CBLS CAN EVALUATE CEMENT INTEGRITY BETWEEN TWO CASING STRINGS

P.E. Pilkington
Conoco Inc.
Houston

Evaluating cement between two strings of casing can be a problem. The signal from the outer string of casing can interfere with amplitude measurement if the inner string is not centralized.

Interference can also occur if a wide amplitude gate is used and when the annular clearance is small.

The interpreter must be aware of the outer string and how it affects the log; otherwise, the log can be misinterpreted.

CEMENT BOND LOGS

The cement bond log (CBL) can identify poorly centered vs. well-centered concentric casing and give a qualitative idea of the bond between properly centered casing strings. The CBL will have problems with small channels and whenever:

The annular gap between the casing strings is thin
Cement transit time drops below 20 m sec
A wide gate is improperly set.

It is always necessary to understand what the CBL tool is measuring and how it is gated.

Some tools will be more apt to have problems because they do not measure peak amplitude with a narrow gate.

A CBL run with both fixed and floating gates, and using a narrow fixed gate, can make the CBL more useful in concentric casing strings.

The CBL has limitations. But the pulse echo tools, which also can be used in dual strings of casing, also have their limitations and will not be discussed in this article.

Good cementing practices, especially centering casing, are required to achieve good vertical isolation in concentric casing strings. A properly run CBL cannot make up for inadequate cementing practices.

INNER-STRING CENTERED

Fig. 1 is a log section from a CBL run in 9 5/8-in. casing above the 133/8-in. shoe. The amplitude curve is gated on the first 95/8-in. casing arrival shown on the Z-axis log in Track 3.

The steady 133/8-in. casing signature indicates that the 9 5/8-in. casing is well centered. The maximum deviation in this well was 48 with a 33 deviation near total depth.

Three centralizers were used on every two joints. The centralizers appear as discontinuities on the Z-axis plot in Track 3 and as small amplitude peaks with a slight decrease in transit time.

The additional metal on the centralizer and stop collar make it more difficult for the cement to attenuate the casing signal. Thus the amplitude increases slightly. Transit time decreases as the amount of stretch occurring is reduced by the increase in amplitude.

Stretch is illustrated in Fig. 2a. The transit time increases as the amplitude curve approaches the detection level, also referred to as threshold, bias, sensitivity, etc.

Most CBL tools use a "floating" gate to measure transit time and are subject to the stretch phenomenon. Fig. 2b shows stretch on the transit-time curve.

The 95/8 -in. pipe signal has been shifted over on Fig. 1 so that the first arrivals for the 9 5/8-in. and 13 3/8-in. signals are lined up. This shift of about 30 m sec, or one third of a chart division, should be about right for the cement transit time. Cement transit time can be roughly estimated as follows:

TTcmt = (ID1 - OD2) Atcmt

WHERE:

TTcmt = Cement transit time, m sec

ID1 = Inside diameter of outer casing, in.

OD2 = Outside diameter of inner casing, in.

Atcm = Interval transit time for cement, 7-11 m sec/in.

In this example, estimated cement transit time would be in the range of 20-31 m sec. This agrees well with the shift shown on Fig. 1.

The estimate ignores Snell's law but is sufficient for very thin cement sheaths. Ignoring Snell's law results in a slightly low estimate of cement transit time.

SAG BETWEEN CENTRALIZERS

Fig. 3 is another CBL section from the same well as Fig. 1 - The 9 5/8-in. casing is indicating sag between two centralizers. The sag could be caused by a dogleg at this depth.

Note how the 13 3/8-in. signal sways in earlier (to the left) between the centralizers.

Also, either a small channel of some cement with different acoustic impedance exists between the depth X092 and X136 ft. Note the increase in amplitude and appearance of a weak 9 5/8-in. casing arrival on the Z-axis plot.

An anomaly appears on the gamma ray at X075 ft and another anomaly on the transit-time curve at X094 ft. These are believed to be caused by power surges.

The logging unit was using rig power during this log run. A separate generator was used on subsequent CBLS, and no more anomalies of this type occurred.

