SPE-176422-MS

Successful Drilling of First CBM Horizontal Wells in Indonesia Through

Meticulous Drilling Design and Geomechanical Analysis
Anzar Syed, Schlumberger; Jason Patterson, Erick Supriyanto, and Ding Hsu, Vico Indonesia;
Ade Surya Setiawan, and Somesh Bahuguna, Schlumberger

Copyright 2015, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE/IATMI Asia Pacific Oil & Gas Conference and Exhibition held in Nusa Dua, Bali, Indonesia, 20 –22 October 2015.

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
Drilling horizontal wells in soft formations can pose significant risks that require careful planning and
execution. During 2014, Virginia Indonesia Co. Limited (Vico) successfully completed two surface-to-
inseam horizontal wells to appraise coalbed methane (CBM) production potential on their Sanga Sanga
license in East Kalimantan, Indonesia. Both wells passed through soft shales and coals that exhibited
breakouts in offset wells, in addition to heavily depleted zones exhibiting large fluid losses in an area well
known for shallow gas blowouts. Geomechanical analysis of information from nearby offset wells was
used to determine the mud-weight window and fluid properties required to effectively manage these
conditions. The results were incorporated into the preliminary well design and planning, and with model
matching carried out in real time while drilling operations were underway, any deviations in wellbore
integrity were managed as they arose.
The drilling success was, in large part, a result of extensive preplanning and a regimented approach to
how the wells would be drilled and any instability dealt with promptly. In addition, real-time monitoring
that identified deviations from model-based predictions provided invaluable information for planning
subsequent wells. The methodical approach to planning and execution of these wells led to their successful
completion, with the results potentially reshaping the Indonesian CBM and energy industry.
Introduction
Indonesia has been blessed with large CBM resources, currently estimated at 534 TCF (Thomas, 2013)
which is sixth in the world ranking. Over the last 5 years considerable development has taken place in
exploiting the coals for gas production—with mixed results. The three main CBM basins in Indonesia are
the South Sumatra basin in Sumatra, and the Kutai and Barito basins in Kalimantan. Vico is active in both
CBM exploration and conventional development within the Sanga Sanga production sharing contracts
located in the Kutai basin area.
Common well completion techniques in CBM plays are loosely based on those already employed in
conventional wells, although they are typically carried out at far shallower depths, requiring lower tolerances
on equipment and more simplified techniques. In 2014 Vico drilled two horizontal wells in one of the coal
seams located in the Mutiara field to test the production potential through inseam horizontal drilling for
2 SPE-176422-MS

comparison with the previous completion techniques. The high tectonic stresses in the Sanga Sanga basin can
potentially lead to drilling issues from coal failures while drilling the horizontal sections. In addition the
overburden requires a proper understanding of rock strengths and stresses for mud design to avoid breakout in
soft shales.
The area has been under production from shallow sands for some time resulting in the lowering of the
fracture gradient, which introduces additional risks from mud losses and associated well-control, and forma-
tion-instability issues. Therefore, to properly assess these risks and incorporate solutions into the drilling design,
Vico decided to conduct a detailed geomechanics study. This paper describes the key steps of the study, the
recommendations on mud weight design, and actual outcome of the drilling, which provides valuable
comparison of geomechanical predictions vs actual drilling experiences and observations.
To conduct the study, logs, drilling, and core data of some of the selected offset wells were studied in detail
to construct the geomechanical model. Fig. 1 shows the location of some of the offset wells and the tectonic
stress direction which can be compared to the regional stress direction from the world stress map (GFZ, 2009).

