SPE 87042
The Arun Gas Field in Indonesia: Resource Management of a Mature Field
Prabodh Pathak, Yan Fidra, Hanifatu Avida, Zulkarnain Kahar, Mark Agnew, ExxonMobil Oil Indonesia Inc. and Dodi
Hidayat, BPMIGAS
Copyright 2004, Society of Petroleum Engineers Inc.
This paper was prepared for presentation at the SPE Asia Pacific Conference on Integrated
Modelling for Asset Management held in Kuala Lumpur, Malaysia, 29-30 March 2004.
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, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes 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 where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836 U.S.A., fax 01-972-952-9435.
Abstract
The Arun gas field is a giant hydrocarbon resource in
Indonesia and has been supporting Liquified Natural
Gas (LNG) sales to the Asian market since 1978. The
ultimate gas recovery is expected to be about 94% of
initial gas in place, and the ultimate condensate recovery
is expected to be a very high 87% of initial condensate in
place. As the field has matured, reservoir pressure has
declined and well productivities have declined.
Mechanical wellbore failures have increased because of
increased downhole stresses. The lower reservoir
pressure has also increased the water vapor content in
the gas, and producing wells have started to loadup as
flow rates have decreased.
The application of new technologies has generated
numerous opportunities to extend field life. Recent work
has focused at maintaining productivity despite declining
reservoir pressures. This paper presents a summary of
Arun history and examples of recent resource
management practices, including compression upgrades
to reduce wellhead pressures, under-balanced drilling
into a pressure-depleted reservoir and problem
well restarts.
Introduction
The Arun gas condensate field in North Sumatra,
Indonesia has been and continues to be a prolific
producer of gas for the Asian LNG market. ExxonMobil
Oil Indonesia Inc. operates the field as a Production
Sharing Contractor to the Indonesian Oil and Gas
Upstream Regulatory Body, BADAN PELAKSANA
KEGIATAN USAHA HULU MINYAK DAN GAS BUMI
(hereinafter called "BPMIGAS"). The discovery of this
massive resource in October 1971 by the drilling and
testing of the Arun-1 discovery well spawned the birth of
the Indonesian LNG industry. Since then, Arun has been
a valuable resource that has been carefully developed
and nurtured to maximize its potential. Field
performance has been excellent and the application of
the latest petroleum technologies has extended field life.
Arun remains a great success story with excellent
recovery and high well productivity. Arun has now
matured and this paper presents the challenges
associated with managing this mature asset to meet
market commitments.
The Arun Field is located in the Aceh Province of North
Sumatra, straddling the coastal plains between the
Straits of Malacca on the north and the Barisan
Mountains on the south (Figure 1). 1 After the Arun-1
discovery well encountered more than 1000 ft of
productive limestone in the thickest part of the reefal
structure, 13 additional delineation wells were drilled to
map the outer limits of the reservoir.
Arun was developed from 4 clusters, and a total of 118
wells have been drilled, most of them deviated holes
(Figure 2). The initial reservoir pressure was 7115 psi
and the temperature 351 oF at a datum depth of 10,050
ft (Table 1). The field is approximately 18.5 km long and
5 km wide and covers approximately 23,240 acres. The
average thickness of the Arun limestone is 495 ft, and
the formation has an average porosity of 16.1%, with a
low connate water saturation of 10.7%.
The Arun reservoir fluid is a retrograde gas condensate
with 50.4 barrels per million standard cubic feet
(b/MMscf) of 48o API condensate and a dew point of
4450 psi - which is 2665 psi below original reservoir
pressure (Figure 3). An extensive fluid collection and
analysis program was conducted to characterize the
reservoir fluids and compositional simulations were
performed using state-of-the-art Equation of State (EOS)
simulation models.2,3 The importance of the water
component was recognized early - both as a source of
dissolved gas in the liquid connate water phase and the
2 SPE 87042
aquifer, and also as a vapor in the free gas phase.3,4 The
initial water fraction in the gas phase is 4.1 mole % and
the gas contains about 14.1 mole % carbon dioxide and
other non-hydrocarbons.
Field development planning began in mid-1972 in
conjunction with the LNG marketing efforts.2 A high
degree of reliability was designed into the system to
ensure a dependable supply of LNG to the buyers. Also,
gas re-cycling was recognized as being essential to
increase the recovery of condensates from the field.
First gas was produced in May 1977, gas recycling
began in July 1977, the first tanker of condensate was
exported in October 1977 and the first LNG tanker was
loaded in October 1978. Field production continued to
ramp up until the maximum condensate production rate
of 130,000 bpd was reached in 1989 and the maximum
gas production rate of 3,500 MMscf/D in 1995 (Figure 4).
At the peak in 1993, about 900 MMscf/D of dry gas was
reinjected back into the reservoir. An NGL plant was
commissioned in 1988 and NGL production continued
until 1998 when gas reinjection was discontinued. The
field has been declining since 1997 and currently
produces around 700 MMscf/D with 22,500 bpd of
stabilized condensate.
The Arun reservoir had an initial gas in place of 16.8
TCF (dry gas) and an initial condensate in place of 840
million barrels (Table 2). Currently about 90% of the gas
has been produced. The ultimate recovery is expected to
be 15.7 TCF dry gas and 730 million barrels condensate,
or 94% of the initial gas in place and 87% of the initial
condensate in place. This ranks Arun as one of the
highest recoveries worldwide for a gas condensate field.
The unique combination of reservoir management,
excellent reservoir quality, state-of-the-art drilling and
subsurface engineering technologies, and surface
facilities design and modifications has yielded these
impressive recovery factors.
Reservoir depletion has been managed with the
industry's first compositional simulator that modeled all
aspects of the production chain - reservoir, wellbore,
surface facilities and LNG production.5 Thus, production
optimization for changing market conditions, operational
difficulties or facilities modifications could be
accomplished using a single model. In addition, the
important reservoir mechanisms - gas expansion, gas
liberation from the connate water and aquifer, increased
water vaporization from the aquifer into the gas phase,
and limited aquifer movement were all identified and
modeled to optimize the depletion plan. Simulator
development has continued to add new features and
simulation results have been used extensively for the
past 25 years in field development decision making.
Drilling and subsurface engineering technologies were
applied to enhance productivity and improve condensate
recovery. 6-8 Innovative drilling techniques were devised
to handle the over-pressured Arun reservoir in the initial
drilling program. As productivity decreased later in the
field life, big bore wells were drilled to reduce frictional
pressure losses for gas flow. Horizontal wells were
drilled in the condensate-rich parts of the formation to
increase productivity and condensate recovery.9 Coiled
tubing workovers were designed and implemented for
increasing productivity while ensuring wellbore stability
and improving flow rates.
The challenges have changed as Arun has matured, but
the application of new technologies has generated
numerous opportunities to extend field life. Recent work
has focused at maintaining productivity despite rapidly
declining reservoir pressures. The results of this recent
work - including compression upgrades to reduce
wellhead pressures, under-balanced drilling into low-
pressure depleted reservoirs and problem well restarts
are presented here.
Geological Description
The Arun limestone is a reefal carbonate complex of
Early to Middle Miocene age.10,11 The first sediments to
