Sie sind auf Seite 1von 10

SPE/IADC 97282

Expandable Sand Screens Selection, Performance, and Reliability: A Review of the


First 340 Installations
C. Jones, M. Tollefsen, P. Metcalfe, J. Cameron, D. Hillis, and Q. Morgan, Weatherford Intl.

Copyright 2005, SPE/IADC Middle East Drilling Technology Conference & Exhibition
This paper was prepared for presentation at the SPE/IADC Middle East Drilling Technology
Conference & Exhibition held in Dubai, U.A.E., 1214 September 2005.
This paper was selected for presentation by an SPE/IADC 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 or the
International Association of Drilling Contractors and are subject to correction by the author(s).
The material, as presented, does not necessarily reflect any position of the SPE, IADC, their
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers or
the International Association of Drilling Contractors 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
Expandable sand screens (ESS.) are a relatively new sand
control system, which combines many of the properties of
gravel packs with the ease of installation of a stand-alone
screen. Although they have been used in a wide variety of
applications, they are not considered a panacea and have an
operational envelope, which is becoming clearer with time.
Weatherfords ESS system currently (June 2005) has 340
installations and over 700 years of combined production. A
recent survey of the installations was analyzed in terms of
performance and reliability.
The productivity performance of the ESS has been shown
to be very good, with an average skin of 0.3 being achieved in
recent openhole applications. ESS completions generally
perform better than the baseline models. Where field
comparisons were possible, they also performed better than
alternative sand control completions.
Over the 340 ESS wells, ESS has a reliability comparable
with other sand control systems, with initial failures less than
5% and a production failure rate of 0.021 failures/well.year.
This gives a projected survival rate at 20 years of greater than
90%. This rate is expected to get better with improving
operations, designs, systems and application selection.
Introduction
When the first ESS was launched in 1999 it was a radical
departure from convention, it introduced the concept of direct
screen contact with the formation as a means of increasing
productivity, sand control and reliability.
ESS was designed and aimed specifically at openhole
applications, despite the subsequent use of it in cased-hole
applications. Features such as large exposed filter area and
variable ESS borehole contact (becoming known as compliant
.

Registered Trademark of Weatherford in the UK

expansion) were included to provide gravel pack functionality


with the operational simplicity of a stand-alone screen. ESS
has been used to replace openhole gravel packs (OHGP),
cased hole gravel packs (CHGP), cased hole frac and packs
(CHFP) and standalone screens (SAS).
The design premise was that a compliantly expanded filter
that could eliminate as much of the annular gap as was
practicable would also promote rapid formation stabilization
and minimize the movement of sand particles around the
screen during initial transient sand production period. The
large directly exposed filter surface was designed to minimize
the pressure drops in the screen and sand pack composite
caused by mud particles and formation fines. Both features
were aimed at improving productivity and reliability of the
sand-face completion1. The role of the reduction in the
annular gap in increasing completion reliability has been
discussed in Helland et al2.
With 340 ESS installations and 65 km now installed, the
statistics are becoming significant and it is now possible with
confidence to determine whether ESS does indeed offer the
improvements claimed by the designers and to formulate
selection criteria for their use. This paper presents data on
performance of ESS completions together with data on longterm reliability and failure rates. The ESS application
selection process is used to show how failure rates can be
reduced and long-term reliability improved.
ESS Performance
There have been a large number of case studies published
which have looked at ESS performance, in a wide variety of
well types, vertical, horizontal, gas, gas-condensate, water and
oil wells. The current maximum fluid production rate is
25,000 bpd, the maximum gas rate is 290 MMscf/d, and the
maximum water injection rate is 40,000 bwpd. Some specific
published examples are given below.
Weekse et al3 documented the installation and performance
of three long horizontal gas wells in the Brigantine field, in the
Southern North Sea. The wells had up to 40% improved
production over expectation. They were completed 32 days
ahead of schedule with a saving of $13.5M. These wells have
been producing for over 4 years.
The performance of the ESS completed wells in the gas
condensate Scoter Field was also very good with no evidence
of mechanical skin or formation damage4. These wells have
been producing for over two years.
The Xijang Field in China is a mature field with high water
cut. The ESS was fitted to a number of wells including a

SPE/IADC 97282

Customer Supplied Openhole Skin Data

40

# Cone OH Skin
# Compliant OH Skin

35

# of skins

30
25
20
15
10
5
>9

>8

>7

>6

>5

>4

>3

>2

>1

>0

>-1

0
Skin Value (Range)

Figure 1 Compilation of Openhole Skin Values for Compliant and


Non-compliant ESS

There is as interesting split between applications expanded


with a fixed cone, which will not be fully compliant and
applications expanded with a compliant expander, which will
be generally fully compliant. The non-compliantly expanded
applications have an average skin of 2.3. The fully compliant
applications have and average skin of 0.3. This shows the
performance benefits of compliant expansion. The reason for
this may be due to the compliantly expanded ESS stabilizing
the formation and limiting formation movement and yield.
This could theoretically lead to a slightly lower skin18.