CASING NOT CENTRALIZED

Fig. 4 is a log from 5 1/2-in. casing inside 7-in. casing in a vertical well bore. Note the irregular wavy pattern behind the first cycle of the 5 1/2-in. casing signal. This results from the 5 1/2-in. casing not being centralized. Good cement jobs can be difficult to obtain under these conditions.

INSUFFICIENT CENTRALIZERS

Casing centralizing requirements are even more stringent in directional wells. Fig. 5 is a section of a CBL run in 9 5/8-in. casing inside 13 3/8-in. casing. Well bore deviation is 32 in this section.

Note the repetitive pattern on the Z-axis plot with amplitude peaks in the middle of each joint of 9 -1/8-in. casing.

The chevron patterns on the Z-axis plot that accompany the amplitude peaks are typical of casing-centralizer response. One centralizer per joint is obviously inadequate for this hole angle as the 9 5/8-in. casing is sagging between the centralizers and the interference from the 13 3/8-in. casing signal reduces the amplitude of the 9 5/8-in. casing signal.

FLOATING-GATE AMPLITUDE

The log shown in Fig. 6 was run in uncentralized 7-in. casing cemented inside the 9 5/8-in., 40 deviated well bore. The irregular pattern on the Z-axis plot is typical; however, the amplitude curve reaction is different.

Two amplitude gates, fixed and floating, are logged. Both gates are measuring the 7-in. casing arrivals where the two curves overlay. When the amplitude of the 7-in. casing signal decreases, transit time increased. When amplitude drops below the detection level of the floating amplitude gate, the amplitude curves diverge.

The fixed gate still measures the amplitude of the 7-in. casing arrival while the floating amplitude gate measures the amplitude of the first arrival from the 9 %-in. casing string. The floating gate CBL does appear to have an application under these conditions.

OUTER-STRING INTERFERENCE

Fig. 7 is a section of CBL with the 7 in. not centered inside 9 5/8-in casing. Two sections of scope-picture blowups shown on the CBL illustrate the double peak caused by the outer string of casing interfering with the signal from the inner string.

The upper interval shows the double peak above XX100 ft, but the two peaks merge below XX100 ft where the 7 in. lies against the 9 1/8-in. casing.

The interval above XX150 ft shows distinct separation of the 7 in. and 9 5/8-in. first positive half-cycle, El, peaks which converge below XX150 ft as the 7 in. goes further off center.

CENTERED LINER

Fig. 8a is a CBL section from another directional well with a deviation of 50. Centering used on the 7-in. casing included one centralizer per joint immediately below the 9 5/8-in. shoe and two per joint in the liner overlap section. The drop in amplitude in the overlap section is probably due to interference between the 7-in. El casing arrival and the forerunner from the 9 5/8-in. casing.

The 9 5/8-in. casing arrivals were uniform up to the liner top. Transit time increases at the 9 5/8-in. casing shoe where the 7-in. casing signal drops below the bias level and the floating transit-time gate starts triggering on the 9 5/8-in. casing arrival.

Fig. 8b is a section of the log run over the overlap section with the fixed amplitude gate reset to measure the 9 5/8-in. E1 casing arrival. This is possible only with good annular cement and well-centered concentric casing strings.

IMPORTANCE OF GATING

Gating is important when logging in two strings of casing. Fig. 9 shows two runs of a CBL in concentric casing strings.

The amplitude increase over various intervals is caused by the increase in pressure reducing the fluid transit time and variations in casing centering moving the 9 -5/8-in. casing signal intermittently into the 7-in. pipe gate. This is illustrated by the schematic shown in Fig. 10.

This problem can be readily eliminated by reducing the gate width so that it simply straddles the 7-in. El arrival. Jutten1 suggests a 20 [L sec wide gate set to open on the lead edge of El to minimize the interference from the outer string of casing. This will help when the inner string is properly centered.

Gate width can vary as long as the gate is set to close 20 [L sec after free-pipe transit time. Tools that can't be adjusted to these criteria should not be used in multiple-casing-string cement evaluation.

ACKNOWLEDGMENT

The author wishes to thank Conoco for permission to publish this article and to recognize the contributions of the Conoco engineers who have provided the examples contained in the article. S.F. Stine, who contributed much to my knowledge and understanding of CBLS, deserves special recognition.