be laid down on top of the basement were thin, localized
Oligocene sandstones of the Bruksah Formation
followed by the shales of the Bampo formation. The Arun
reef complex developed on top of the Bampo shales. A
major unconformity occurred on top of the Arun
limestone and a 1200 ft thick sequence of marine shales
called the Baong Formation was deposited on top of the
Arun formation. The Arun reef was surrounded on all
sides by overpressured marine shales that provided an
effective seal to allow the Arun gas to accumulate. In
addition, these seals are most probably responsible for
the extremely weak water drive in the reservoir.
Twenty-five wells were cored, most through part of the
Arun Formation and a few through the entire interval.
These cores and the related core analyses provide a
valuable source of detailed information about the
lithofacies and petrology. Log correlations have been
difficult within the reservoir due to the lack of shales or
other marker beds in the thick carbonate section.
However, a detailed core-to-log correlation analysis did
show the presence of thin (3-9 inch thick) marker beds of
wavy to laminated skeletal wackestones that appear to
correlate with log response across larger distances in
the field.
Lithologic analysis shows five main depositional facies -
Reef, Near-Reef, Inter-reef lagoonal, Slope, and Middle
Shelf environments (Figure 5). Reef facies occur
primarily along the center spine of the buildup while
flanking bodies of Near-Reef facies are developed on
the wide, southern part of the complex. Lagoonal facies
are found along the sides of the buildup or in large inter-
reef channels.
SPE 87042
The porosity distribution within Arun is related to
carbonate diagenesis that was initially controlled by the
lithofacies distribution. The buildup experienced strong
meteoric diagenesis in which two main pore types were
created - micropores formed by re-crystallization and
macropores formed by dissolution. Vuggy macroporosity
is found mainly in Reef Facies while moldic
macroporosity occurs mainly in Lagoonal Facies.
Approximately 25% of the total porosity is due to the
facies-controlled macroporosity. The Reef facies have
the highest porosity and permeability. The most
pervasive porosity is intercrystalline microporosity that is
found throughout the reservoir and accounts for the
remaining 75% of total porosity. Inter-particle porosity is
found mainly in Near-Reef facies and a small fraction of
fracture porosity is found in all the facies.
The high porosity Reef facies predominates at the top of
the reservoir while the lower porosity Middle Shelf and
Lagoonal facies are present deeper and at the edges.
Permeabilities in the Reef facies that have well-
developed macroporosity are an order of magnitude
higher than other facies in which micropores dominate.
Thus, the Arun reservoir generally has higher
permeabilities at the top throughout the reservoir (Figure
5). This characteristic has contributed tremendously to
the high recoveries in Arun but it also impacts drilling
operations, workover effectiveness and wellbore
stability, as discussed later in this paper.
The Arun Field has closure in all directions with a
bounding fault and a steep reef edge on the east, a
structural plunge on the north and north-west, and a gas-
water contact on the south and west. The gas-water
contact appears to be lower towards the south, with a
dip of about 650 ft from the north end to the south. The
cause of this tilt has been discussed extensively. 10,11
Hydrodynamics has been ruled out because of the lack
of access to reservoir in the north and an outlet in the
south, and also because the influx of fresh water would
have flushed out the salty aquifer water in the Arun. Two
other theories have been proposed - a fault-block
controlled concept with differing gas-water contacts in
each fault-block, and a lithofacies controlled concept
with more microporosity in the north leading to larger
transition zones and higher contacts. There is little direct
evidence for substantiating one of these theories and
there is most probably an element of both lithofacies and
fault control in determining the gas-water contact.
The simple reservoir description of the Arun limestone
was modified as the lean gas injection was history-
matched, and the existence of reservoir heterogeneities
became evident. First, the permeability of the laterally-
extensive Reef facies had to be modified to match the
north-south pressure gradient.5 Second, tracking the
movement of the injected gas by mapping the dew point
pressures of the produced gas indicated the presence of
a north-south barrier in the western part of the field and
3
a high permeability conduit for gas movement in the
south-east part. Lean gas injected in west injection wells
fingered through to Cluster II in the north and Cluster IV
in the south without having any impact on Cluster III
wells. Also, gas injected in the south-east wells broke
through to wells in the eastern part of Cluster IV very
fast. In the big picture, the field does behave like one
high-permeability, well-connected carbonate deposit, but
there are local discontinuities that impact only certain
portions of the reservoir.
Challenges in Managing a Mature Arun Field
Arun currently produces at an average well rate of 11
MMscf/D from 65 wells. This is, of course, much less
than the maximum well productivities of more than 200
MMscf/D and the average well productivity of 130
MMscf/D reached in 1978 (Figure 6). The primary cause
of the lower productivity is reduced reservoir pressure.
The average reservoir pressure in Arun has fallen from
the initial 7115 psi to about 700 psi currently, and as
shown in Figure 7, the compositional reservoir simulation
model matches the field data very well. There is a
reservoir pressure gradient in Arun with pressures in the
northern clusters (Clusters 1 and 2) measuring 500-600
psi, and the southern clusters (Clusters 3 and 4) having
higher pressures, about 700-800 psi. Differential
depletion and injection have caused this pressure
distribution across the reservoir. Higher pressures exist
in the south, in part, because of higher gas injection in
previous years.
In addition to declining reservoir pressure, Arun
production has also declined because of wellbore
failures and well loading. The number of producing wells
started to decline in the early 1990's as wellbores failed
due to increased mechanical stresses. Well loading has
become a critical issue recently as the declining
reservoir energy is not sufficient to lift condensed liquids
out of the tubing.
Wellbore Failures
Most Arun wells are deviated from a cluster surface
location. The primary drilling phase in the field was
conducted using 7" tubing that was considered the
largest practical tubing size in the industry in the late
1970's. The 9-5/8" tubing completions in the 1992-1995
Big Bore drilling program were the largest gas
completions in the world at that time.12 Additionally,
horizontal completions were used to increase
productivity from low-permeability zones that contained
pockets of rich gas that had been bypassed by the
injected gas.
Two major wellbore failure mechanisms have been
identified - subsidence and Arun pay zone failure. The
subsidence failures occurred in the overpressured
Baong shale above the Arun limestone. Subsidence
4 SPE 87042
failures are diagnosed by shale production, severe hole
shift, and wireline tag-ups above the top of the Arun. Pay
zone failures occur when the Arun limestone itself
collapses. Pay zone failures are normally indicated by
limestone production, and wireline diagnostics indicate
no hole shift and the tubing is clear to the formation. The
root cause of both failure mechanisms is the limited
strength of the Arun limestone and the severe
mechanical stresses induced by the reduction in
reservoir pressure.
A total of 34 wells have failed, and subsidence failures
dominated in the 1994-1999 timeframe. In later years,
more formation failures have been recorded, but there
have also been a few wells where the failure mechanism
was most probably a combination of both subsidence
and formation failure. In 2000, the highest number of
well failures were recorded, but well failures have
