SPE 133286

Remedial Cementing in Limestone Formation Using Bradenhead Squeeze:

A Case History of KS-X22
E.C. Aprilianto, SPE, PT Medco E&P Indonesia

Copyright 2010, Society of Petroleum Engineers

This paper was prepared for presentation at the 2010 SPE Russian Oil & Gas Technical Conference and Exhibition held in Moscow, Russia, 26–28 October 2010.

This paper was selected for presentation by an SPE 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 and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, 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 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 copyright.

Abstract

The high porosity of the limestone formation of Kaji-Semoga Field, Indonesia, caused a vast amount of lost circulation during
the waterflood drilling campaign. An example of major lost circulation was at KS-X22, which caused a major failure in the
primary cementing. This failure resulted in a significant increase in water cut, from 45 % to 96 %, within 6 months.

To make KS-X22 economically viable again, it was decided to improve the primary cement quality by remedial cementing. The
job was designed as low-pressure balanced plug placement, with 15.8 ppg G class cement pumped down on open-ended
tubing. The limited capacity of the onsite batch mixer restricted the cementing process, which could not be done in one stage.
But, to our advantage, this limitation enabled us to control the amount of cement required for the following stage, based on the
results of the previous stage.

When the job was finished, the cement bond log showed that it had been done perfectly, indicated by the significant change in
the free pipe reading (62 mV) to an excellent cement bond (<20mV). The producing phase after the job showed a slower rate
in the water cut increase, and currently the well is still producing using ESP, more than 2 years after the remedial cementing
job.

This paper will discuss the planning, implementation, and results of remedial cementing at KS-X22 and the lessons learned

Introduction

Kaji-Semoga Field is located onshore on South Sumatra, Indonesia. This field mainly produces from limestone, called
‘Baturaja Formation’, and two other sandstone formations. The average porosity of Baturaja limestone is 16 %, with 65 mD
permeability and an average reservoir depth of 2,700 ft TVD SS.

Baturaja Formation has been developed aggressively since its discovery in 1996, and in 2006 a waterflood project was begun
to sweep the remaining oil. This project was implemented through two major activities:

1. Water injection pattern re-alignment

2. Up to 67 infill drilling in 2006, and then continued by 37 more infill drilling in 2007.

Unfortunately, in the early stage of its development, the pressure maintenance of the Baturaja formation was not well
established. Therefore the reservoir pressure decreased to lower than bubble point. In addition, a combination of its porous
formation and decreased reservoir pressure created several lost circulation problems in many parts of the field, which has
created challenges during primary or squeeze cementing.

As mentioned above, one major lost circulation occurred during infill drilling in KS-X22, located in the crest of Baturaja
reservoir, which had an average pressure of 800 psi.
2 SPE 133286

Drilling History

KS-X22 was designed as a vertical well. In the early stages, drilling activity was executed normally. But after setting 9-5/8”
casing at 2,665 ft KBMD and beginning to drill using 8-1/2” rock bit through top of Baturaja at 2,765 ft KBMD, lost
circulation occurred all the way to TD at 3,033 ft KBMD.

Several measures were taken, but in the end this well had taken up to 5,792 barrels of mud and LCM. Considering this, it was
decided not to run any open-hole logging in order to avoid downhole problems. We continued by running 7” production
casing. Its completion would be based on data from offset well and cased-hole GR-CNL log interpretation.

Primary cementing in this well was also very difficult. 30 bbls of LCM was pumped before 56 barrels of cement, consisting of
130 sacks of cement weighing 11.5 ppg. However, lost circulation still occurred. This could be seen from annulus observation
and was a sign of bad primary cementing. The drilling rig was then moved out to another well objective, leaving the well to be
completed with a service rig.

Completion History

Four months after the drilling rig was moved out, a service rig was moved in to perform completion. GR-CCL-CBL-VDL log
was run and it was found that interval 2,800 ft KBMD to 2,880 ft KBMD showed an average value of 60 mV value, which was
summarized as pipe without cement behind casing (Figure 1 – CBL-VDL Before), compared to free pipe reading at 62 mV
(Figure 2 – Free Pipe Reading). The VDL reading was also confirmed by showing a faint chevron effect related to the casing
collar. Nevertheless, KS-X22 was still completed by perforating interval 2,820-2,830 ft KBMD, for the following reasons

1. Interval 2,820-2,830 ft KBMD showed a promising readout based on its adjacent well
2. Looking for its initial testing in current condition, and
3. If remedial cementing was required, the perforation would be used for cement path.