Figure 1—Location of offset wells and maximum horizontal directions from study of wellbore breakouts and sonic shear anisotropy.
Inset shows the regional stress direction from World Stress Map (GFZ, 2009)

Geomechanical Model Construction

Construction of the geomechanical model consisted of several steps. The first step was to study the
available data in the offset wells, reviewing the data quality and its applicability for the geomechanical
study. The available data in this case consisted of detailed log data including borehole images, sonic
compression and shear, sonic shear anisotropy, core triaxial tests in coals and shales, and mini-frac data
from earlier hydraulic fracturing operations. The overburden stratigraphy consists of sand/shale and coal
sequences. To construct a correct geomechanical model, it is imperative to properly study the mechanical
properties of each facie separately. Fig. 2 shows typical stratigraphy and log responses of the area.
SPE-176422-MS 3

Figure 2—Typical log response in overburden and target coal layer. Shallow sands, some of which are depleted, together with soft
shales and laminated coal layers pose drilling risks when building well angle in overburden
4 SPE-176422-MS

Fig. 3 shows results of the drilling event review of one of the offset wells. In general, some tight spots
and minor mud losses were usual in coals, even when drilling vertical wells. Several offset wells also
exhibited shale breakouts on borehole images.

Figure 3—Typical drilling events observed in an offset well. Losses in depleted sands and tight hole when pulling through coal seams
was observed. These events were used to calibrate and validate the geomechanical model

Rock Properties Construction

Elastic properties represent the elastic deformation behaviour of the rock, which is dependent upon the
rock type. For isotropic elastic materials, the stress-strain behaviour can be described by two parameters,
typically Young’s modulus and Poisson’s ratio, which can be determined from the stress-strain curves of
laboratory triaxial tests. However, as the core tests are only available at discrete depths, it is necessary to
use the log data to derive the dynamic elastic properties. Dipole sonic measurements provide the
compressional and shear slowness; and together with the bulk density log, dynamic elastic properties can
be calculated using the equations below:
Dynamic shear modulus
(1)

Dynamic Poisson’s ratio

(2)

Dynamic Young’s modulus

(3)
SPE-176422-MS 5

The dynamic Young’s modulus is typically higher than the test-derived static Young’s modulus, due
to lower and rapid loading strains imposed on the rock by sonic logging tools. However, the core-derived
Poisson’s ratio is usually smaller than log-derived Poisson’s ratio. The log-derived dynamic properties are
then converted to static elastic properties and calibrated with core data.
Calibration with Core Data
Extensive triaxial tests on shale and coal cores were conducted to provide the rock Young’s modulus and
Poisson’s ratio. The dynamic Young’s modulus and Poisson’s ratio were derived from the available log
data as described above, which were then calibrated to core test data. Fig. 4 shows the measured core
elastic properties vs depth. Wide variations in shale properties were seen between wells and depths.
However, coal mechanical properties were more consistent. Horsrud’s (2001) public correlation for shale
Young’s modulus was calibrated with core data to derive the static Young’s modulus of shales. Fig. 5
shows one such correlation for shales. Similar correlations were developed for sands and coals. Fig. 6
shows typical elastic and strength properties obtained in the target interval in one well.

Figure 4 —Coal and shale Young’s modulus and Poisson’s ratio in Mutiara area

Figure 5—Regression between shale static Young’s moduli from core test data and correlation. Shale properties derived from public
correlations were calibrated to core test data. The scatter in measurements shows the heterogeneity of the overburden shales, which
was handled though sensitivity analysis of input parameters on the mud weight window
6 SPE-176422-MS

Figure 6 —Calibrated rock elastic and strength properties in coal, sand, and shale sequence in an offset well

Stresses and Pore Pressure

Pore pressure in the area is generally normal at shallow depths. The pore pressure in coals is difficult to
measure, so it was taken as normal for this study. The vertical stress is defined by the weight of rock and
can be determined by integrating the density log. The overburden stress at the target depths is slightly
lower than 1 psi/ft. The maximum horizontal stress direction can be obtained from the orientation of
breakouts and/or drilling-induced fractures, or from the sonic shear anisotropy. In this case the maximum
horizontal stress direction was very well constrained from a number of shear sonic anisotropy logs (Sinha
et al., 2000) and is shown as a reference in Fig.1.
Minifrac operations conducted during hydraulic fracturing provided critical data on fracture closure stress
in coals which were very beneficial in constraining and calibrating the minimum horizontal stress. Recent
SPE-176422-MS 7