declined during the past 3 years to about 2-3 failures per
year. The ratio between wellbore failures and formation
pay zone failures is now about equal. Eight of the failed
wells have been restored and are still producing.
Rock mechanics studies of the Arun limestone indicate
that the compressibilities are in some cases as high as
100 mips. In addition, the high porosity rocks, which are
present in greater abundance in the Reef and Near-Reef
facies, are also susceptible to pore collapse. Pore
collapse increases formation subsidence and also
accentuates formation failure. The large
compressibilities induce increased subsidence with
decreasing reservoir pressure. As the Arun limestone
subsides, the overlying overpressured Baong shale also
moves creating high shear stresses in the casing in the
Baong. Subsidence also increases the hoop stresses in
the borehole and the borehole wall starts to fail.
Stimulation treatments in which acid was injected down
the tubing, and in which most of the acid dissolved the
upper high permeability zone may have also contributed
to increased subsidence and pay zone collapse.8
An Arun wellbore review team was formed to analyze
the well failures with key members coming from
ExxonMobil Oil Indonesia Inc., ExxonMobil Upstream
Research Company and ExxonMobil Production
Company Central Technology organizations.
Geomechanical models were developed to analyze
wellbore stresses and predict future failure rates. A
Three-Dimensional Analytical Subsidence and
Compaction (TASC) model was developed in
conjunction with a Quantitative Risk Analysis (QRA)
model for Arun wells. TASC uses analytical solutions for
analyzing subsidence in reservoir and overburden grid
cells. The Arun reservoir properties, geometry and
pressure for each well are imported from the reservoir
simulation model and the TASC model analyzes the
bending strains, reservoir subsidence, principal strains,
and shear stresses. The mechanical loads are then
combined with casing strength data in the QRA model to
quantify the risk of well failure.
TASC/QRA model results show that bending strains
caused by subsidence did not increase appreciably for
most wells until 1985. Figure 8 shows that for typical
Arun wells, the onset of well deformations, kinks in the
tubing and casing strings, and wellbore restrictions were
first noticed in the 1987-1990 time frame. The maximum
strains are reached around 1998-2000, and this time
period also corresponds to the largest number of well
failures. The study also identifies the remaining Arun
wells that are at risk of subsidence or pay zone failure.
Another important observation is that wells drilled
parallel to the natural fracture slip plane dip may be less
likely to fail. Cores taken in the Baong shale indicate that
the natural fractures are oriented in the north-west
direction. Thus, wells drilled in the north-west or west
direction have a reduced risk for failure, although they
have some of the highest bending strains in the field.
The Arun wellbore team recommended that new well
locations should be at least 1000 ft away from existing
downhole locations to avoid excessive localized
subsidence. Wellbores in the north-east end of the field
may have failed because of excessive clustering and
reduced spacing. In addition, the team recommended
that all new wells should have a liner penetrate at least
the top 30-50 ft of the high porosity, high compressibility
Arun limestone to prevent the overpressured Baong
shale from entering the wellbore as the formation
subsided. Highly deviated and horizontal wells were not
recommended because of the risk for pay zone failure.
Well Loading
Liquid loading in gas wells occurs late in field life when
gas flow rates decrease, and gas velocities in the
wellbore are not sufficient to lift the condensed liquids to
the surface. Liquids start to accumulate in the wellbore
until the flow ceases when gas pressure cannot lift the
column of liquid. 13 It was recognized early in the Arun
development that liquid loading would become
increasingly important as the field matured. 5 A series of
studies were performed to analyze and quantify the
mechanism of liquid loading and well tests were
performed to check the applicability to Arun conditions.
Both hydrocarbon and water condense from the gas
phase in the wellbore. The fraction of each liquid phase
that condenses depends on the PVT properties of the
gas-condensate-water components and the flow
properties of the mixture. Both pressure and temperature
decrease as the fluids move up the wellbore, and the
dominant liquid fraction will depend on the phase
equilibrium properties at the local conditions along the
wellbore. This paper presents the results from the most
recent compositional wellbore hydraulics study.
SPE 87042
The fraction of water in the gas phase increases
dramatically as reservoir pressures fall below 1000 psi in
the high-temperature Arun reservoir environment (Figure
9). Arun gas in the reservoir equilibrates with the two
sources of water in the reservoir - the aquifer and the
connate water. The vaporization of water at the gas-
water contact may impact aquifer movement. Also, the
vaporization of the connate water may improve gas
permeability. The field water production is also shown in
Figure 9 and it closely matches the water content in the
equilibrated gas produced. Thus, the produced Arun gas
is saturated with water at reservoir conditions. Chemical
analyses also indicate that the produced water is very
fresh compared with aquifer water salinities.
A series of well tests were performed in 1999 to
measure well loading rates. In each well test, wellhead
parameters such as pressure, temperature and rate
were measured as the well was choked back and
production rates decreased. The well was produced at
each reduced rate for a few days to ensure that the rate
was stable. Once the well loaded up, it was quickly
unloaded by reducing the wellhead pressure and
removing the wellbore liquids. Figure 10 shows the
results for a typical Cluster I well that was producing 10
MMscf/D before the test began. The well was initially
produced at a lower rate of 8 MMscf/D for a few days.
The wellhead pressure decreased by about 60 psi after
about 100 hours because an extra compressor was
turned on to meet increased gas demand from the field.
Wellhead pressure was increased to 350 psi at 320
hours and the gas rate initially decreased but then
stabilized around 6.5 MMscf/D. Further increases in
wellhead pressure reduced the gas rate dramatically to 2
MMscf/D. The well loaded up as both the well rate and
wellhead pressure continued to decrease. Tests on other
wells showed similar behavior - the well would flow at a
stable rate for many days but a small reduction in rate or
increase in pressure would trigger the instability and load
up the well in a few hours.
A compositional wellbore hydraulics model was
developed to analyze well loading. The model uses a 9-
component Equation-of-State (EOS) characterization,
which is similar to the characterization used in the
reservoir simulator. The original PVT model was updated
to reflect current reservoir conditions. The EOS
characterization was input into a well performance model
that models both the fluid flow and the heat transfer
around the wellbore. The tubing component sizes and
other properties were entered into the wellbore model
and the overall length of the wellbore was discretized
into 21 individual sections of approximately 500 ft length
each. The flow properties and liquid holdup in different
sections of the tubing could then be analyzed. The
OLGA steady-state correlation was used to model the
fluid flow although it slowed the simulation runs
compared with other correlations. Reservoir flow was not
modeled in this study and the simulation was confined to
5
the tubing. Because the fluid flow and heat transfer
equations are being solved simultaneously, model
stability improves if the bottomhole pressure and
temperature are specified for a particular flow rate and
the model calculates the resulting surface temperature
and pressure. Attempts at specifying both bottomhole
and surface conditions usually made the model unstable.
Results show that at high flow rates the pressure and
