. Thus, ovalization of the casing cross-section would not be avoided. In addition, larger
pads might help to improve self-centralization of the tool. These ideas need further evaluation, with the aim
of reducing the AVT's sensitivity to centralization.
Visual inspection of the cement sheath after logging the samples did not reveal any damage due to the
forces imposed on the casing inner wall. However, the possibility of micro-scale damage like cracks or
debonding is not discarded.

Potential of the AVT to complement cement sheath evaluation

The most important advancement provided by the AVT appears to be the ability to quantify the size of a full
microannulus at the casing-cement interface. Today, the presence of this microannulus is often detectable,
but no tools are able to measure its size. Hence, the AVT enables a more comprehensive evaluation of
the microannulus effect on the integrity of the cement sheath, given the significant effect it may have on
potential leakages.
Furthermore, the AVT concept may be particularly beneficial in the case of a gas-filled microannulus,
where the acoustic tools face a challenge to detect the cement sheath, as they offer a response similar to
that of a free pipe.
The AVT concept, in combination with the acoustic logging tools use today, may potentially provide
valuable insights into P&A operations, where:

• The decision-making of using an annular cement sheath as permanent barrier, along a designated
well internal, has a substantial impact on the operation cost, and thus should be taken based upon
the best possible foundation of data.
SPE/IADC-194135-MS 13

• Creeping/swelling of shale formations towards the casing is considered as a potential barrier, since
status of shale barrier and horizontal stresses are uncertain.
• The annular cement sheath of a well interval can be re-establish by a system termed Perforate,
Wash and Cement (PWC), (Ferg et al., 2011). Existing logging tools are difficult to use in this case,
and the AVT may be an alternative for verification of the cement quality.
Concerning the challenges that may be encountered when using the AVT concept, it is noteworthy the
following:

• Improper tool centralization, as it results in contact mismatch between the tool's pads and the casing
inner wall.
• The AVT measurements have inherent limitations to distinguish azimuthal variations, such as
channeled cement, and should not be carried out in such intervals.
• Irregular casing wall and accumulation of solids, as they may affect the effective pads-to-casing
contact area.
• The nature of the AVT demand stationary measurements, and should then be performed in any
intervals where a microannulus is suspected, based on the acoustic logs.
• Cement-formation microannulus, as it may affect the quantification of the casing-cement
microannulus to some extent, particularly in the case of lightweight cements.

Field implementation
For cement evaluation, the AVT may be integrated into the same wireline tool string used for the acoustic
tools. However, while acoustic logging is performed when pulling the string upwards at a constant logging
speed, the AVT demands stationary measurements. Hence, the number of AVT measurement points and
the location should be thoroughly evaluated. This may be done based on real-time interpretation of the
acoustic logs, and a casing collar locator (CCL) to avoid logging casing collars. Following this, two-way
real-time communication with the downhole tools may be needed. This enables real-time interpretation of
the acoustic logging data, as well as the possibility to communicate with the AVT when measurements are
to be taken. It also allows for real-time interpretation of the logging results from the AVT. This does not
seem like an issue, as the acoustic logs have long been presented in real time, using electrical wireline (Guo
et al., 2015; Sheives et al., 1986).
To make the AVT fit for downhole application, several changes are necessary compared to the prototype
constructed for this initial experimental testing. This would comprise integration into a typical wireline
logging string, hydraulic power unit, fail-safe mechanisms to preserve the tool's mechanical integrity and
operational features.

Future work
Experimental testing of the AVT should be continued to further evaluate the uncertainty related to
microannulus size estimations. Testing on more complex casing-cement-formation samples, like channeled
cement, partial microannuli, casing stand-off, should also be performed. In addition, the effect of a cement-
formation microannulus should be evaluated. This work may be partially supported by extensive finite
element numerical simulations. The effect of improper tool centralization on the measured microannulus
size should further be investigated, as well as the use of larger pads. In a longer perspective, the AVT should
be made ready for testing a more representative downhole setting.

Conclusions
A novel mechanical-based approach for cement evaluation – the Annulus Verification Tool (AVT) has been
introduced, as well as the laboratory set-up for testing of a prototype. This comprised the construction of full-
14 SPE/IADC-194135-MS

scale diameter samples representing a typical 9 5/8-in. production casing cement job, with the possibility
to generate a uniform and controllable microannulus at the casing-cement interface.
1. The tool proved to be able of quantifying the size of a uniform casing-cement microannulus with an
uncertainty of ±10 μm. This was achieved by radially loading the casing inner wall with a pair of
pads in the tool, recording their relative displacement and then identifying variations in stiffness as
the casing makes contact with the annular cement sheath.
2. Visual inspection of the annular cement sheath after logging did not reveal any damage imposed by
the radial loading imposed on the casing inner wall.
3. Centralization of the AVT within the casing should be ensure, as improper alignment between the
pads and the casing may impact the stiffness estimates, and thus microannulus quantification.
4. The potential insights of the AVT into the cement sheath barrier may be particularly useful in cases
where a gas-filled microannulus exists, as well as in the context of P&A operations.
5. The AVT may be an alternative for verification of cement quality when using the Perforate, Wash
and cement (PWC) method.
6. The AVT may be used for verification of creeping/swelling shale formations towards the casing and
potentially measurements of minimum and maximum horizontal stresses.