developments in advanced borehole acoustics have enabled computation of horizontal stresses from shear
radial profiles (Sinha et al., 2010). The shallow sands in the area exhibited very high shear sonic anisotropy,
as shown in the green shaded area in Track 6 of Fig. 7. Minimum and maximum horizontal stress magnitudes
in sands were inverted from the sonic data, which are shown as pink and blue dots in Track 7. The overall stress
profiles were obtained with poro-elastic horizontal stress method (Blanton and Olson 1999).

Figure 7—Stress model in an offset well

8 SPE-176422-MS

The stress anisotropy in the area varied between 5 and 35% depending on the lithology. Sands
exhibited the highest stress anisotropy, followed by shales, and the lowest stress anisotropy was observed
in coals. Overall, the stress regime was normal at the shallow depths where the target coals were to be
drilled. Fig. 7 shows the stress profile across some shallow horizons.

Wellbore Stability Analysis

The constructed stresses and rock strengths profiles were used to conduct the wellbore stability analysis
of the planned horizontal wells. The mud-weight windows for the build-up and horizontal sections were
computed and drilling risks assessed. Due to the narrow stable mud-weight window and the desire to
reduce reservoir damage from using too high mud weight, a probabilistic drilling risk analysis based on
Depth of Damage (DoD) methodology was conducted (Higgins et al.’s 2013). This risk-based approach
classifies the risks of drilling with mud weight lower than the breakout mud weight. The traditional
breakout curve is labeled as 0% damage mud weight. Two more shear failure curves are calculated for 5%
and 10% depth of damage of the borehole radius (%).
The mud-weight window of one of the planned wells is shown in Track 6 of Fig. 8. The ideal mud
weight is above the breakout curve and lower than the mud-loss gradient; that is, in the white region.
However, as can be seen, the depletion in sands at 2,132 ft has reduced the mud-loss gradient, whereas
the shales require a high mud weight for maintaining stability. Consequently, no single mud weight in the
section can satisfy all wellbore stability requirements. The risks of using mud weights lower than breakout
mud weight can be assessed through the Depth of Damage (DoD) curves. The deeper the failure extension
into the formation, the more failed rock material will fall into the wellbore and the higher the risk of
encountering drilling problems.
SPE-176422-MS 9

Figure 8 —A section of wellbore stability model showing mud-weight window

In Fig. 8, the area between breakout and 5% DoD mud weight is shaded yellow, between 5% and 10%
DoD mud weight is shaded orange, and less than 10% DoD mud weight is shaded red. Mud weight in the
yellow region means that shear failure has initiated at the wellbore wall and will extend up to 5% of
borehole radius (using liner elasticity assumptions). Similarly, the orange and red shaded areas denote
rock failure extension up to 10% and greater than 10% of borehole radius, respectively. Thus, the three
shaded mud weight areas—yellow, orange and red— can denote low, medium and high risk of wellbore
collapse when using mud weights lower than breakout mud weight. The black curve is the recommended
10 SPE-176422-MS

mud weight, which is a compromise between wellbore collapse and mud losses and it is kept above the
orange area (medium risk).
The breakout, 5% DoD, and 10% DoD mud weights were defined for the buildup and coal sections,
as shown in Table 1. The wellbore stability modeling also highlighted some drilling risks which needed
to be addressed adequately in the initial planning stage. The identified drilling risks are listed in Table 2.