Using gas lift as artificial lift, the initial testing gave 134 BOPD with 45 % water cut. Because of this result it was decided to
produce this well as it was.

Production History

By November 2006 the water cut at KS-X22 had rapidly increased to 96% and the well only produced 14.2 BOPD. Re-
evaluating the well condition, the main suspect was the water channeling, because of bad cement bonding. It was then decided
to do a workover to shift up the current perforation after performing remedial cementing, based on the following
considerations:

1. Oil water contact at KS-X22 was still 60 ft TVD SS below current opened interval
2. KS-X22 was located up-dip than offset wells
3. It was considered that remedial cementing could prevent water channeling behind casing and prolong the life of this
well.

Remedial Cementing Planning

Key parameters for planning are outlined below:

1. Pressure Calcuation
With formation depth of 2,800 ft KBMD and assuming 0.7 psi/ft fracture gradient, formation fracture pressure will be
around 1,960 psi. Subtracting hydrostatic pressure of 1,200 psi, it will obtain an approximate maximum surface
pressure of 760-psi, which is far below the well head rating of 3,000 psi.

From the calculations above combined with the drilling history and field experience, it was decided to use packerless
low pressure squeeze cementing, often called Bradenhead cementing, since the applied pressure will be held by the
well head, not by the packer (Smith, 1976)

2. Cement Placement
We did not consider block squeeze for remedial cementing in this well because:
SPE 133286 3

a. Block squeeze requires additional perforation for cement path. Adding perforations results in extra cost
(Worzel, 1990)

b. Our objective was to fill up the voidage behind casing while applying pressure to filtrate water into formation.
Therefore, we did not require additional perforations, since the existing ones would be enough to deliver
cement.

c. LCM was not considered at the initial stage, since there was a possibility that LCM would create void and
channel, which cannot be filled in with cement.

3. Cement Composition

Considering the objective was to seal the free pipe zone by filling voidage behind casing, special cement was not
required. Another reason for not using special cement was that we did not know how much cement would be required
on site to perform remedial cement. An insufficient amount would delay the remedial cementing progress, which
would create an extra cost for the rig. Too much stock would also result in create extra cost.

From the logistics perspective, using regular G-class cement was more convenient since we were conducting a
drilling campaign at the time. Plenty of stock was available at the field and could be brought to the well site at short
notice

Cement slurry was based on regular squeeze cement composition, which was G-class cement with additives, a
combination of fluid loss agent, retarder, defoamer, and friction loss reducer. Properties of cement slurry were:

- Fluid loss : 88 ml/ 30 sec

- Thickening time ; 4.09 hrs
- Cement weight : 15.8 ppg
- Slurry yield : 1.16 cu ft/sack
- Compressive strength : 2,951 psi, after 24 hours

The exact cement amount could not be determined precisely, but from another squeeze cementing experience in this
field, it is considered that usually 25 – 80 sacks of cement will do the job. We therefore would deliver cement
gradually, and then evaluate the well response before mixing and pumping the next cement batch if required. This
method would compromise the limitations of the available batch mixer capacity in the cementing unit.

An advantage of using batching system is to minimize the drill out cement process, which in the end will optimize
material usage and save rig time. Another difficulty was that if we pumped too much cement and created a very high
top of cement, it was possibile that the cement would reach its thickening time before the tubing was raised up above
the top of cement, which could have resulted in the loss of the well.

4. Cementing Equipment

Since applied pressure would be below the well head rating of 3,000 psi, a special cement packer for pressure
isolation was not required. A cost reduction was obtained by eliminating the need for a special packer.

At the outset, the limited capacity of the batch mixer became an issue, but eventually this issue was resolved, since
cement placement method did not require a huge capacity of batch mixer. Summarizing the above criteria, the
existing equipment, which was 400 HP cementing unit with 80 sacks cement mixing capability, would do the job

2-7/8” size, EUE, J-55 grade tubing was used for working string, since it was common tubing at Kaji-Semoga field.
No additional tubing handling equipment was required for this job.