Filtercake cleanup and mixing with failed formation material


will also have a large role.
Figure 2 shows the openhole skin values broken down for
vertical (<15), deviated (>15 & <85) and horizontal (>85)
wells. The data shows no negative bias towards the deviated
and horizontal wells, which suggests that the skins are in fact
true formation/completion skins.
Openhole Customer Skin Data Analysis

15.00

40

Data Wells
All OH Skins
10.00

30

5.00

20

0.00

10

-5.00

# Data Wells

Skin Value

Cone
Horizontal

Compliant
Horizontal

All Horizontal
(>85)

Cone Deviated

Compliant
Deviated

Cone Straight

All OH Deviated
(<85)

Compliant
Straight

All OH
straight(<15)

multilateral5. Each lateral produced two to three times more


than the alternative sand control completions, which were frac
and pack and gravel packs.
Due to its small running outside diameter and large final
inside diameter (ID) the ESS is particularly applicable to
situations where reductions in final ID must be avoided, such
as in workovers and sidetracks. Oluwatosin et al6 show that
ESS completed sidetracks produced between 60-99% more
than the predicted rates.
There have been many others, which have noted increased
production and a reduction in costs7,8,9,10,11. ESS has also
been used as replacements for single and multi-zone
CHGP12,13,14. There have also been theoretical papers, which
have shown the benefits of the large ID of the ESS in
improving sweep efficiency and the recovery factor of a
reservoir15.
All the references cited above have looked at the
performance of the ESS completed wells shortly after
installation. Typically a well test will be performed just after
completion to determine the wells performance and to check if
any further clean up or stimulation is needed for the well to
achieve its potential. A number of operators have shared the
skin values determined from the well tests. Data is available
on 98 wells; some of this data has already been published16.
Figure 1 shows an updated compilation of the openhole skin
values. Although we have no control on how the skins are
calculated, we have been assured that they are
formation/completion skin values as defined by Van
Everdingen and Hurst17, the effect of deviation and horizontal
wells has been removed. The data show that for the openhole
wells there is an average skin of 0.9. This is a very low value
and compares favourably with other openhole sand control
completions.

Figure 2 Openhole skins for vertical, deviated and horizontal wells

The oldest ESS completed wells have now been producing


for close to six years. Case studies are now being published
looking at long-term performance (Mason et al19). This case
study analysed the long-term performance of an installation in
Q4 2000, which has now more than four years of production
history. The ESS was compared to a SAS and an OHGP in
heterogeneous unconsolidated sands. The ESS had much
higher productivity and lower skin. Also the ESS appeared to
maintain a low skin over the study period, whereas the SAS
became rapidly impaired. The OHGP also declined but to a
lesser extent. Normally in a sand prone environment electrical
submersible pumps (ESP) need changing regularly, but the
well with ESS has had the same ESP for more than four years.
This is a strong testament to the long-term sand integrity of the
ESS.
To summarise the production performance of the ESS: the
published data shows that the ESS if correctly applied gives
high productivity or low skin wells, which maintain their
productivity over time. Where equitable comparisons are
made, the ESS is better and often much better than the
alternative sand control completion option.
Production performance is not the only selection criteria
used in choosing the type of sand control to fit to a well.
Long-term reliability is also very important.

SPE/IADC 97282

ESS Reliability and Failure Rates


An understanding of reliability and failure rates is central to
the design of any system. There has been a recent survey by
King et al20 published in 2003, which studied the reliability of
a wide range of sand control completions; SAS, OHGP,
CHGP, CHFP and ESS. The failures were split into four
categories: design failures, such as errors in sizing; application
failures, which in the context of ESS relates to mechanical
damage; infant failures defined as failures within 30 days of
the start of production and production failures, which were
failures after 30 days of production. The first three failure
categories can be classed as initial failures. The data from the
paper is shown in Table 1.
Type
of
Comp
letion
SAS
OHGP
ESS
CHGP
CHFP

No.of
Wells

Well.
Years

Design
failures
(%)

183
175
194
369
844

783
507
255
1514
3369

0.6
0
1
0
1.69

Applica
tion
Failures
(%)
0
9.7
3
2.2
2.4

Infant
failures
(%)

Production
Failures/
well/yr

0.6
0.57
1
0.8
0.24

0.056
0.020
0.016
0.011
0.004

Table 1 Reliability Summary from SPE 84262

The data shows that ESS has similar reliability to OHGP,


and CHGP both of which it typically replaces. Since the
publication of the King paper, there have been many more
ESS applications by Weatherford. Weatherford maintains a
database of all ESS installations in which is captured a wide
range of data including failure information. The database is
frequently updated by contacting operators to determine how
the wells are performing. A survey was performed recently by
sending out simple questionnaires on each installation. This
resulted in an update on > 90% of the wells. Using this
information we are able to provide up to date statistics on
performance, reliability and failures. The analysis in this
paper is for installations done up to the end of June 2005.
Figure 3 shows the success rates within 30 days of
installation. Of the 340 installations, 266 or 78% were
straightforward installations, 55 or 16% had NPT (downtime)
or required some remedial actions. Eighteen wells (5.6%) were
either recompleted, sidetracked or abandoned and one well
was completely lost. The last two categories would be failures
in the King publication.
Installation Categories

