The Essentials of Cement Evaluation M-081004

Download as pdf or txt
Download as pdf or txt
You are on page 1of 14

THE ESSENTIALS

of cement evaluation

MARCH 1976
CEMENTEVALUATION

TABLE OF CONTENTS

Pages

CHAPTER I Principle of operation of Cament Bond


Log and VariaMe Density Log 1

I-1 Cement Bond Log (CBL) 1

Transit time measurement 3

Amplitude measurement 3

l-2 Variable Density Log (VDL) 7

CHAPTER II Interpretation of CB UVD L 8

I I-1 Qualitative interpretation 8

II-2 Quantitative interpretation 10


CEMENTEVALUATION

I - PRINCIPLEOF OPERATION OF THE CEMENTBOND LO6 AND OF THE


VARIABLE DENSITY LO6 (CBL-VDL)

Among the many factors influencing the


acoustic properties of a cemented casing is the
quality of the bond from casing to cement : the
waves travelling along the casing are attenuated
when energy is lost to the environment from the
casing, i.e. when the bond is good.

The sonic tool used to evaluate the cement


TRANSMlTlfR
quality is schematized in figure I-1 : the transmitter
emits acoustic wave trains of short duration. The
signal travels through the casing, cement and
formation before it reaches two receivers, 3 ft and
5 ft from the transmitter.

The Cement Bond Log (CBL) is a recording of


the amplitude of the first arrival of energy on the
3 ft receiver.

The Variable Density Log (VDL) is optional and


supplements the information given by the CBL ; it
is a full-wave display of the 5 ft receiver signal. Figure l-l : CB L-VDL tool

The three-foot spacing was determined experi-


I - 1 - CEMENTBONDLO6
mentally to allow logging of the full range of
cement conditions. The compressional wave traveling through the
casing is generally the first one to reach the 3ft
The five-foot spacing allows a time separation receiver ; the first three peaks of the signal at the
of the fast and slow arrivals before their receiver are labeled El, E2, E3 as shown in
presentation. Fig. l-2.

-l-
When the bond between casing and cement is
good, the amplitudes of Et, Eq, E3 decrease
(Fig. i-2).

9. CIRC”WLRLNCE BONDED
1
Figure l-4 : Per Cent attenuation vs Per Cent
circumference bonded.
i
lgure l-2 : Schematic receiver output signal with un- On a Cement Bond Log, two quantities are
bonded casing and with bonded casing. measured :

The amplitudes of ET, Eq, E3 are proportional - the transit time At : it is the time taken for
to the attenuation rate, for a given level of El to reach the receiver (see Fig. l-2)
transmitter signal. The attenuation rate, when
there is a good bond from casing to cement, - the amplitude of Et.
depends on the compressive strength of the cement
and on the percentage of circumference bonded. Fig. l-5 shows a typical Cement Bond Log,
(see Fig. 1-3 and 14. This has been shown by with the scales usually used for At and for the
laboratory experiments. amplitude.

1: 5 l/Z’- 17tc casing

P 4’
It
a’ 0.~. cement

.--J NEAT W/40% WATER 85-E a *cm psi


CI aENTON,TE 8% 85-F 0 800 DII
.-* POZMfX A 50 - 50 85-F 0 800 PSI
10
2 4 6 ~db/ff.‘0 I2 ‘4
gure l-3 : Compressive strength v-s attenuati
Figure l-5 : Typical Cement Bond Log
rates for cements shown.

-2-
Transit time measurement Cycle skipping

At the moment the sound pulse is transmitted In the case when the bond leads to an amplitude
from the tool, an electronic clock starts of ET so low that it is below the detection level,
counting. the transit time measurement would be triggered
by E3, instead of ET ; this is refered to as
The bias-setting is adjusted so that when ET cycle-skipping (see Fig. l-8).
reaches the receiver, it is detected and the clock is
stopped. At is the elapsed time recorded by the
clock. (see Fig. l-6).

-. .- - . .
I-rgure l-6 : I ransrt time measurement

The detection level must be high enough so that


the background noise does not trigger the system. Figure l-8 : Cycle skipping
This method of measurement exhibits two
distinctive characteristics when the amplitude of I% a summary, a good cement bond appears on
ET decreases due to a good bond from casing to the transit time measurement as an increase in At,
cement: stretching and cycle skipping. either a small one (stretch) or a larger one (cycle
skipping). Stretch and cycle skipping can be seen
Stretch on the example logs in Fig. I l-2 and I l-3.
The stretch effect is illustrated in Fig. l-7. A
Figure I-11 a is a typical C8L log showing free
decrease in the amplitude of Et appears as a stretch
pipe and good bond.
of the transit time, since the detection level is

Amplitude measurement

To measure El, an electronic “gate” is cpened


for a short time and the maximum value of the
signal received within the gate is recorded.

