Openhole Expandable-Sand-Screen

Completions in Brunei
Hon Chung Lau, SPE, Jacques van Vliet, SPE, Mike Ward, SPE, David Morin, Arifun Djamil, SPE, and Paul Kuhnert,
Brunei Shell Petroleum, and Walter Aldaz, SPE, and Steven Shanks, Weatherford Completions Systems

Summary have a history of sand production, making sand control a necessity.

Sand control in openhole completions transecting multiple sand
Expandable sand screens were deployed successfully as a sand-
layers is particularly challenging because of shale-stability issues.
control device in two horizontal openhole completions in two geo-
This field had a history of unsuccessful openhole gravel packs
logically different offshore fields in Brunei. A critical success
thought to be associated with shale collapse once the openhole was
factor was the choice of a drill-in fluid (DIF) that maintained hole
displaced to brine. Caliper logs showed that with increasing open-
stability and gave a near-gauge hole while running and expanding
hole time, the shale sections tended to be overgauge, while tight
the sand screen. Field experience and extensive research led to the
spots developed in the sand sections. Following several unsuccess-
choice of a synthetic-based DIF for one well but a water-based one
ful openhole gravel packs, extensive laboratory research was con-
for the other. Challenges unique to the deployment of expandable
ducted by the operators technology and application group in
sand screens in these wells are discussed.
Houston to determine if expandable sand screens could be used as
an alternative sand-control method. The following sections discuss
Introduction the major findings of this research.
The expandable sand screen is an innovative technology that is Mud/Shale Interaction Test Results. Samples of shale cored
gaining wide acceptance in the industry as a simple and reliable from several wells were analyzed in the laboratory. X-ray diffrac-
sand-control technique.1,2 The slim nature of the design facilitates tion data showed a complete absence of smectites and an abun-
screen deployment in various openhole applications, including dance of illites, kaolinites, and chlorites in the clay components.
high dogleg severity and horizontal wells. After deployment, the The mud/shale interaction tests conducted included a capillary
screen is expanded to eliminate the annulus, making gravel-
suction test, linear swell meter test, cuttings dispersion, pore-
packing operations unnecessary in reservoirs that carry risks, such
pressure penetration tests, and visual stability observations of shale
as reactive shale, low fracture gradient, fractures, or faults.
in contact with various brines, water-based polymer, and ester-
An expandable sand screen consists of multiple overlapping
rectangular sheets of metal-weave filters attached to an expandable based DIF.
base pipe and encased within a protective metal shroud (see Visual observations at room temperature and pressure showed
Fig. 1). It can be expanded by a solid expansion cone followed by that Ampa shales disperse quickly in KCl and CaCl2 brine, slowly
a compliant expansion tool. During expansion, the base pipe and in water-based DIFs, and not at all in an ester-based DIF. Pore-
the protective shrouds slots open to expose a flow area through pressure-transmission tests conducted on Pierre shale showed that
the metal-weave filters, which accommodate the expansion by both CaCl2 brine and water-based DIFs have almost no resistance
sliding away from each other while maintaining a tight overlap at to pore-pressure penetration. However, an ester-based DIF reduces
all times. The effective filter aperture remains constant and unaf- the effective shale permeability by two orders of magnitudefrom
fected by the expansion process. 33 to 0.37 nDand gives the greatest resistance to pore-pressure
The expandable sand screen has a number of unique advan- transmission. These tests showed that the brine in water-based
tages. First, it offers a large inflow area that minimizes screen DIFs invades the shale, causing pore-pressure equilibration and
plugging and erosion. Second, it is operationally simple to install. leading to shale failure with time. The driving forces for filtrate
Third, it offers a larger internal diameter than most sand-control invasion are hydraulic overbalance and osmotic-driven diffusion.
screens, thus facilitating tubular installation for zonal isolation. On the other hand, an ester-based DIF shows almost no pore-
Fourth, in openhole applications, an expandable sand screen elimi- pressure penetration because of the high capillary entry pressure of
nates the annulus between the screen and the sandface. Therefore, it the oil phase; hence, the shale is stable.
stabilizes the sandface and minimizes sand movement, thus reducing
the risk of sand failure and screen erosion caused by sand production. Borehole-Stability Analysis. The mud weight required for me-
To date, Brunei Shell has completed several cased-hole3 and chanical stability of the shale during drilling was calculated using
openhole oil and gas wells with an expandable sand screen as the the operators proprietary wellbore-stability model. Specific well
sand-control device. This paper discusses the first two applications input, such as true vertical depth (TVD), inclination, pore pressure,
in long openhole completions in two different offshore fields in minimum and maximum horizontal stresses, vertical stress,
Brunei. The first application was in the Champion field, and the Youngs modulus, Poisson ratio, and formation-specific data, such
second was in the southwest Ampa field. These wells are discussed as the friction angle and cohesion extracted from shale samples,
in reverse order here. was used in the calculations. The model showed that a mud weight
equal to the pore pressure plus a trip margin is not enough to
stabilize the hole when drilling weak formations, such as shales at
Expandable-Sand-Screen Application in the high deviation. This is because in a deviated hole, the wellbore
Southwest Ampa Field starts to carry part of the vertical stress. Therefore, in a weak
History. The southwest Ampa field consists of many thin, stacked formation like shales, a higher mud weight is needed to stabilize
sand layers separated by shale stringers. Many wells in this field the wellbore.