Cementing Program Design

Based on the planning above, remedial cementing for KS-X22 was designed as follows (as shown at Figure 3 – Squeeze
Cementing Process)
4 SPE 133286

1. Run open ended tubing to 2,850 ft KBMD (or 20 ft below lowest perforation depth). Using an extra 10-20 ft was
common practice to ensure tubing placement was below perforation.
2. Conduct injectivity test using salt water to measure pump rate and pressure
3. Mix and pump desired amount of cement based on injectivity test result into balanced condition inside tubing and
annulus between tubing & casing. For this case, 80 sacks of cement will be used for the first stage.
4. Raise up open-ended tubing above predicted Top of Cement
5. Observe and apply pressure for displacement carefully using 8.4-ppg salt water.
6. For field practice, if we cannot reach final pressure at certain barrels of displaced slurry, clean up well and assume all
slurry has gone into formation and no cement has been left inside casing and tubing.
7. Repeat above steps until final pressure obtained.
8. If final pressure is finally obtained, shut-in well for 12 hrs curing time.
9. Drill out cement then perform positive test by applying pressure into well, and perform negative test by swabbing
well into under-balance condition.
10. When positive and negative tests have been passed, job is done; otherwise we should re-do the complete process

Field implementation

Since the schedule was tight, the service rig could not be moved to KS-X22 until June 2007. Field implementation for
remedial cementing is described below and summarized in Table 1 – Remedial Cementing Summary, while cementing charts
for this process are shown in figures 4 to 8 respectively.

a. Stage-1

Injectivity test gave us 5 BPM rate at 200 psi. As planned, 80 sacks of cement were mixed and pumped down using
gravity, since surface pressure is not necessarily a prerequisite of a successful squeeze job (Worzel, 1993). Final
pressure was not obtained and well still taking fluid 0.7 BPM. Rock bit then run to check if there was remaining
cement in formation, but found nothing, which indicates that all cement went into formation.

b. Stage-2

Injectivity test after 1st stage resulted in 4.5 BPM at 800 psi. Because it still indicated heavy loss, 80 sacks of cement
were pumped down using gravity. Final pressure was still not obtained. Open-ended string was then lowered down to
prepare for 3rd stage.

c. Stage-3

Injectivity test for 3rd stage resulted in 3.1 BPM at 800-psi. Compared with previous stages, injectivity result was
gradually decreased. But because it still considered high loss, 80 sacks cement was also used. After delivering 12 bbls
of 16.5 bbls, we decided to close tubing valve to hold cement slurry into its position. Rock bit then run and TOC was
obtained at 2,837 ft KBMD, but it was still below lowest perforation depth at 2,830 ft KBMD. Drilling out cement
was performed, followed by an injectivity test, which resulted in 1.7 BPM at 800 psi.

d. Stage-4

Since 3rd stage successfully decreased the injectivity result, for this stage 10 bbls lost circulation material (LCM) was
used ahead followed by 25 sacks cement. Final pressure was obtained for this stage and it was decided to hold
pressure in 800 psi until thickening time was reached. Unfortunately the pressure decreased to zero during waiting
time period, indicating that this stage failed. Our suspicion is that this failure was due to too much LCM usage which
blocked perforation entry hole and made it difficult for the cement slurry to enter the perforation hole. Even though
TOC was obtained at 2,812 ft KBMD, this cementing stage failed due to leaking at positive test. The injectivity rate
after this stage resulted in 2-BPM at 800-psi.

e. Stage-5

Analyzing and learning from the previous stage, the 5th stage was executed using 80 sacks of cement. This amount
was used to compensate elimination of LCM. Hesitate squeeze was done for after 4 hours until the cement almost
reached its thickening time. Unfortunately, final pressure could not be obtained. Tubing valve was then closed and
pressure observation was performed. Unlike the previous step, pressure was holding. Top of cement at this stage was
SPE 133286 5

found at 2,741 ft KBMD, and after cement was drilled out, finally the result of this stage was able to withstand
positive and negative tests.

Remedial Cementing Result

The CBL-VDL result after the final stage is shown in Figure 9 - CBL-VDL After. Compared to the initial CBL-VDL log, an
improvement in the cement bond reading was obtained. To evaluate whether the result is acceptable, CBL-VDL was analyzed
as below:

1. Interval 2,770 – 2,790 ft KBMD showed slightly longer transit time, which should correspond to good bonding. The
Bond Index (BI) calculation from this interval gives us Bond Index of 70%. The VDL result at track 3 also confirmed
low amplitude. Figure 10 – Free Pipe after Remedial Cementing was used as transit time baseline.

2. Interval 2,790 – 2,816 ft KBMD showed shorter transit time. Nelson (1990) said that 4 ms decrease was acceptable,
but this average 4 ms reading should be correlated to improved readability of log display. Combined with good
attenuation at track 2 and low amplitude indication at track 3, it can be said that this zone is also well cemented and
probably encountered fast formation. In fast formation it is not possible to quantify the result with Bond Index,
therefore only qualitative result is available for this interval.