Two gating systems can be used in the CBL


tools :

- the floating gate : the gate is triggered by the


transit time mmsurement ; the amplitude
Figure l-7 : Stretch mmsured is that of the first peak above the
detection level of At.

-3-
- the fixed gate : the time the gate opens is fixed If the transit time measurement is triggered
by the operator ; the amplitude recorded is the by ET, both systems give the same value.
maximum of whatever signal is present at the
receiver output during the opening of the gate ;
the operator sets the gate so that the amplitude If there is cycle skipping, the two gating systems
of ET is recorded. give different results :

The two gating systems are schematized in


Fig. 1-9. - the fixed gate reads ET, which is very small

- the floating gate reads E3, instead of ET.

In Fig. I-10, the same log recorded with the two


I rz::::“.
rat gating systems is shown.

A :::‘“.““zL. Table I-1 summarizes the different possibilities


Figure l-9 : Gating systems and their influences on the transit time and on the
amplitude.

TRANSIT TIME TRANSIT TIME

FIXED GATE FLOATING GATE


Figure I-10 : Same log with fixed gate and floating gate.

-4-
Table l-l compares the two systems ; with the
fixed gate, a good bond always corresponds to a
low amplitude reading. The floating gate must be
interpreted with greater care, but it does, however,
eliminate gate-setting errors due to changes in the
sound velocity in mud, etc.

In any case, the gating system usad should


be known when interpreting a CBL.

In figure I-11 is a typical CBL log, run with


Table I-1 : Comparison between gating systems fixed gates showing free pipe and good bond.
Stretch occurs at one place (see Fig. l-11).

Figure I-11 : CBL log showing free pipe down to 1350 ft and good bond below.
Stretch occurs at 1375 ft and 1398 ft.

-5-
CBL in deviated holes : effect of eccentering

The centering of the tool is very critical for the


response of the CBL. The effect of eccentering is
I TRANSIT TIME
detailed in figure l-12 : two different paths for the
acoustic waves are shown, one shorter than the
other. This path difference spreads out the signal at
the receiver, as shown in Fig. I-13.

LONG PATH SHORT PATH

Figure l-12 : Eccentering of the CBL tool.

I LFFECTon at
I
lure l-13 : Effect of eccentering on the acoustic
. . .
slanal a* me recewer.

The effect of eccentering on the CBL log is


twofold :

- the transit time At decreases,

- the amplitude of El decreases(l/2” eccentering


reduces Et by more than 50%)

The simultaneity of these two phenomena on a Figure l-4 : Effect of on a CBL log.
CBL log indicates eccentering (see Fig. i-14).

-6-
I - 2 - VARIABLE DENSITY LO6 [VOL) contrast depending on the amplitude of the
positive peaks.

The different parts of the wave train can be


identified on the VDL log : the casing arrivals show
The principle of the Variable Density Log is as regular stripes, while the formation arrivals are
explained in Fig. l-15 : the full wave-train is characterised by their irregularity. This will be
displayed on the film as light and dark stripes, the illustrated in the example of Section II.

Figure l-14 : Principle of operation of the Variable Density Log.

-7-
II - INTERPRETATIONOF CM-VDL

QUALITATIVE INTERPRETATION Good casing bond and good formation acoustical


&piing
As mentioned in the introduction, the VDL is
not always recorded with the CBL. However, it is a Acoustic energy is then transmitted to the
very useful complement to the information yielded formation. This results in weak casing signals
by the CBL. Therefore, we shall compare the coupled with strong formation signals, depending
responses of the two recordings in different on formation characteristics.
situations.
CBL : low amplitude (Et) ; when amplitude is
In Fig. 11-1, II-2 and II-3 are reproduced three very low, stretching and even skipping
CBL - VDL logs run in the same well at different may occur in the transit time curve.
times all using the fixed gate system.
VDL: weak casing signal ; strong formation
Fig. I I-1 shows the logs that were obtained four arrivals if formation attenuation is not too
days after the initial cementing of the 7” casing in high.
the 8 l/2” hole, with class G cement.
Section B in Fig. II-2 and II-3 (from 7800
The CBL - VDL in Fig. II-2 was run after a to 7888 ft) is an example of good bond,
squeeze job, and in Fig. 11.3, the logs were made with stretching and cycle skipping in places.
with the casing under pressure.
Remarks : very fast formation may result in a
Such an example illustrates clearly the different formation signal reaching the receiver before the
conditions that the CBL - VDL can encounter. casing signal does. Then At decreases and the
amplitude increases.
Uncemented pipe
Good casing bond but poor formation acoustical
Most of -the acoustic energy travels through the coupling
casing to the receiver, with very little coupling to
the formation The cement attenuates the acoustic energy, but
the energy transmitted into and received from the
C8L : At constant, High amplitude IElI formation is very low.