temperature decrease uniformly as the gas travels up
the tubing (Figure 11). Wellhead temperatures are high
(about 270 oF) and wellhead pressures are
approximately 200 psi less than bottomhole pressures.
The fraction of liquid dropout in the wellbore is small,
less than 1%, and the fluid flow behavior is similar to
single-phase gas flow. As the gas flow rate decreases,
liquid holdup in the wellbore increases significantly, and
the flowing gas phase has to overcome frictional
resistance from the liquid held in the wellbore. The two-
phase flow regime is annular flow at high flow rates but
as the liquid holdup increases it becomes slug flow
(Figure 12). Wellhead pressures decrease until the
minimum stable flow of 3 MMscf/D is reached. At this
low flow rate, liquid holdup is around 10% for the top
7000 ft of the tubing but it increases to more than 20% at
depths around 10,000 ft (Figure 13). The water phase
fraction is almost 90% in the top 6000 ft and thus, water
is condensing more than the hydrocarbon phase (Figure
13). At depths greater than 7000 ft, the liquids are a
mixture of both water and hydrocarbon phases. The gas
produced from the reservoir is in the two-phase
envelope of the PVT diagram initially at higher depths,
and the liquid condensate phase does condense at the
higher pressures and temperatures, but the liquid
condensate phase revaporizes at the lower pressures in
the upper part of the tubing. Wellhead temperatures
continue to decrease as the flow rate is decreased, and
wellhead temperatures below 200 oF generally indicate
a well that has loaded-up. Well loadup rates increase as
the tubing size is increased, so Big Bore wells with 10"
tubing are expected to load up at 3.75 MMscf/D
(Figure 14).
The well loading rates are very low compared to most of
the industry because wellhead pressures are very low -
about 90 psi. The well loading rates are also lower than
the Turner correlation results of 4-6 MMscf/D because of
the simplifying assumptions made in the Turner
correlation about spherical droplet shapes and viscous
drag forces. In Arun, a substantial portion of the current
gas production is produced from low rate gas wells.
Almost 80 MMscf/D of the total 700 MMscf/D is
produced by wells making less than 6 MMscf/D.
Managing a Mature Arun Field for the Future
Arun is a critical field for meeting LNG delivery. Much
effort has been put into maintaining the production,
managing the reservoir and stemming the production
6 SPE 87042
decline. Three major efforts are discussed in this paper -
the new drill well program, the compressor and
dehydration upgrade project, and the problem well
restart program. Other initiatives, such as gas well
dewatering are also being considered to further
maximize recovery.
New Drill Well Program
Wellbore failures made it imperative that the failed wells
be replaced by sidetracking existing wells or by drilling
new ones to supply the required gas capacity. Seven
Arun wells were re-entered and sidetracked in 1997-
1999 to maintain productivity. However, more wellbores
were needed to meet supply quantities and a program
was initiated in 2000 to evaluate new wells. An important
factor in the decision making process was that Arun had
to be drilled underbalanced to prevent formation damage
in a reservoir where pressures had decreased
substantially. Drilling underbalanced at low pressures in
a high temperature environment at depths exceeding
10,000 ft presents a formidable challenge. Also, the
reservoir pressure continues to decrease by about 8 psi
per month at current depletion rates. Thus, decisions
about when to drill and the pace of drilling are essential
in program planning.
The reserves produced by Arun wells can be correlated
with their productivity and the time that the wells were
put on production. Thus, wells with high productivity, as
measured by the product of permeability and thickness
(Kh), usually produce more reserves. The Kh maps
generated from core, log and well test data are essential
in identifying the high quality reservoir areas, i.e., the
"sweet spots" on the Kh maps (Figure 15). Two other
factors were considered in choosing well locations. First,
clusters with higher reservoir pressures were given a
higher priority because the risks associated with
underbalanced drilling would be lower. Also, the
productivities would be higher there because of the
higher pressures. Finally, the probability of well failure
due to subsidence was lowest around Cluster III,
followed by Cluster IV, II and I. This ranking system
eliminated most of the locations around Cluster I and II,
leaving possible locations only around Cluster III & IV.
Once the bottomhole locations for the new wellbores had
been identified, the next major decision was whether to
sidetrack an existing well or to drill a new well. The
sidetrack could be drilled by cutting a window in the 9-
5/8" casing, and drilling into the Arun with a 7" or 4-1/2"
completion. Figure 16 shows the nodal analysis results
for the sidetrack completions and for a new Big Bore 9-
5/8" completion. There is little pressure drop in the
reservoir and most of the frictional resistance to gas flow
comes from the tubing. The Big Bore well produces
about four times what the 4-1/2" sidetrack completion
would make and 50% more than the 7" sidetrack. The
recovery for each well type would be proportional to its
initial rate, making the 9-5/8" Big Bore program much
more attractive. The capital investment to drill a
sidetrack is significantly less than the capital required to
drill a new well, but the risks associated with formation
subsidence around a sidetrack openhole completion are
also higher. The economics for all three cases yielded
positive results but the Big Bore well provided superior
economics. In addition, it provided the highest
productivity for meeting market demand, and thus, the
Big Bore new well program was selected instead of the
sidetrack option.
Seven new drill wells were selected for the drilling
program. A new optimized Big Bore well design was
used.12 The major innovations included the use of a 10"
tubingless completion string to replace the old assembly
of a 10" liner, 13-3/8" production packer and 9-5/8"
floating seal assembly. After the 10" liner is set, a 7" drill-
in liner is used to drill across the bottom section of the
Baong shale and case it off across the 8000 psi
transition into the Arun limestone. The underbalanced
drilling of the Arun limestone at 10,000 ft is
accomplished using a nitrogen-water mist as drilling
fluid. The well suspension procedures were modified to
ensure minimal formation damage. Directional profiles of
the wells were kept below 25O to maximize
wellbore stability.
The drilling program began on 18 June 2002 with the
spudding of the first well and was concluded on 6 Sept.
2003 with the completion of the final well. The program
showed conclusively that under-balanced drilling (UBD)
operations could be conducted at very low pressures -
for instance, the static bottomhole pressure during UBD
operations in one of the Cluster III wells was 560 psi.
The nitrogen-water mist was effective at removing
cuttings at these low pressures and wellbore damage
was minimized.
Eight wells have been drilled, of which seven successful
wells are currently being produced. One well had to be
re-drilled because of mechanical problems. Table 3
summarizes the initial rates through the wellbore. The
wells are performing as expected and the overall drilling
program has been successfully concluded. These wells
are probably the last wells to be drilled in Arun, but the
drilling technologies that were developed and refined in
Arun will continue to be used in other locations.
Compressor and Dehydration Upgrade
Booster compressors were commissioned in Cluster II
and Cluster IV in 1995 to reduce wellhead pressures. All
the gas from Cluster I is cooled, dehydrated and sent to
a gas-liquid separator before flowing to Cluster II where
it is combined with treated Cluster II gas for
compression. Similarly, all the Cluster III and IV treated
dry gas is compressed at Cluster IV for delivery. The
SPE 87042
compressed gas from Cluster II and IV is sent to PT
Arun LNG facilities through a 42" pipeline.