Table 1—Suggested mud weights for planned wells

Buildup Section Horizontal Section

Mud Weight Limits ppg ppg

Breakout Mud Weight 12.8 11.0

5% Depth of Damage Mud Weight 11.9 10.0
10% Depth of Damage Mud Weight 10.9 9.0
Minimum Horizontal Stress Gradient with Depletion 9.8 15.0
Minimum Horizontal Stress Gradient without Depletion 13.5 N/A
Breakdown Gradient 14.9 18.5

Table 2—Key drilling risks and suggested mitigation measures

Description Risk Suggested mitigation measures

Shale instability in shallow section Severe Use appropriate mud weight. Avoid swab& surge
Mud losses in shallow depleted sands Severe Mud design to control losses
Differential sticking in depleted sands Moderate Mud and drilling design to reduce differential sticking
Cavings production and instability in coals due to low mud weight Moderate Maintain proper mud weight, adopt gentle drilling practices and
monitor for wellbore instability in real-time
Coal cavings due to mud invasion, pressurization and lubrication Moderate If possible minimize mud invasion through mud design. Otherwise
maintain good hole cleaning

Drilling Performance vs Prediction

For the build-up section of both wells, a variable and gradual increase in mud weight approach was
chosen. We began with 10.5 ppg for drilling through shallow depleted sands and a gradual increase to 12.5
ppg with lost circulation material (LCM) to stabilize the well while penetrating the lower shale sections.
Referring to Table 2, the uppermost mud weight of 12.5 ppg was 0.6 ppg above the 5% DoD mud weight
and 0.3 ppg below the 0% breakout limit, so as to allow for a swab margin above the 5% DoD mud weight.
While drilling this section in either well, minimal mud losses were experienced, although it is unknown
whether the sands were not as depleted as expected or the pre-inclusion and treatment with LCM
prevented any notable losses. The wells remained stable throughout drilling of these sections, and minimal
cavings were observed. The only notable drilling event was located around the depleted sands and
transition from shale to coal at the targeted seam entry locations. The cavings were predominantly platy
and angular shale cavings which stopped quickly with the correct treatment and related drilling proce-
dures.
For drilling the horizontal inseam section of each well, a brine-based mud was used up to the saturation
point, with the remaining weighting agent being a semi-invasive material that was to be later removed by
acidizing. The first well started at 10.2 ppg (referring to Table 2), which was 0.2 ppg above the 5% DoD
mud weight and 0.8 ppg below that of the 0% breakout limit with allowance for a ⫹/-0.5ppg swab/surge.
Approximately halfway through the drilling of the inseam section, the first signs of caving were observed,
which were used to identify the wellbore failure modes (Aadnoy et al., 2009, pp. 381-391). These cavings
were indicative of failure along weak fracture planes, but were considered minor and drilling continued
without adjustment in mud weight or any other treatment.
SPE-176422-MS 11

Shortly after this observation, larger chunks of coal were observed, some of which were angular and
others platy. Fig. 9 shows some of the cavings observed in the returns. The angular cavings indicated shear
failure and mud weight lower than breakout mud weight, whereas the platy cavings were a result of failure
of the weak fracture planes. In view of this information, the mud weight was gradually increased to 10.5
ppg (0.5 ppg above the 5% DoD mud weight) and the well was drilled and TD’d with no further drilling
issues.

Figure 9 —Cavings observed in inseam section

In the second well, the coal stability was more challenging. The section drilling started with mud
weight of 10.5 ppg. However, it had to be quickly increased to 10.7 ppg, 10.9 ppg, and then 11.2 ppg at
TD to stabilize the well for final completion. Referring to Table 2, the 11.2-ppg mud weight was 0.2 ppg
above the 0% breakout mud weight, although it was required to stabilize the wellbore for final completion.
The large amount of platy coal cavings seen during drilling cannot be explained based on shear failure
of the coal alone. The cavings primarily were due to failure of cleats, which resulted in sliding of coal
material into the wellbore. Such weak plane failures are difficult to mitigate just by mud weight alone.
Wells drilled at relatively low angles to cleat planes are more susceptible to this type of failure; mitigating
them requires mud weights that can be higher than the breakout mud weight. On the other hand, wells
drilled perpendicular or at high angles to cleats planes are less susceptible to weak plane failures. Proper
simulation of the weak plane failures can be conducted using a suitable weak plane failure model (Yan
et al., 2014). Adding loss-control material and plugging additives can help reduce the failure severity
through plugging the fractures and by preventing pressurization and lubrication of fracture planes.
However, such measures need to be carefully evaluated against potential formation damage that they may
cause and their impact on the overall production potential of the well.
Conclusions
The study highlighted various drilling challenges in both the planning and execution of the horizontal
wells in Kalimantan coal seams which was integral in the overall success of this project.
Mud losses from the depleted sands introduced additional risks in an area well known for shallow gas
hazards. Some shale collapse instability was encountered in the build section. Coal cavings from
12 SPE-176422-MS

