Sie sind auf Seite 1von 15

ACOUSTIC CEMENT BOND LOGGING

Warren L. Anderson and Terry Walker


Welex, A Division of Halliburton Company
The opinions expressed in this paper are those of the
authors and are not necessarily the opinions of the Society
of Professional Well Log Analysts or its members.

Publication
Rights Reserved
.

This paper is to be presented at the Second Annual


Meeting of the Society of Professional Well Log Analysts,
May 18-19, 1961, in Dallas, Texas, and is considered the
property of the Society. Permission to publish is hereby
restricted to an abstract of no more than 300 words, with
no illustrations,unless the paper is specifically released
to the press by the Secretary or the Editor of the Society
of Professional Well Log Analysts.

ACOUSTIC CEMENT BOND LOGGING


by
Warren L. Anderson and Terry Walker
Welex, A Division of Halliburton Company
ABSTRACT
Laboratory investigationsand test well oscilloscope
pictures were made to determine the application of acoustic
signals for evaluation of effectiveness of casing cementing.
This work indicated the possibilities of measuring the
amplitude of acoustic signals from the pipe and from the
formation both in o en and cased hole to Indicate the bonding to ipe and to $ormation. Field tests have shown the
applica% ilit of the system, but experience and further
model tests tave shown that a number of variables affect
the interpretation. Some of these variables, such as type
of cement, quantity of admixes, sheath thickness, and time
of 10 ging after cementing, are being investigated and
resulfs to date are included. Results to date indicate that
it will be possible to evaluate the effectiveness of cementing under most conditions.

INTRODUCTION
When reduction.tests do not a ;:: with 10 core analysis,
or ot[er formation evaluation i
the effactiveness of
casing cementing is usually questioned. Even in wells
which have been produced for some time and then begin to
produce water or excess gas, the isolation by cement behind
the casin has been found to be incomplete by a squeeze and
reperfora!ion. The answers to these questions in the past
have been quite expensive and often damaging to formation
productivity during the process of repair.
Recently considerable research effort has been expended by
service and oil companies on the cement bond log, which
offers the possibility of evaluating effectiveness of
cementing. The process involves the measurement of the
amplitude of an acoustic signal, the quality of the cement
bonding along with some other variables, affecting the
signal amplitude. At this point a definition of the term
bond should be established. For the purposes of this
paper bond is defined as coupling between the cement and
pipe or formation such that there 1s no space intervening
and no density discontinuity except that from cement to
steel or formation.
-1-

ACOUSTIC CEMENT BOND LOGGING


EXPERIMENTAL INVESTIGATION
In order to completely evaluate the possibilities of
determining the effectiveness of cement isolation with
acoustic signals a series of test holes were set up with
varied cement conditions, using Ideal Portland 15#/gal.
cement. Oscilloscope pictures were taken of various spaced
receiver signals under a number of conditions. As expected,
it was found that the signal through the pi e depends u on
the cement bond to the pipe. Also, it was !ound that tEe
amplitude of the formation signal depended upon the cement
bonding to the formation. The pipe signal can be considered
as a vibration of the pipe which has an arrival time of 57
microseconds per foot, the bonded pipe having a ver low
amplitude signal because of the damping effect of tie cement
sheath. The amplitude of the formation signal depends upon
the presence of a bonded path for the travel of the acoustic
energ . If the cement is not bonded to the formation the
signsT would have to cross four additional solid-liquld
interfaces with the resulting loss of energy at each. There
would of course, be some loss of ener y because of an
density interface at cement bonded to !ormation; but tK is
loss would be much less than for the unbended condition.
the scope pictures from this study (three feet
:::w~en transmitter and receiver) are presented in Fi ures 1
Figure la shows the high am litude pipe signa!7from
uncem;nted pipe with a very low amp!itude formation signal
superimposed on the ringout of the pipe vibration. Fi ures
lb and lC show the same low amplitude in bonded pipe afthough
the section of pipe in Figure lC was sandblasted to remove
mill scale etc. In both Figures lb and lC the hi h amplitude formation signal indicates the good bond to tEe
formation. Figures ld, le and lf show different widths of
channels and thicknesses o+ cement sheaths. This and other
data indicate that the thickness of the cement sheath
affects the amplitude up to about a two inch sheath (for
Portland cement), but after that point essentially no
change in amplitude can be distinguished. The channels in
these three pictures were rectangular, which is probabl
not too common in the field--the more common situation zeing
decentralized pipe with insufficient cement between the pipe
and formation on one side. This condition will be shown
later.