The dehydration facilities installed in 1992 were
designed to reduce the gas dew point to 150 psi. The
facilities used propane refrigeration to cool the gas to 40
o
F to avoid water dropout in the carbon steel
compression equipment and pipelines. Arun gas
contains about 14.1% carbon dioxide and thus, any
water dropout can lead to rapid corrosion. A joint
ExxonMobil Oil Indonesia Inc. and ExxonMobil
Production Company Central Technology task force was
commissioned in 2000 to evaluate alternatives for
reducing the operating suction pressure to 50 psi. A
number of alternatives were considered including the
installation of a new glycol dehydration plant, but the
most economic was the upgrade of the existing propane
dehydration system. This alternative had the added
benefit of a faster installation schedule and lower training
requirements because operations personnel were
already familiar with this technology.
In conjunction with the dehydration upgrade, the
compression facilities were also re-evaluated. The
original compression facilities consisted of six
compressor trains, three each at Cluster II and Cluster
IV. The compressor trains were designed to operate a
minimum suction pressure of 150 psi, and although the
suction pressure could be drawn down to 75-80 psi, the
compressors became increasingly inefficient at the lower
suction pressures. A number of alternatives were
considered including the re-wheeling of both compressor
stages, and re-wheeling or replacing the low pressure
stage. The task force recommended the replacement of
the low pressure compressor casing and the installation
of larger wheels to increase throughput and reduce
suction pressures to 50 psi. Because Arun gas rates are
decreasing and the Cluster II compressors have much
lower throughput than the Cluster IV compressors, the
upgrade of all six compressors was not recommended.
The optimum number of compressor upgrades was
found to be three, i.e., two compressors on Cluster IV
and one compressor on Cluster II. The upgrades allow
the early retirement of old compressor trains, which
helps reduce compressor fuel usage, turbine overhaul
costs, operating and maintenance costs, and emissions
of carbon dioxide and other gases.
The dehydration project and compressor upgrades were
completed in 2002. Figure 17 compares the performance
curves for the old compressor and the new modified
compressor. At comparable suction and discharge
pressures, the new compressor capacity is almost 75%
higher. For instance, at a suction pressure of 95 psi and
discharge pressure of 950 psi, the old compressor
throughput was limited to 120 MMscf/D while the new
compressor can handle 210 MMscf/D. The compressor
and dehydration upgrades have reduced anticipated
reservoir abandonment pressures to 270-290 psi and
increased Arun ultimate recovery by about 130 BCF.
7
An active facilities and reservoir surveillance team
monitors compressor performance. This interface of
reservoir performance and compressor performance is
essential in understanding the overall field performance.
There are limited tools available to integrate the
reservoir with the compressors. Most reservoir
simulators, including the Arun compositional model,
have a simplified compressor performance capability. On
the other hand, accurate compressor performance
predicting tools have limited reservoir simulation
capabilities. Although the Arun simulation model does
include simplified correlations for the compressor
performance, the short-term optimization of field
performance is done using a material balance model that
incorporates the subsurface nodal models for all the
wells and the surface network for collecting, treating and
transporting the gas to the compressor suction header.
This material balance model has been developed over
the past few years and all the Arun wells have been
history-matched. Multiple operating scenarios can be
evaluated quickly using this tool.
The compressor performance curves from Figure 17
have to be overlayed with the reservoir performance
curve in order to analyze the production rates. At Cluster
II, one old compressor and one new compressor work in
parallel to compress the Cluster I & II gas. The
compressor performance curve for this configuration of
compressors in Cluster II is shown in Figure 18 at an
outlet discharge pressure of 950 psi. Thus, as the
suction pressure increases, the compressor capacity
also increases. The corresponding reservoir
performance curve at the current operating conditions,
which is calculated from the material balance model, is
also shown in Figure 18. In the case of the reservoir, as
the suction pressure increases, the flow rate decreases
as the backpressure on the wells increases. The
intersection of the reservoir and the compressor curve
determines the daily operating capacity of the cluster,
and as shown in Figure 18, the material balance and
compressor models accurately predict field performance.
The reservoir performance curve is constantly changing
as reservoir pressure continues to decrease while the
compressor performance remains reasonably invariant
unless operating conditions change. Figure 19 shows
that the reservoir performance curve shifts to the right as
the drilling of the new wells increases productivity. The
decision to run two new upgraded compressors in
Cluster IV or to run the third old compressor also
depends on market demand, the additional rate from the
third compressor and the fuel usage in the compressor.
As shown in Figure 19, the new wells have increased
field deliverability and the analysis shows that 20
MMscf/D of additional gas can be produced if the third
compressor is also run.
At this time, all three compressors in Cluster IV are
being operated to compress the gas being produced
8 SPE 87042
from Clusters III and IV. The two compressors in Cluster
II on the other hand are not fully loaded and have
additional compression capacity. The Arun reservoir and
facilities team looked for innovative ways to move gas
from Cluster III to Cluster II to better utilize all the
compressors. A review of the field facilities identified
pipelines that were not currently being used. A pipeline
that had previously been used for gas injection was
recommended as the best way to move gas between
Cluster III and Cluster II. Preliminary field testing
indicates that about 15 MMscf/D of Cluster III gas can be
moved through this line to Cluster II for compression and
thus, the need to run the third compressor at Cluster IV
has been reduced.
Problem Well Restart Program
From the start of Arun production in 1977, the field has
had the luxury of excess production capacity. In the early
years, the gas re-injection program was used not only to
increase condensate recovery but to also act as a buffer
in case additional gas was needed for LNG production.
Thus, gas re-injection was curtailed for short periods if
LNG demand was high. In later years, Arun wells
remained so prolific that market demand could be met by
increasing production from low condensate-gas ratio
wells that had been choked back. In the current stage of
field life, significant excess capacity does not exist.
Consequently, low rate wells need to be watched closely
and immediate corrective actions taken if one of the
wells ceases to flow.
An active well restart program was initiated in 2000. The
first part of the program was a Problem Well Report
(PWR) system for tracking problem wells and simplifying
the process for restoring them. A PWR is issued to
identify production problems and is then used to track
progress on restoring the well. PWR's are generated by
reservoir engineers, subsurface engineers or operations
personnel for any unexpected change in well
performance - sudden drop in gas rate, solids
production, high/low annulus pressures or increased
water production. The PWR is discussed in bi-weekly
surveillance meetings and a multi-disciplinary team of
reservoir, subsurface, production operations and
facilities engineers analyzes the problem. The analysis
may involve nodal analysis, wellbore stress modeling or
sub-surface equipment checks. The team recommends
corrective action after the analysis is completed. When
the program was started, the time taken to recommend