VDL : Casing signals only, showing as regular. CBL : low amplitude


well contrasted stripes.
VDL: no formation arrivals.

Channelling and microannulus


Note that the casing collars introduce
inhomogeneities in the path of the sonic wave. - Microannulus : a small gap is formed between
They appear in the CBL (increase in At, decrease in casing and cement in a well cemented casing.
Amplitude) and in the VDL (“chevron pattern”).
The section A in Fig. 11-1, II-3 illustrates the - Channelling : cement is present, but does not
CBL - VDL response to free pipe. surround the casing completely.

-8-
In the case of a microannulus, a hydraulic seal Section C in Fig. Ii-2 (from 8000 ft to 8100 ft)
probably exists, but for channelling a seal probably indicates channelling or a microannulus.
does not exist. However, similar waveform and log
results are observed for channelling and
microannulus : If a microannulus occurs, pressuring the casing
improves the bond; a comparison of Sections 5
- C8L : Moderate Amplitude (El), At constant and C, Fig. II-2 and Fig. 11-3, shows that a micro-
- VDL: Moderate casing arrivals and moderate annulus is present around 7815 ft, 8050 ft and
formation arrivals. 8100 ft.

after cement after squeeze with pressurized casing.

-9-
QUANTITATIVE INTERPRETATION The attenuation corresponding to a given
amplitude can be determined with the aid of
Determination of the compressive strength of the Fig. 114.
cement This bond index is, in practice, equal to the
proportion of the circumference of the casing
The attenuation of a sound wave is a function which is bonded.
of the cement bonding, and consequently is related
to the compressive strength of the cement. In the A bond index of 1. mdicates complete bonding.
case where the cement completely surrounds the
casing, the amplitude of Et can be used to estimate Incomplete bonding is indicated by a bond
this compressive strength with the aid of index less than 1.
Fig. 114.
The minimum value of Bond Index necessary for
good hydraulic seal varies depending on local
conditions. In practice, a figure of 0.8 generally
indicates good results. However, the Bond Index,
by itself, is not sufficient to guarantee zone
isolation. The length of the bonded interval must
also be considered. Field experience indicates that
the minimum bonded interval needed for effective
zone isolation depends on the casing size.

Fig. II-5 was derived from observations and


isolation tests in field wells. The figure shows
casing size versus the interval of Bond Index of 0.8
required in order to insure zone isolation. The
larger the casing, the more bonded interval is
required.

c
FOR BOND INDEX=D.S

I
17”
gure II4 : CBL interpretation chaq - c
^_
toot only ; itI-toot spacmg.
/

i5%
/
An indication of the quantity of the
cementation (and whether isolation can be
expected) is given by the value of the Bond Index.

The Bond Index is defined as :


i”
CASI NG SIZE
-
I lure I l-5 : Length of ten nented interval requiredfc
zone isolation (for Bond Index = 0.8).

-lO-
1. Grosmangin, M., Kokesh, F.P., and Majani, - A Study of Cement and Casing Variables,”
P. : “A Sonic Method for Analyzing the J. Pet. Tech. (May 1963) 545-555; Trans.,
Quality of Cementation of Borehole Casings,” AIME, 228.
J. Pet. Tech. (Feb. 1961) 165171 ; Trans.,
AIME, 222. 3. Brown, H. D., Grijalva, V. E., and Aaymer,
L. L. : “New Developments in Sonic Wave
2. Pardue, G. H., Morris, R. L., Gollwitzer, Train Display and Analysis in Cased Holes,”
L. H., and Moran, J. H. : “Cement Bond Log The Log Analyst (Jan. - Feb. 1971) 27-40.

-ll-
NOTES

You might also like