Flow-Through Tests. Experiments were conducted to determine

whether an ester-based DIF plugged an expandable sand screen
Copyright 2004 Society of Petroleum Engineers
during run-in and expansion. Results showed that an uncondi-
This paper (SPE 87896) was revised for publication from paper SPE 72131, first presented tioned DIF has a high tendency to plug an expandable sand screen
at the 2001 SPE Asia Pacific Improved Oil Recovery Conference, Kuala Lumpur, 89
October. Original manuscript received for review 11 January 2002. Revised manuscript
whereas a DIF that has gone through an appropriately sized shaker
received 12 December 2003. Paper peer approved 27 January 2004. screen has the least tendency to plug the sand screen.

46 March 2004 SPE Drilling & Completion

 Requirement for remedial zonal isolation for the different sands.
A near-gauge 6-in. hole required through the sand bodies that
are separated by sensitive shales.
A low-skin openhole completion with sand control.
A 958-in. production casing, which is required for 5-in. gas-
lifted tubing with the potential for major hole-cleaning problems.
Research results suggested using an ester-based DIF with a
mud weight of 12 kPa/m to maintain shale stability. Pore-
penetration tests showed that the hole would be stable for an in-
definite period of time with this DIF. The DIF, conditioned
through a 250-mesh or finer shaker, gave a high return permeabil-
ity (in excess of 75%), with a liftoff pressure of less than 10 psi.
Finite-element modeling showed that the expandable sand screen
can easily sustain a pressure drop caused by reservoir depletion
with no risk of collapse. Calculations also showed that under typi-
cal production rates, screen erosion is not a concern. An expand-
able sand screen with a 150-m filter medium was chosen on the
basis of D10 of the formation sands. Remedial zonal isolation of
different sands was achieved with an expandable isolation sleeve
across the intersand shales. An expandable isolation sleeve con-
sists of a conventional expandable-sand-screen joint that has an
elastomer coating bonded to the exterior surface. This converts it
from a sand-control screen to a nonflowing isolation device. In
Fig. 1Expandable-sand-screen features. the future, when one of the sands gases out, a scab liner, an
expandable isolation sleeve, or a solid expandable liner can be run
Return Permeability Tests. Return permeability tests were car-
between the expandable isolation sleeves in the adjoining shales to
ried out with various water- and synthetic-based DIFs on alumi-
isolate that sand. If the bottom sand gases out, an inflatable bridge
num oxide disks with a permeability mimicking that of the forma-
plug can be used. The expandable-sand-screen sandface comple-
tion. In these tests, a DIF filter cake was formed on top of the
tion was designed to allow the use of a fixed cone for primary
aluminum oxide disk at reservoir temperature for a 15-hour period
expansion followed by the compliant expansion tool for secondary
in a high-pressure, high-temperature (HP/HT) cell. The cell was
expansion. Expansion by the compliant expansion tool allows the
then opened, and a shroud and filter medium was pressed on top of
screen to conform to the shape of the formation, thus eliminating
the filter cake. A mineral oil was then flowed into the cell to
the annulus between the screen and the formation.
determine the return permeability. These tests showed that a barite-
weighted ester-based DIF conditioned through an appropriately
sized shaker screen has a liftoff pressure of less than 10 psi and a Results
return permeability greater than 70%.
A total of 543 m of 4-in. expandable screen screen/expandable
isolation sleeve was successfully run and expanded in a 6-in. hori-
Expandable-Sand-Screen Deformation Modeling. An expand-