corrective action could take up to 3 months, but as the
program has matured the timing has been reduced to
under 10 days.
A total of 30 PWR's have been issued for Arun wells, of
which 14 have been executed in the field with varying
degrees of success. The major well work
recommendations that have been implemented include
unloading the well to flare in order to clean the wellbore
of condensed liquids, installing flow-control valves to
minimize solids production, and re-starting wells that
were thought to have failed. Some of the PWR
recommendations were not successful because of fill in
the tubing or because of the catastrophic mechanical
failure of the casing. When the well work is completed,
the PWR is closed. If unsuccessful, a new PWR is
generated, the problem is re-studied and new
recommendations are made. The PWR process is a
dynamic method for monitoring well work, increasing the
responsiveness of Jakarta staff to the Arun production
operations, and reducing the time required to execute
work programs. The PWR program has paid off in an
additional 100 MMscf/D that has been returned
to production.
The second part of the Arun problem well restart
program is a change in the operating philosophy of the
field. Arun is part of a larger supply source that feeds the
PT Arun LNG facility. Other fields in the area, such as
Pase and South Lhoksukon, were brought on-stream in
1996-1998 to provide additional supply to meet gas
demand. Arun still remained at the core of the
operations, but the production from the other fields was
increased to balance the depletion between the core and
the satellite fields. In case of short-term operational
difficulties or demand fluctuations, Arun was used to
balance the production requirements. Thus, Arun wells
would be choked back or shut-in during periods of low
demand. As Arun has matured and reservoir pressure
declined, it has become more difficult to lift condensed
liquids out of the choked-back wells. Since late 2002, the
area operating strategy has been changed, such that the
satellite fields are used to balance demand. This
ensures the Arun wells are kept on-line at all times
thereby reducing well load-up associated with system
pressure variability.
The third part of the Arun well restart program was the
start of production from the Gas Injection Wells (GIW).
The GIW's are located on the western edge of the field
and are the lowest wells on the depth structure maps.
There was some concern that aquifer movement may
have flooded the productive gas zone in these wells. The
movement of the aquifer has been monitored by running
cased-hole neutron logs in observation wells that are
located between the GIW's and the main production
Clusters (Figure 2). The aquifer has moved 66 ft
between the startup of the field and 1998 based on the
logs in the observation well (Figure 20). Extrapolating
this trend to the current time indicated that some of the
GIW's could still be produced if they are perforated high
on the structure. Thus, a program to test the GIW's was
started in 1999 and well tests indicated that some of the
GIW's had high productivities. For instance, a GIW on
the south-west side of the field made 25-30 MMscf/D in
the 1999 well test, which was comparable to some of the
better wells in the field at that time. A program to repair
SPE 87042
and refurbish the GIW's was started in 2002 and
currently two GIW's have been put on production.
Conclusions
The Arun gas field development has been one of the
most technically challenging and also one of the most
successful gas developments in the world. The
technologies developed for the field have proven to be
extremely important in ensuring excellent ultimate
recoveries from a deep, high-temperature, high-pressure
reservoir. The success of Arun is based on the insight,
hard-work and efforts of engineers, operations
personnel, geoscientists, and business professionals
over the past 32 years.
The challenges have changed as Arun has matured, and
the lessons learned in optimizing the production from the
field include:
1. The initial design of the field development is critical in
ensuring the long-term success of the field. In the case
of the Arun development, most of the important reservoir
mechanisms and operational issues were identified
before the development started. For instance, the
decision to recycle gas from the onset of field production
has contributed to the very high ultimate condensate
recovery of 87% from the field.
2. Compositional reservoir simulators are extremely
useful throughout field life. The development and use of
compositional simulation has been essential in making
early predictions of rates and recoveries, in monitoring
the performance of the development, in understanding
the recovery mechanisms and in the later stages of field
life to analyze well loading.
3. Geoscientists need to be continually involved in field
management - from the early development planning to
the later infill drilling. The insights provided by the
detailed lithologic analysis have been essential in
understanding the movement of hydrocarbons within the
reservoir and in deciding the locations of the new wells
to maximize gas recovery.
4. Geomechanical modeling of the reservoir should be
done early in field life to understand the long-term impact
of reservoir depletion on wellbore stability and formation
subsidence. Accurate formation compressibility data
should be collected routinely as part of the core analysis
package. Wellbore design and tubular selection should
be based on formation subsidence and shale
movement analyses.
5. Drilling into depleted low-pressure reservoirs requires
state-of-the-art under-balanced drilling technology to
minimize reservoir damage.
9
6. Innovative methods for utilizing all existing
infrastructure in the field need to be evaluated.
Upgrading existing compressors and dehydration
facilities can reduce wellhead pressures and keep low
rate wells flowing longer. Periodic reviews of pipelines,
separation facilities and compressors can provide
innovative solutions to facilities debottlenecking and
improving operational efficiency.
7. A well-organized Problem Well Report system is
essential in maintaining field production. All operational
groups need to be involved in this effort - field
operations, reservoir engineering, subsurface
engineering, facilities engineering and wellhead and
wireline operations. A regular meeting schedule with all
personnel maintains excellent communication and fast
resolution of problems as they occur.
Acknowledgements
The authors would like to thank BPMIGAS, ExxonMobil
Oil Indonesia Inc., ExxonMobil Production Company,
and ExxonMobil Upstream Research Company for
permission to publish this paper. This paper represents
the work of many dedicated ExxonMobil* work-teams
over the past 10 years. In particular, the authors would
like to thank Alex Isnya, David Baxendale, John Turner,
Jon Benesch, Tim Busing, Tumiran, Mark Hoefner,
Rakhmadi Avianto, John Crane, Sarbini, Clif Jordan, Dwi
Sulistio, Jerry Fugate, Purwanto Widjaya, Zakaria
Zainun, Mohammad Hafizh, Terry Templeman, Teuku
Mirza, Dale Pittman, Meg O'Neill, Arlene Pollin, Michael
Kroeger, Ram Seetharam, and Ravi Vaidya for
their help.
* The term ExxonMobil as used in this material may refer to Exxon
Mobil Corporation, to one of its divisions, or to the companies affiliated
with Exxon Mobil Corporation, or to any one or more of the foregoing.
The shorter term is used merely for convenience and simplicity.
Nothing in this material is intended to override the corporate
separateness of those entities.
References:
1. Myers, S. D., "Arun Field : Model for Successful Gas
Development", SPE Paper No. 30150, Presented at SPE
PetroVietnam 1995 held in Ho Chi Minh City, Vietnam,
1-3 March 1995.
2. Seipel, J. R., "Arun Field Development Phase", SPE
Paper No. 10483, Presented at the Offshore South-East
Asia 1982 Conference, Singapore, 9-12 Feb. 1982.
3. Sutan-Assin, T., S. C. Rastogi, M. Abdullah, D.
Hidayat, S. Bette and R. F. Heineman, "Use of
Compositional Simulation in the Management of Arun
Gas Condensate Reservoir", SPE Paper No. 17683,
Presented at the 7th Offshore South-East Asia
Conference, Singapore, 2-5 Feb. 1988.
10 SPE 87042