3. Interval 2,816 – 2,824 ft KBMD showed skips, with a transit time of 15 – 20 ms longer than baseline. This indicates
good bond exists between cement and casing. Attenuation and amplitude at track 2 and 3 also confirms this reading.
Bond Index rule was not applicable in this interval due to energy reflections at the cement external interface.

4. Interval 2,824 – 2,890 showed varying transit time, fast formation was the most possible influence in this interval.
Therefore it can be said qualitatively that the cement quality is quite good.

Tracs.com (2007) explained that BI of 0.8 is generally considered enough to give a good hydraulic seal. A minimum length of
10 ft should be required for BI of 80% at 7” casing, shown at Figure 11 – Minimum Cemented Interval. This value was clearly
achieved at interval 2,816-2,824 ft KBMD,

For other segments, Bond Index was less than perfect or could not be measured. But, according to Bassiouni (2006),
reasonable assurance of isolation may be achieved with a long enough interval less than perfect bond. Considering the length
of improved cement bonding and the log analysis, it can be concluded that the job was completed successfully.

Re-perforation was then conducted for checking the possibility of producing this well again at the same interval after remedial
cementing. The result was 100% WC, which indicated that water breakthrough had occurred or oil-water contact was higher
than our initial suspicion. Considering the re-perforation result, it was decided to shift up from current perforation interval.
Squeeze cementing for this perforation interval was performed perfectly in one stage using 60 sacks of cement. A new interval
in Baturaja at 2,782 – 2,790 ft KBMD was then perforated, giving a good initial test result of 346 BOPD, 15 % WC using gas
lift. The result of the previous remedial cementing job also ensured the sealing reliability of this well. In August 2009 this well
was converted into ESP and is still producing today. After years, the water cut increment was not as rapid as before, as shown
in Figure 12 – Water Cut Monitoring.

Conclusions

1. Remedial cementing in KS-X22 was successful, even though only simple techniques and resources were used
2. Careful placement & calculation of cement slurry will deliver a satisfactory result
3. Using cement with higher fluid loss formula is recommended to build faster filter cake.
4. Remedial cementing time for this process could be improved by:
a. Eliminating RIH rock bit to check TOC at first step
b. Re-evaluating the amount of LCM being used, since too many LCM tend to bridge perforation entry hole
and block cement slurry entering perforation

Acknowledgements

The author would like to thank the Management of PT Medco E&P Indonesia for their permission to publish this paper.
Acknowledgement is also extended to colleagues, advisors, and field crew who provided support during the planning,
execution, and evaluation phase of this job.
6 SPE 133286

References

Bassiouni, Z. Production and Cased Hole Logging Training Handout. 2006. Jakarta: LDI Training

Logging Through Casing. Tracs International Training, www.tracs.com/Graphics/Products/. Downloaded May 14, 2010

Nelson, E.B. ed. 1990. Well Cementing. Sugarland, Texas: Schlumberger

Smith, D.K. 1976. Cementing. Monograph Series, SPE, Richardson, Texas

Worzel, H.C. Subsurface Engineering School - Operational Squeeze Cementing. 1990. Houston: Exxon Production Department
Training

Table

Table 1 – Remedial Cementing Summary

REMEDIAL CEMENTING
Injectivity Amount of cement Cement Volume Displaced slurry
Stage TOC Remarks
psi bpm (sacks) (bbls) (bbls)
1 200 5.0 80 16.52 All none No test
2 800 4.5 80 16.52 All none No test
3 800 3.1 80 16.52 16.01 2,837 TOC below perforation
4 800 0.0 25 5.16 3.67 2,812 10-bbls LCM ahead, Leak on positive test
5 800 2.5 80 16.52 12.24 2,741 Positive and negative test OK

Figure

Figure 1 – CBL-VDL Before

SPE 133286 7

Figure 2 – Free Pipe Reading for CBL-VDL before

1 2 3 4

1 : RIH open ended tubing

2 : Spot balanced plug cement
3 : Raise up tubing, reverse circulate, hesitate squeeze
4 : Drill Out Cement, positive & negative test

Figure 3 – Squeeze Cementing Process

8 SPE 133286

Figure 4 –Stage 1

Figure 5 –Stage 2 Figure 6 – Stage 3

Figure 7 – Stage 4
SPE 133286 9

Figure 8 – Stage 5

Figure 9 – CBL-VDL After

10 SPE 133286

Figure 10– Free Pipe after Remedial Cementing

Figure 11 – Minimum Cemented Interval

Figure 12 – Water Cut Monitoring