References
Bellabarba M, Bulte-Loyer H, Froelich B, Le Roy-Delage S, Kujik R, Zerouy S, Guillot D, Meroni N, Pastor S, & Zanchi
A. 2008. Ensuring zonal isolation beyond the life of the well. Schlumberger Oil Field Review, 18-31, April.
Catala G., V. de Montmollin, A. Hayman, R. Hutin, R. Gilles, D. Guillot, J. Jutten, U. Qureshi, B. Kelly, B. Piot, T. Simien,
and I. Toma, 1991. Modernizing Well Cementing Design and Evaluation. Oilfield Review 3(2) pp. 55–71.
Ellis D. V. and Singer J. M., 2007. Well Logging for Earth Scientists. Springer, 2 ed.
Ferg T. E., H.-J. Lund, D. Mueller, M. Myhre, A. Larsen, P. Andersen, G. Lende, C. Hudson, C. Prestegaard, and D.
Field., 2011. Novel Approach to More Effective Plug and Abandonment Cementing Techniques. Paper SPE-148640-
RU presented at the SPE Arctic and Extreme Environments Conference & Exhibition, Moscow, October 18-20.
Guo Y., J. Qi, Y. S. Zheng, S. Taoutaou, R. Wang, Y. An, and H. Guo, 2015. Cementing Optimization through an Enhanced
Ultrasonic Imaging Tool. Paper SPE-176069-MS presented at the Asia Pacific Oil & Gas Conference and Exhibition,
Bali, October 20-22.
Hayden R., C. Russell, A. Vereide, P. Babasick, P. Shaposhnikov, and D. May, 2011. Case Studies in Evaluation of Cement
with Wireline Logs in a Deep Water Environment. Paper presented at the SPWLA 52nd Annual Logging Symposium,
Colorado, May 14-18.
ISO 10426-1:2009/API 10A Petroleum and natural gas industries - cement and materials for well cementing. Part 1:
Specification. American Petroleum Institute.
Jutten J. J. and Hayman A. J., 1993. Microannulus Effect on Cementation logs: Experiments and Case Histories. Paper
SPE- 25377-MS presented at the SPE Asia Pacific Oil & Gas Conference & Exhibition, Singapore, February 8-10.
Kuijk R. van, S. Zeroug, B. Froelich, M. Allouche, S. Bose, D. Miller, J.-L. Le Calvez, V. Schoepf, and A. Pagnin, 2005.
A Novel Ultrasonic Cased-Hole Imager for Enhanced Cement Evaluation. Paper IPTC-10546-MS presented at the
International Petroleum Technology Conference, Doha, November 21-23.
Morris C, Sabbagh L, Wydrinski R, Hupp J, van Kuijk R and Froelich B. 2007. Application of Enhanced Ultrasonic
Measurements for Cement and Casing Evaluation. Paper SPE/IADC 105648 presented at the SPE/IADC Drilling
Conference, Amsterdam, February 20–22.
Myrseth V., G. A. Perez-Valdes, S. J. Bakker, K. T. Midthun, and M. Torsæter, 2017. Development of a Norwegian
Open-Source Plug-and-Abandonment Database with Applications. SPE Economics & Management, vol. 9, No. 01,
pp. 27–31.
National Instruments. Measuring Position and Displacement with LVDTs. http://www.ni.com/white-paper/3638/en/
(accessed 7 September 2018)
NORSOK D-010:2013, Well integrity in drilling and well operations. Norwegian Oil Industry Association (OLF) and
Federation of Norwegian Manufacturing Industries (TBL). August 2013, Rev 4.
Oil and Gas UK. 2012, OP071 - Guidelines for the suspension and abandonment of wells, Issue 4, July 2012 and Guidelines
on qualification of materials for the suspension and abandonment of wells, Issue 1, July 2012. [Online], Available for
purchase: http://www.oilandgasuk.co.uk/publications/viewpub.cfm?frmPubID=447
Pilkington P. E., 1988. Pressure needed to reduce microannulus effect on CBL. Oil & Gas Journal 86(22), pp. 68–74.
SPE/IADC-194135-MS 15

Pilkington P. E., 1992. Cement Evaluation - Past, Present, and Future. Journal of Petroleum Technology 44 (02), pp.
132–140.
Sheives T., L. Tello, M. V.E., T. Standley, and T. Blankinship, 1986. A Comparison of New Ultrasonic Cement and Casing
Evaluation Logs with Standard Cement Bond Logs. Paper SPE- 15436-MS presented at the SPE Annual Technical
Conference and Exhibition, New Orleans, October 5-8.
Vignes, B. 2011. Contribution to well integrity and increased focus on well barriers from a life cycle aspect. PhD thesis,
University of Stavanger, Faculty of Science and Technology