Some of

Figure 2a is taken under the condition of cement bonded to


the pipe but not to the formation--the pipe signal is,
therefore, low amplitude as is the formation signal because
-2-

of the additional solid-liquid interfaces. Figure 2b is


the reverse condition--greasedpi e to prevent bonding to
the pipe but good bonding to the !ormatlon. In this case
the pipe amplltude is high (but not clean cycles)? indicating the free pipe* and the formation signal still appears
to be present si~ce the formation signal amplitude is about
the same as for good cement bond. This may be the result of
of the pipe to cement through the thin
mechanical couplm
coat of grease. TE is also shows that the damping of the
signal by the cement can be considered as a clamping action
on the pipe. Figure 2C shows the condition of the tool
centered at the cement-water interface. The pipe signal
amplitude is high but again is not a clean si nal, probably
the result of intermittent bonding in the 1 172 feet of
cement affecting the tool. The formation si nal is of
intermediate amplitude which is to be expec!ed since the
si nal had only two soiid-liquid interfaces tc cross instead
of%our. Figures 2d and 2e show decentralized pipe uncemented and cemented. The pipe signal is greatly reduced in
the cemented condition but still of sufficient amplitude to
indicate the channel resulting from no bonding where the
pipe is held against the formation. Again the free pipe
signal in Figure 2d does not show clean cycles, indicating
the slight damping because of the pipe restin against the
formation. The formation arrival appears to Be quite large
in Figure 2d, but it is probably because of pressure coupling
of the pipe to formation. The pressure with which the pipe
is forced against the formation is difficult to evaluate in
this case.
Figure 2f shows the formation signal with no pipe--the tool
was lyin in a trench filled with water, thereby givin the
relativefy lower amplitude. Superimposed is the signaf
received in 5 1/2 inch casing in 4 foot diameter cement.
The cement arrival can be distinguished from pipe arrival
and has not been observed in bond logging under normal conditions to date.

LOGGING TOOL DESIGN


Study of these investi ations indicated that by measuring
the amplitude of both $ormation and ipe arrivals, the
troublesome cementing conditions cou!?
d be resolved. Accordingly, a tool was designed to meet these requirements and
to test the system it was run in open hole and after pipe
was set to check the feasibility of formation amplitude
measurements. A four foot transmitter to receiver spacing
was used in this design. Figures 3a and 3b are a section
of this well which is on a salt dome with lime present,
-3-

giving a wide variation in formation arrival times. As


shown the formation amplitude in open and cased hole is
essentially the same, indicating good bond except for
specific points to be discussed later. This correlative
appearance shows the validit of cased hole formation
amplitude measurements and siows possibilities for bond
logging under all formation veloclty conditions.
Over the entire section shown in Figures 3a and 3b the pipe
amplitude curve is low (compared to the value marked A,
which is the value recorded for free pipe up the hole)
except for readings in the lime zones marked B, C, and D.
In these sections the amplitude of the formation si~nal,
open and cased hole indicates good cement bond. Since
the acoustic veloci~ in lime is of the same order as in
steel, the somewhat Iigher amplitude of the pipe signal is
the result of a formation arrival. In the section marked G
the thick mud cake indicated by the caliper appears to have
prevented bonding to the formation. The low pipe amplitude
indicates bonded pipe, but the cased hole formation signal
is much lower than that in open hole, indicating poor
coupling of the signal, in this case poor bonding of cement
to formation. In the section marked F the cased hole jffmation amplitude is higher than that in the open hole.
caliper over that section shows an erratic character.
Acoustic coupling in the open hole would be poor because of
the large number of travel paths and resulting interference
arrivals at the receiver. When pipe was cemented, the
coupling improved.
The sand marked E shows good pipe bond, but the formation
amplitude is very low in both open and cased hole. This is
an unconsolidated sand that must be gravel packed to be
produced.

VARIABLES AFFECTING ACOUSTIC


CEMENT BOND LOGGING
Since there are so man variables in cementing operations,
it would be expected tiat there would be a number of
variables affecting bond logging. Since the cement appears
to have a dampin action on the amplitude of the acoustic
signal throu h tEe pipe? the effect of cement sheath thickness should Eave a considerable effect. Figure 4 presents
the result of tests on 5 1/2 inch, 17# casing with various
thicknesses of cement sheath. As expected increasing
sheath thicknesses reduce the amplitude. these tests were
made with the pipe suspended in waterl and the bond to the
formation will radically alter this picture. For instance,
-4-

one-inch sheath bonded to a consolidated formation should


attenuate the signal in ap roximately the same manner as a
thick sheath. There have Been cases where the pipe signal
amplitude was very low, but communication still existed
between completion zones, indicating a channel behind the
cement or no formation bond. Under such circumstances the
sheath thickness would be a factor, making knowledge of hole
size necessary before pipe amplitude could be used as an
indicator of cementing qualit . It is possible that the
formation signal amplltude wi$1 resolve such problems or
in some cases, the comparison of open and cased hole $ormation amplitude may be required.
a

uite a number of admixes are being used with cement to gain


esired properties, and these factors will affect the damping
of the pipe signal. Tests are now in progress on these
effects, and the data will be presented as soon as possible.