zontal reservoir section, 1416 May 2001, in Well SWA-290.
able sand screen in contact with the formation is subjected to
Fig. 2 gives a schematic of the completion design.
wellbore stresses. As the pore pressure decreases because of res-
Of the 559 m of 6-in. hole, 254 m was sand, with some large
ervoir depletion, the screen will experience increasing wellbore
shale packages adding up to 305 m of shale. Drilling took longer
stresses and may collapse. Finite-element modeling of expandable-
than planned because the rotary assembly could not be controlled
sand-screen deformation under reservoir depletion was carried out
in the dipping sand/shale beds. However, no hole-stability issues
with specific sand, shale (thick-wall cylinder strength, uniaxial
were encountered, despite 7 days of openhole time. No differential
compressive strength, cohesion, friction angle, and bulk and shear
sticking was seen owing to excellent filter-cake properties.
moduli), and well properties (hole size, total vertical depth, vertical
Figs. 3 and 4 show the as drilled section and plan views,
and horizontal stresses, and mud weight). The results indicated that
respectively, of the wellbore trajectories for Well SWA-290. The
expandable sand screens would not collapse under the expected
close proximity of the target intervals to the drilling platform ne-
load conditions.
cessitated a demanding S-type well path. To achieve the horizontal
targets, continuous build rates of 3.6 to 4.0/30 m were required for
Expandable-Sand-Screen Erosion Tests. Impingement of oil or
the last 1200 m of hole.
gas with entrained sand particles has the potential to erode the
Modeling indicated that excessive doglegs or well-path tortu-
expandable sand screen. The erosion rate is a function of flow
osity would compromise the ability to expand the sand screen.
velocity and particle concentration and is accentuated if an annular
Therefore, to achieve a relatively smooth trajectory, a rotary-
gap exists between the screen and the sandface because of incom-
steerable directional assembly was used to drill the final build
plete expansion or overgauged hole. The annular gap permits axial
section and to land the well.
flow in addition to radial flow. The experiments to predict the
The reservoir sands were penetrated to confirm the geological
erosion rate under various flow rates and sand concentrations indi-
model and the presence of hydrocarbons. A casing size of 958 in.
cated that this application was suitable for expandable sand screens.
was run with the shoe at a 90 inclination. Drilling a 6-in. hole out
Planning for Well SWA-290 of the 958-in. production casing enabled a near-monobore-type
The aforementioned research results were used to design an off- completion with 5-in. tubing and 5-in. expanded inside diameter
shore oil well, SWA-290. An openhole completion was preferred (ID) on the expanded screen. This design will facilitate through-
vs. a cased-hole completion because of the expected lower skin. tubing running of clads or plugs, if required, for remedial zonal
An expandable sand screen was chosen vs. gravel packing because isolation later in the life of the well.
of previous shale-related problems with openhole gravel packing. The objectives while drilling the 6-in. reservoir interval were to
Unique challenges of this well included: Maintain a gauge hole that would enable effective contact of
A demanding belly-shaped well profile with a 600-m hori- the expandable sand screen with the wellbore.
zontal reservoir section. Minimize doglegs and avoid excessive drag when running
A horizontal section consisting of four sand bodies de- the screen.
pleted to 4 kPa/m, resulting in a minimum of 13 000 kPa (1,900 Optimize hole cleaning and leave a cuttings-free interval be-
psi) overbalance. fore running the screens.