4. Hasan, N., S. C. Rastogi, H. J. Ng, and N. R. Table 1

Nagarajan, "Phase Behavior of Retrograde Gas Arun Initial Reservoir Properties
Condensate-Water System Under High Pressure and
Temperature Conditions", OSEA No. 88203, Paper Initial Reservoir Pressure at Datum 7,115o psi
Presented at the 7th Offshore South-East Asia Reservoir Temperature at Datum 351 F
Conference, Singapore, 2-5 Feb. 1988. Datum Depth 10,050 ft
Productive Area 23,240 acres
5. Bette, S. and R. F. Heinemann, "Compositional Average Thickness 495 ft
Modeling of High-Temperature Gas-Condensate Average Porosity 16.1%
Reservoirs with Water Vaporization", SPE Paper No. Average Water Saturation 10.7%
18422, Presented at the SPE Symposium on Reservoir Initial Water Vapor Content 4.1 mole%
Simulation, Houston, TX, Feb. 6-8, 1989. CO2 and Other Non-Hydrocarbons 14.1 mole%

6. Bolt, L. H., M. Soepardi and D. Suherman, " Drilling of

Table 2
Arun Gas Field", J. Pet. Tech., pp. 771-778, May 1984. Arun Ultimate Gas and Condensate Recovery

7. Aladyrus, M. U. and T. P. Bordeleon, "Boosting

Initial Gas in Place 16.8 TCF
Deliverability in the Giant Arun Gas Field", SPE Paper #
Free Gas 16.3 TCF
22997, Presented at SPE Asia Pacific Meeting, Perth, 4- Solution Gas 0.5 TCF
7 Nov. 1991.
Gross Ultimate Gas Reserves 15.7 TCF
8. Bunnell, F. D. and M. M. Daud, "Coiled Tubing Gas Recovery Factor 94%
Simulations Eliminate Hole Failures and Condensate
Losses in Arun Field", SPE Paper No. 30681, Presented Initial Condensate in Place 840 million barrels
at SPE Annual Technical Conf., Dallas, 22-25 Oct. 1995.
Gross Ultimate Condensate Reserves 730 million barrels
9. Sunaryo, A. C., A. Mulia and M. D. Sulitra, "Horizontal
Condensate Recovery Factor 87%
Wells in Management of the Arun Reservoir", Proc.
Indonesian Petroleum Association, IPA 92-23.33, 21st
Annual Convention, Jakarta, October 1992.
Table 3
10. Soeparjadi, R.,"Geology of the Arun Gas Field", J. New Drill Well Performance
Pet. Tech., pp. 1163-1172, June 1983.
Initial Rate

11. Jordan, C. F. Jr. and M. Abdullah, "Arun Field- MMscf /D

Indonesia, North Sumatra Basin, Sumatra", in Foster, N. Well Pre-Drill Actual
F. and E. A. Beaumont, eds. "Stratigraphic Traps III,
Atlas of Oil and Gas Fields", AAPG, Treatise of 1 22 27
Petroleum Geology, pp. 1-39, 1992. 2 27 27
12. Benesch, J., Nazri Nor and M. M. Ngatijan, 3 29 26
"Optimization of Big Bore HTHP Wells to Exploit a Low- 4 29 30
Pressure Reservoir in Indonesia", SPE Paper # 87171,
To Be Presented at IADC/SPE Drilling Conference, 5 27 33
Dallas, 2004. 6 22 21

13. Turner, R. G., M. G. Hubbard and A. E. Dukler, 7 25 31

"Analysis and Prediction of Minimum Flow Rate for the Total 181 195
Continuous Removal of Liquids from Gas Wells", J. Pet.
Tech., pp. 1475-1482, Nov. 1969.
SPE 87042 11

MALACCA STRAIT 'A'