At present it is difficult to assess the bond requirements-is two or three hundred feet of mediocre bond as good as
five feet of good bond? Figure 5 shows the electric log and
tm::r:: over a sand. The high pipe amplitude in the sand
f indicated incomplete cementing which was verified
by an excessive gas-oil ratio on completion, the gas-oil
contact being in the upper part of the sand. A squeeze and
reperforation resulted in oil production with a normal ratio.
Figure 6 is an upper completion in the same well as the
previous figure. AgainH:::v:+gh pipe amplitude indicates
lncom lete cementing.
in this case the production
for tEe zone was normal with fiowater. The pipe amplitudes
in the section between this oil sand and the water sand
above indicate only a few intermittent intervals of lower
am litude which still are not as low as the lowest amplitudes
taEen from other parts of the log above and below. In this
case these thicknesses of indicated intermediate cementing
were sufficient for isolation. Considerable work and study
will be necessary to learn these requirements for isolation.
Another variable is the apparent losing of cement bond which
has occurred a sufficient number of times to be a problem.
~npexa~:~na~f this is in Figure 7 which shows the caliper,
induction reslstivity, and four runs of the
pip~amplitude log. This well is in a field in Loui~~:na
where primary cement jobs are difficult to obtain.
pipe amplitude run 17 hours after cementing indicates a
good cement job with only occasi~::mv:~gheramplitu~e spots
over the lower interval shown.
40 hours a ter
primar cementing, the amplitude curve indicates a considersz le amount of the apparent bond was lost, especially
over the lower part of the section shown. No operations
were performed on the well between these two runs. The well
was then squeezed and a log run 24 hours after the squeeze.
-5-

This run indicated good bonding for 50-70 either side of


the s ueeze point. Another run 48 hours after the squeeze
at 4190 feet
showe 8 no change except for a small interval
where the amplltude
increased
greatly.
Over the upper
section
shown there was relatively
small change on any of
the four runs.
therefore,
The optimum time for running this type of 10
~me depends
is a roblem requiring study, since setting !!
upon !emperature, ressure, admixes, and contaminatl:n.
Over and above set!in time there is the problem of losing
bond to complicate tEe choice of optimum time. This
problem is now being studied by repeat logging of pipe with
the various additives in the cement, the data to be published as soon as possible.

SUMMARY AND CONCLUSIONS


Acoustic cement bond logging offers the possibility of
determining the effectiveness of cementing under down hole
conditions, which has not been possible previously except
by testing. There are a large number of variables which
affect the interpretation of the logs, but present work
indicates that it will be possible to evaluate effectiveness
of cementing under most conditions.
These variables are being studied and some of the results
are included in the paper and others will be available at
the time of presentation. Further field work comparing
amplitudes of formation signal in open and cased hole,
especially in hard rock country will also be available for
the paper presentation.

-6-

FIGURE I
ACOUSTIC SIGNAL
CONDITIONS.
+

&

100

IN CASED

HOLE

IN VARIOUS

CEMENTING

MICROSECONDS

.-.

--U-J.*V

1
VI

1.2VOLT

T
1

I-A

: FREE PIPE - 5-1/2


9 HOLE.

,.

+00

CASING

I-D : CEMENTED PIPE , L75CEMENT


SHEATH -1.37
CHANNEL, 5-1/2
CASING IN 9 HOLE.

IN

MICROSECONDS

t* -----

b-

100 MICROSECONDS

1.1 v
15 v

-L

II

.6 VOLT

VOLT
1

T
1

I-B

: CEMENTED
IN

7- 7/s

klOo

PIPE -

5-1/2

I-E

CSG.

HOLE .

: CEMENTED
PIPE , 1 CEMENT
SHEATH -.75
CHANNEL,
7
C&lNG
IN 9 HOLE,

MICROSECONDS

.08 w ---

.5 VOLT

I-C : CEMENTED
PIPE (SAND BLASTED)
5-1/2 CASING IN 7-7/B HOLE.

I-F : CEMENTED
PIPE ; 2.75 CEMENT
SHEATH -1.37 CHANNEL,
5-1/2
CASING
IN II HOLE .

FIGURE

2:

ACOUSTIC
SIGNALS IN 5-1/2 PIPE $ET
IN VAR1OUS
CEMENTING
CONDITIONS.