March 2004 SPE Drilling & Completion 47

 Fig. 3Well SWA-290 section view.

section, and they processed a maximum of 1.8 m3/min without

losses. Conditioning the mud reduced the mud weight from 12.4 to
12.0 kPa/m.
After conditioning the mud and at regular intervals throughout
the hole section, a simple flow-through test was conducted with
an HP/HT filtrate test with discs of 150-m expandable sand
screen. The flow-through test was considered a Pass when 95% of
the mud in the HP/HT cell was able to flow through screen mesh
with a 30-psi pressure differential. This test was a check on the
mud quality to ensure it could be produced back without plugging
the screen and had the same specifications as those required by the
laboratory return permeability testing carried out beforehand.
While drilling the 6-in. hole section, the mud properties were
modeled and the flow rate restricted to avoid hole erosion caused
by turbulent flow at the bottomhole assembly. Efforts were also made
to minimize stopping and circulating with the bit in the open hole.
After reaching total depth (TD) and before running the expand-
able sand screen, the mud was further conditioned through 325-
Fig. 2Well SWA-290 horizontal openhole expandable sand mesh screens to reduce the particle size in the mud for running and
screen. expanding the screens. A D90 particle-size analysis of 42 to 45 m
was achieved. Using five shale shakers with fine screens, the maxi-
Leave a filter cake that could be produced back without plug- mum circulation rate over the shakers was limited to 0.8 m3/min
ging the screen. without mud losses. This additional mud conditioning further re-
From the test results discussed previously, an ester-based DIF duced the risk of mud plugging the expandable sand screen when
with a target mud weight of 12 kPa/m was selected to drill the flowing back the well after completion.
reservoir interval. Before drilling out the 958-in. casing, the mud On the final trip out before running the screen, the assembly
was weighted with barite to 12.4 kPa/m and conditioned through was backreamed out of the hole to sweep the wellbore clear of
250-mesh screens. The objective was to reduce the particle size in any remaining cuttings bed or possible cavings created from trip-
the mud to a D90 of less than 50 m. In practice, a D90 ranging ping out. A differential caliper log indicated that the openhole was
from 52 to 59 m was achieved. Given these results, smaller in gauge in the sand sections and slightly overgauge in the some
screen sizes were considered. However, the 250-mesh screens shale sections.
were deemed to be the smallest practical screen size, given their Running the expandable sand screen itself went very well, with
conductance and screen durability when handling drill cuttings. average running speed of more than 7 joints/hr. Primary expansion
Five parallel linear-motion shale shakers were used to drill the with a 4.875-ft fixed cone to TD went very smoothly and was

48 March 2004 SPE Drilling & Completion

 Fig. 4Well SWA-290 plan view.

completed in 1.25 hours at an average rate of 7 m/min. It expanded

the screen body to an outside diameter (OD) of approximately 5.8
in. The force required for the fixed-cone expansion was 15 kdaN
(33,700 lbf). No difference in expansion force was seen between
the expandable sand screen and the expandable-isolation-sleeve
joints. The secondary expansion with a compliant expansion tool
also went smoothly and took 6 hours using a 338-in. mud motor at
an average rate of 1.5 m/min. It expanded the screen OD to a
maximum of 6.25 in. Both the cone and the compliant expansion
tool came out fine. The low drag observed indicated that there was
no helical buckling, and the back-calculated casing/drillpipe coef-
ficient of friction was 0.14, compared with the conservative figure
of 0.25 used when planning the job.

Expandable-Sand-Screen Application in
Champion Field
Well CP-306 is 1 of 10 offshore shallow horizontal infill wells in
the Champion field, which is geologically different from the south-
west Ampa field. Unlike Ampa shales, Champion shales are stable
to KCl brine. More than a dozen openhole horizontal gravel packs4
have been successfully installed in this field. The pay zone in these
horizontal wells has been drilled with a water-based polymer DIF
weighted with sized calcium carbonate. Hole stability was not an
issue. Consequently, the same water-based DIF was used for Well
CP-306.
There are several unique features in this well. First, it is rather
shallow, with a TVD of 1100 m at the heel of the well. The shallow
TVD allows only limited weight on bit for expanding the sand
screen. Second, the reservoir section actually goes uphill, with the
heel of the well at 1100 m TVD, and rises at a maximum inclina-
tion of 102 to 1047 m TVD at the toe of the well. Third, because
of the benign nature of the formation shale, a water-based sized
calcium carbonate DIF was used to drill the pay zone. After drill-
ing to TD, the DIF in the entire well was replaced by a solids-free
DIF before running and expanding the screen. Fourth, before the
well was put on production, the DIF filter cake was removed by
acidizing with hydrochloric acid and coiled tubing. Fig. 5 gives a
schematic of the completion design. Figs. 6 and 7 give the section
and plan views, respectively, of the wellbore trajectories.
On 24 April 2001, 344 m of 4-in., 175-m expandable sand
screen was successfully run and expanded in Well CP-306. The
expansion-screen/isolation-sleeve string contained two joints of
expandable isolation sleeve, included to provide a seal inside the
previous casing and to straddle off a gas-bearing zone at the bot-
tom of the completed interval. The screen was run on a string of
312-in., 15.5-lbm/ft drillpipe to the desired depth without prob- Fig. 5Well CP-306 horizontal openhole expandable sand screen.