AREA MAP- Arun Field Ar ea NSO

FIELD

LNG PLANT

ARUN FIELD

ACEH
Bar isa SLS/PASE
n Mou
0 Km 25
ntai ns FIELDS

ACEH PROVINCE - NORTH SUMATRA

SU

KALIMANTAN
M
AT
R
A

SULAWESI

JAKARTA
300 MILES
JAVA
BALI

Figure 1 Arun is located in Aceh in North Sumatra

12 SPE 87042

OBS-3

OBS-2

OBS-1

GAS INJECTION WELLS

OBSERVATION WELLS

-9000’- -9500’
-9500’- -10000’
-10000’- -10500’
-10500’- -11000’
-11000’- -11500’
-11500’- -12000’
0 2Km s
-12500’- -12000’

Figure 2 Arun Structure Map Showing Cluster Locations

SPE 87042 13

Fluid PVT Diagram

Original Res. Pres. 7115 psi.,

Dew Point - Reservoir Temp. 351 oF
4450 psi

SINGLE Critical SINGLE

Pressure, psi
Point
PHASE PHASE

TWO PHASES

Current Res Pres. 600 - 900 psi

Temperature, oF

Figure 3 Arun Gas PVT Diagram

Arun Production History

3750 250
Sep Gas

, Mbpdbpd
3000 200

thousand
Separator Gas, MMSCFD
Sep Gas, MMscf/D

Water

2250 Condensate 150

/ Water,
Condensate/

1500 100
Condensate

750 50
Water

0 0
1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003

Figure 4 Arun Separator Gas, Condensate and Water Production Rates

14 SPE 87042

Arun Lithology and Permeability

West-East Cross-section

Figure 5 Arun Lithofacies and Permeability Distribution

SPE 87042 15

140

Number of Production Wells

120
Productivity per Well, MMscf/D

100

80

60

40

20

0
1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003

Figure 6 Number of Production Wells and Productivity

Average Field Reservoir Pressure

8,000

Actual
7,000 Simulation

6,000
Average Reservoir Pressure, psi

5,000

4,000

3,000

2,000

1,000

0
1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002

Figure 7 Average Field Reservoir Pressure Decline

16 SPE 87042

C-I-7 TASC Bending Strain

5,000

Casing Failed
4,000
TASC Bending Strain

3,000

First Noticed Restriction

2,000

Casing Installed
1,000

0
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004

Year

Figure 8 Increase in Casing Bending Strains

Water Production Comparison

70

60 Actual
Simulation
Correlation
50

40
% H2O

30

20

10

0
0 500 1000 1500 2000
Pressure, psi

Figure 9 Water Equilibration at Reduced Reservoir Pressures

SPE 87042 17

Figure 10 Well Loading Test

High Gas Rate Pressure & Temperature Profiles

0

2000

Pressure
Measured Depth, ft

4000

Temperature
10 MMscf/D
Annular & Stratified Flow
6000

8000

10000

0 100 200 300 400 500 600 700 800 900

Pressure, psi / Temperature oF

Figure 11 Pressure and Temperature Profiles in Wellbore at 10 MMscf/D

18 SPE 87042

Figure 12 Pressure Profiles in Wellbore at Low Flow Rates

Figure 13 Liquid Holdup and Water Fraction in Wellbore

SPE 87042 19

Impact of Tubing Size on Loading Rate

4

3.5
Minimum Stable Gas Rate, MMscf/D

2.5

1.5

0.5

0
0 1 2 3 4 5 6 7 8 9 10
Tubing Size, inch

Figure 14 . Impact of Tubing Size on Well Loading Rate

C3-2 C4-1
Cluster-3 KH WELL TEST MAP Cluster-4 12
KH WELL TEST MAP
38
2 29
29
20 3
3

15
0

1 40 Well 3
0 5
1
20

38 25
1
7
7
40

21 28
Well 7
65
26 17
15
38

Well 5 27
33 52
Well 2
6 57 74
27 20-40 DarcyFt 34

20-40 DarcyFt 40-60 DarcyFt Well 6

60

40-60 DarcyFt 30 60-80 DarcyFt

107Well 4
60-80 DarcyFt
Well 10

0 500 METERS
0 500 METERS

Figure 15 New Wells Are Located in High Kh Areas

20 SPE 87042

Inflow / Outflow Curves

4-1/2” comp.
curve

Reservoir 7” comp.
Inflow Curve
Reservoir curve
inflow curve
Pressure, psi

9-5/8” comp.
curve

C-III-16C New Drill Well

Kh = 40,183 md. ft
9-5/8” Comp : Initial Rate 27 MMscf/d
7” Comp : Initial Rate 18 MMscf/d
4-1/2” Comp : Initial Rate 7 MMscf/d

Gas Rate, MMscf/D

Figure 16 Nodal Analysis of Sidetrack Options and New Big Bore Well

Cluster II Reservoir & Compressor Performance

Discharge Pressure - 950 psi
250

Old New
Compressor Compressor
200 Combined
Suction Pressure, psi

150

100

50

0
0 100 200 300 400 500 600

Gas Flow Rate, MMscf/D

Figure 17 Cluster II Reservoir and Compressor Performance

SPE 87042 21

Cluster II Compressor Performance Model

Discharge Pressure - 950 psi
250

Cluster II Reservoir
200
Performance
psi(PSIG)

150
Pressure,
Suction Pressure,
Suction

100

(1 New + 1 Old)

50
Actual field measured data

0
260 265 270 275 280 285 290 295 300 305 310

Sep Gas Rate,

Gas Flow Rate,MMscf/D
(MSCFD)

Figure 18 Cluster II Compressor Performance Model

Cluster IV Compressor Performance Model

Discharge Pressure - 950 psi
250

Base + 1 New Drill Base + 7 New Drill

200 Base Well Wells
Suction Pressure, psi

150

2 Compressors
100 (2 New)

3 Compressors
(2 New + 1 old)
50 Matched Points
Actual performance with 2 compressors
Actual Performance with New Well and 3 compressors online

0
420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 580

Gas Flow Rate, MMscf/D

Figure 19 Cluster IV Compressor Performance Model

22 SPE 87042

Gas-Water Contact Movement in Observation Well

GR 2000 CNTC 1987 22000 2000 CNTC 1993 22000 2000 CNTC 1998 22000
0 150 CFTC 1987 CFTC 1993 CFTC 1998
0 11000 0 11000 0 11000
FEET
TVD

GWC @ 10530’ in 1998

GWC @ 10535’ in 1993

GWC @ 10596’ in 1987

Figure 20 Aquifer Movement at Observation Well