100

IN

7-7/@

HOLE

MICROSECONDS

.03 v

-.2V

L
.5 VOLT

LT

2-A : CEMENT
BONDED

2-D : PIPE
MECHANICALLY
NO CEMENT.

BONDED TO PIPE ; NOT


TO FORMATION ,

100

MICROSECOND
~1.5

, +-

loO

MICROSECONDS

.5VOLT

.5 VOLT

T
1
I
I
I
2-B : CEMENT
BONDED TO FORMATION;
NOT
BONDED TO PIPE.
I

DECENTRALIZED

,,

2-E

I.ev

: PIPE
MECHANICAUY
CEMENTED.

DECENTRALIZED.

H- --FT-Y-PF

35 v

,5 VOLT

2-C : TOOL
CENTERED
CEMENT.

1
AT

TOP

VOLT

-1.ov
C

2-F : FORMATION SIGNAL


CEMENT
SIGNAL

ONLY
ONLY.

AND

FIGURE 3-A:
FIELD EXAMPLE, AMPLITUDE LOGS,OPEN AND CASED HOLE,

ACOUSTIC AMPLITUDE
MILLIVOLTS

GAMMA RAY
o
RADIATION

INTENSITY

-1+

INCREASES

FORMATION

OPEN HOLE

200

HOLE
.
--- 200 .
zoo
.PIPE
..

0
FORMATION
-----

0
..

CASED

r
\

\.

,
\

\
\

..\.
>

.>
\,

------*
.:- ---f
-----

.=-2.
)
.*
&
4 >1 --- o
--a

CALIPER
r

fORXO

CALIPER

17n

FIGURE 3-B :
FIELD EXAMPLE , AMPLITUDE LOGS, OPEN

GAMMA

ACOUSTIC AMPLITUDE
MILLIVOLTS

RAY
O

RAOIATION

INTENSITY

AND CASED HOLE.

INCREASES

FORMATION

0
FORMATION
CASEO
.

o
-..

-144-

200

OPEN HOLE

PIPE
..

HOLE
.

..

20C

20C
.

,?
.-.
)

(.

Sandy
Shale

(
j>
(

?.

? :3
.:.-

I
(
J

;;
---.

\
!4

v
,

,\,

t:-

CALIPER
7

FORXO

CALIPER

17

FIGURE 4:
EFFECT
OF NEAT
CEMENT
SHEATH
THICKNESS
oN
PIPE
AMPLITUOE.

1.5

u)
i 1.0
~
\
~
w
n
g
F!
=
a

.5
L

<
z
~
(n

CEMENT

I
SHEATH

THICKNESS,

INCHES

FIGURE
HIND

5:
LOG

EXAMPLE , LOWER

SP

--120p -

ZONE

OF

DUAL

NORMAL

COMPLETION.

RES.

10

IND.RES.
0 ----------- ----- ----- -- ---1o
:f
:

o PIPE AMPLITUDE 150


Log Recorded 120 Houn
After Cementing.

:
:

7 Pipe
Slo-Set

*
.
.}
,
;
;
:
*
:*
*
s
i*.

Cement

WS-OIL CONTACT

_-

.-*
,0b

###-
:t
~- !
I
*1
t*
?
t:
m
9
B

NITIAL PERE PRODUCED


49 BOPD, GOR 20,650
SQUEEZED, REPERF PR(X1.
?86 BOPD, GOR 1510
)

:
i
;
P

NOTE Free pipe


amplitude225 MV.

MV

FIGURE 6:
BOND LOG EXAMPLE, UPPER

ZONE

OF DUAL

COMPLETION.

NOR. RES
SP

-1201 -

PIPE

AMPLITUDE

5 o~150

o
IND. RES.
o -------5

150

.og Recoded 120 I-+ours


After Cementing.
7 Pipe
S10- Set Cement

I
I
;

3
I

ii
:
I
i
I
I
t
I
I
1

} NOTE:

Free
220

pipe amplitude
MV.
/

FIWRE

REPEAT

RWS

AFTER

PRIMARY

CEKNTING

MD

SWEEZtZ .

PIPE

POTSNTIAL- MILLIWLTS
410k. . .. -- C4LIPER
---------------Avwsss~mhsh81

AMPLITUOZ
Mllllvslm

PIPC

P!PE

AMPLITuOE
Mllllwm

so

1(

AMPLITUOE

Mlllvelm

RINARY CEKNT
JW
W coo , Sso s&2Ns CM
t WIJRS

A?TtR

CWSNTINS

IO HWRS

AfTCR

CEAISNTINS

WURS

AFTCR

SOIXCZ
E

IS IOJRS

>

33+

%==

SOUEI?ZE

INTERVAL

mER

sQuEEZE

Das könnte Ihnen auch gefallen