insufficient mud weight, failure of fracture planes and fracture pressurization and lubrication resulted in
a host of additional challenges in successfully completing the inseam horizontal sections within the coals.
The case study showed that incorporating a reliable geomechanics assessment and model at the
planning stage is integral to reducing the risks from these drilling challenges. The mud weight window
will define the mud weight program and ECD limits, which will impact bottom-hole assembly selection,
mud design, tripping speeds, etc. For successful drilling, the geomechanical limits need to be understood
at the rig site so that the drilling problems can be dealt with in real-time through onsite monitoring of
wellbore instability indicators like cavings, torque and drag and other drilling parameters. The mud weight
program needs to consider both shear failure of coals and failure along fracture planes. Production of
excessive cavings can lead to many drilling problems, including stuck pipe. It should be note that each
geographical area has its own characteristics that need to be properly understood through proper data
acquisition planning, core testing and geomechanical study.

Acknowledgements
The authors would like to express their gratitude to the management of Vico Indonesia and Schlumberger
for permission to publish this paper. Special thanks to the Unconventional Team in Vico which provided
data and information at various phases of this study to ensure the operational success and preparation of
this paper.

References
Aadnoy, B., Cooper, I., Miska S., et al. 2009. Advanced Drilling Technology. SPE.
Blanton, T.L. and Olson, J.E. 1999. Stress Magnitudes From Logs—Effects of Tectonic Strains and
Temperature. Reservoir Evaluation & Engineering 2 (1): 62–68. SPE-54653-PA
GFZ German Research Center for Geosciences. 2009. World Stress Map. http://dc-app3-14.gfz-
potsdam.de/
Higgins S., Krepp T., Frydman, M., et al. 2013. New Approach to Geomechanics Solve Serious
Horizontal Drilling Problems in Challenging Unconventional Plays. Unconventional Resource
Technology Conference, Denver, CO, USA, 12–14 August. SPE-168675-MS.
Horsrud, P. 2001. Estimating Mechanical Properties of Shale from Empirical Correlations. SPE
Drilling & Completion 16 (2): 68 –73. SPE-56017-PA
Sinha B., Ouellet, A., and Berard, T. 2010. Estimation of Principal Horizontal Stresses Using Radial
Profiles of Shear Slowness Utilizing Sonic Data From a CO2 Storage Site in a Saline Formation
in Germany. SPWLA 51st Annual Logging Symposium, Perth, Australia, 19 –23 June. SPWLA-
2010-32038
Sinha, B.K., Kane, M., and Frignet, B. 2000. Dipole Dispersion Crossover and Sonic Logs in a
Limestone Reservoir. Geophysics 65 (2): 390 –407.
Thomas, M. 2013. Unlocking Indonesia’s CBM Potential. E&P Magazine, February, http://
www.epmag.com/unlocking-indonesias-cbm-potential-689451#p⫽full.
Yan, G., Karpfinger, F., Prioul, R., et al. 2015. Anisotropic Wellbore Stability Model and Its
Application for Drilling Through Challenging Shale Gas Wells, International Petroleum Technol-
ogy Conference, Kuala Lumpur, 10-12 December. IPTC-18143-MS.