March 2004 SPE Drilling & Completion 49

 Fig. 6Section view of Well CP-306.

lems. The screen deployment hanger used was a modified 7-in. this well and the others drilled in the same campaign will provide
packer, with sealing elements removed and an ID machined out to excellent comparison material on the productivity of reservoirs
a minimum of 4.936 in. This packer was hydraulically set and completed with expandable sand screens vs. openhole gravel packs.
mechanically disconnected from the running string.
The primary expansion of the expandable-sand-screen/ Production
expandable-isolation-sleeve string was done with a 4.75-in. solid
To date, both wells have been producing sand-free for more than
expansion cone. The running string consisted of 312-in., 15.5-lbm/
30 months with rates equal to or better than expected. Well SWA-
ft drillpipe, 312-in. heavyweight drillpipe, 434-in. drill collar, and
290 is producing at approximately 800 m3/D gross at a water cut
312-in. drillpipe again. The expansion string was based on torque
of 20% with gas lift. Well CP-306 is producing at approximately
and drag calculations to provide the optimum weight expected for
200 m3/D gross at 2% water cut under natural flow.
expansion. To further reduce drag, the bottom section of 312-in.
drillpipe was fitted with two joints of friction-reducing subs. The
water-based polymer DIF in the hole during expansion contained Conclusions
approximately 5% ester lubricant. The following conclusions can be drawn from the experience of
The initial expansion force was approximately 30,000 kDaN to the two wells discussed in this paper:
get the process going. After that, the expansion went smoothly at 1. Openhole expandable-sand-screen technology has been success-
a speed of 10 to 15 m/min and a weight of 15,000 kDaN. After fully applied in two geologically different offshore fields in
approximately 200 m of expansion, the process became much Brunei. In two separate horizontal wells, CP-306 and SWA-290,
more difficult, and a full slackoff weight of approximately 30,000 344 and 543 m, respectively, of 4-in. expandable sand screen
kDaN had to be used. The expansion slowed to 0.3 m/min for the was successfully run and expanded in 6-in. open holes.
last stand. Occasional low-rate circulating (+/ 600 L/min) and 2. In Well SWA-290, the potential for remedial through-tubing
off-bottom rotating (less than 10 rev/min) to break the drillpipe zonal isolation was achieved by running an expandable isolation
friction before proceeding with the expansion helped keep the sleeve across the shales, separating them into four different
progress going. Upon retrieval, the expansion cone was found to sands. In Well CP-306, an expandable isolation sleeve was used
be in excellent condition without any signs of wear. The most at the toe of the well to isolate the water-bearing sand below the
likely causes of the slow expansion process for the last 200 m of toe.
expanable sand screen are 3. Selecting a DIF that gave a near-gauge hole and hole stability
Additional drag because of the heavyweight drillpipe and while drilling, then running and expanding the expandable sand
drill collars getting into the horizontal part of the well. screen was a critical success factor. On the basis of extensive
Buckling and, therefore, locking up of the expansion string. field experience and laboratory testing, a water-based polymer
The secondary expansion took place with a compliant expan- DIF was chosen for Well CP-306, whereas an ester-based DIF
sion tool driven by a 412-in. mud motor. The initial speed was was chosen for Well SWA-290.
approximately 2 min/stand, the circulation rate approximately 4. Laboratory studies useful for candidate selection included shale/
1000 L/min, and the set-down weight approximately 8,000 kDaN. DIF interaction tests, borehole-stability analysis, screen flow-
The circulation rate and weight on the compliant expansion screen through tests, return permeability tests, screen deformation, and
were kept fairly constant throughout the expansion. Expansion erosion tests.
speed varied from an initial 15 m/min to a 3-m/min speed toward 5. Other critical successful factors included integrated prejob plan-
the end of the expandable section. Upon retrieval, the upper set of ning, including all stakeholders, detailed post-job review, well-
expansion balls clearly showed wear and damage caused by the site supervision, and supportive management.
expansion process. 6. Both wells have been producing sand-free for more than 30
The well has been completed and acidized to remove the DIF months at rates equal to or better than expected, with one well
filter cake before being put on production. Production-test data of having a water cut of 20%.

50 March 2004 SPE Drilling & Completion

 References
1. Metcalfe, P. and Whitelaw, C.: The Development of the First Expand-
able Sand Screen, paper OTC 11032 presented at the 1999 Offshore
Technology Conference, Houston, 36 October.
2. Metcalfe, P.D., Urselmann, R., and Saeby, J: The Global Impact of
Expandable Sand Screens on Reservoir Drilling and Completion, pa-
per SPE 67726 presented at the 2001 SPE/IADC Drilling Conference,
Amsterdam, 27 February1 March.
3. Saeby, J. et al.: The Use of Expandable Sand-Control Technology as
a Step Change for Multiple-Zone SMART Well CompletionA Case
Study, paper SPE 68634 presented at the 2001 SPE Asia Pacific Oil
& Gas Conference and Exhibition, Jakarta, 1719 April.
4. van Vliet, J. et al.: Horizontal Openhole Gravel Packs Boost
Oil Production in Brunei, paper SPE 72130 presented at the 2001
SPE Asia Pacific Improved Oil Recovery Conference, Kuala Lumpur,
89 October.

SI Metric Conversion Factors

ft 3.048* E01 m
in. 2.54* E+00 cm
lbf 4.448 222 E+00 N
psi 6.894 757 E+00 kPa
* Conversion factor is exact.

Hon Chung Lau is a senior staff production engineer at Shell

E&P Solutions in Houston and the Coordinator of Production
Technology in the Integrated Reservoir Modeling Group. He
has held various research positions at Shells E&P laboratories in
Houston and The Netherlands and was a production technolo-
gist and sand-control focal point in Shell Brunei. Lau holds a
PhD degree in chemical engineering from Princeton U. He has
served as Technical Editor for SPE Drilling & Completion.
Jacques van Vliet is a senior production technologist in Al Furat
Petroleum Co., Syria. He joined Shell in 1990 and has worked as
drilling fluids and cement engineer in The Netherlands and Bru-
nei and as production technologist in Brunei and Syria. Van
Vliet holds a PhD degree in physical chemistry from Utrecht U.,
The Netherlands. Mike Ward is a rig superintendent for Brunei
Shell and has worked around the world as a well engineer for
Shell. Ward holds a BS degree (with honors) in physics from the
U. of Exeter, U.K. David Morin worked as well planning and
design engineer for Brunei Shells Technical Limit Drilling Initia-
tive. He had previous drilling, production, and reservoir engi-
neering experience with Canadian energy producers. Moran
holds a BS degree in mechanical engineering from the U. of
Calgary. Arifun Djamil is Drilling Superintendent of Unocal,
based in Sugar Land, Texas. He has worked for Unocal Corp.
and Shell in The Netherlands and Brunei. Djamil holds a BS de-
gree in chemical engineering from the Inst. of Technology
Bandung, Indonesia. No information is available for Paul Kuh-
nert. Walter Aldaz is the Product Line Manager of Expandable
Fig. 7Plan view of Well CP-306. Sand Screens for Weatherford, northern Latin America. He has
been working worldwide in the oil industry for the last 16 years.
Before his recent relocation to Mexico, he was the Expand-
Acknowledgments able-Product-Line Region Manager and south Asia District
Manager for the completion and production systems division
The authors want to thank Brunei Shell, Weatherford, and the of Weatherford.. Aldaz holds a BS degree in mechanical en-
Petroleum Unit for permission to publish this paper. The valuable gineering from the U. of New Orleans. Steven Shanks is South
research on shale stability and DIFs conducted by Shell E&P Asia Operation Manager for Weatherford, based in Kuala Lum-
Technology Application and Research laboratory in Houston and pur. Before his current assignment, he was expandable sand
by Baroid in Aberdeen is gratefully acknowledged. screen focal point for Weatherford in Brunei.

March 2004 SPE Drilling & Completion 51