Loss of Well 1

Combined Initial Failure Rates


Statistics on failure rates as of end June 2005 are shown in
Table 2. They show that there have been 340 installations
producing for a total of 727 years. There have been nineteen
wells or 5.6%, which were failures during installation or
shortly after the start of production. Of these five were initial
design failures, due to the weave selection being based on
poor data, of these one may have had some additional screen
damage. Two of these design failures have been discussed in
a recent SPE paper19. One of the others resulted in the only
installation that to our knowledge resulted in a complete loss
of well.
Nine installations were classed as application failures, with
eight due to parted connections and the other due to screen
damage passing through a poorly dressed window. The
connectors have since been redesigned and have experienced a
100% success rate so far.
There have been two wells, which could be identified as
infant failures. One a high rate gas well which sanded up
immediately on initial startup, the other a low rate oil well,
which also sanded up. In the gas well, there was a large
washout under the casing shoe over which the ESS was run. It
is likely that there would have been a large concentration of
flow at this point, which could have lead to rapid erosion.
This failure could also be considered as a design failure. The
cause of the other infant failure may have been damage at the
top of the screen but this has never been clearly resolved. In
both of the infant failure cases, running blank pipe over the top
sections could have prevented the failure.
Of the nineteen failed wells, three others fell outside these
categories. One oil well produced 100% water. The two
others produced sand because the packers were set below the
top of the perforations. These applications were in no way an
ESS failure. There are two other recent problem wells which
are still under investigation but which are also not ESS
failures. So in reality there are sixteen ESS applications,
which failed at or shortly after installation, this is a combined
initial failure rate of 4.7%.
Number of installations
Number of well years
Design Failures
Application Failures
Infant Failures
Production Failures

340 wells
727 well.years
5 wells
9 wells
2 wells
15 wells
0.021 failures/well.year

Table 2 Failure Statistics as of June 2005


Sidetrack or
Recomplete

18

Remedial Action

19

Unplanned
Downtime

36

Operationally
straightforward

266
0

100

200

Figure 3 ESS Installation Statistics

300

Figure 4 shows the combined design + application + infant


failure rate as a function of time. Between 2000 and 2002,
the failure rate increased until in 2002 there was a
disasterously high failure rate of 10% or 8 out of 80
installations. The reason for this increase in failure rate was
twofold. As the technology became more accepted, it was
applied to ever more demanding applications, with the screen
production rate having to increase dramatically to keep up
with demand. These two conditions combined to give the high

SPE/IADC 97282

failure rate in 2002. The most common failure mode in 2002


was parted connectors, mostly in the 5 size.
As a result of this the 5 ESS was withdrawn from the
market.
Investigation showed that quality issues had
combined with a relatively weak connector design (Mark I) to
cause failure under certain extreme conditions. The connector
was redesigned to make it much more robust and underwent
extensive qualification testing (Mark II) and since this
redesign there have been no Mark II connector failures.
Manufacturing quality control and inspection were also
modified and improved.
These measures have dramatically reduced the combined
initial rate to less than 2% average over 2003 2005. This
corresponds to three wells, one of which sanded up with sand
that was all less than the aperture of the weave. In this case
the weave sizing had been based on inappropriate data. Also
there was one well in which the old 5- ESS with the
relatively weak connector was run (against Weatherfords
recommendation). This failed due to a parted connector. In
the third, the ESS was installed incorrectly and did not fully
cover the perforations.
12.0%

Design+application+infant

10.0%

8.0%

6.0%

4.0%

2.0%

0.0%
1999

2001

2003

2005

Y ear

Figure 4 Combined Design+Installation+Infant Failure Rate with


Time

The improving reliability of ESS as shown in Figure 4 is


also due to improved candidate selection and planning. This
process is discussed later in the paper.

Production Pe rform ance

6%

Initial Failure
Producing at or above
expectation

73%

2%

Not Connected

5%

Depleted

9%

Producing non-optimally

6%

Well shut-in (some issue)


0%

20%

40%

60%

80%

100%

Figure 5 Status of All Installations

Production Failure Rates


Figure 5 shows the current status of all 340 ESS
installations. As described in the previous section around
5.6% of the wells were not produced, of which 4.7% could be
considered to have failed. 73% are producing at or above
expectation. New wells take some time to be connected up to
pipelines so 2% are not currently connected. Reservoirs
deplete or water out, sometimes unexpectedly quickly so 5%
are shut in due to depletion or high water cut. 9% are
producing non-optimally, either poor productivity, restricted
access or high fines.
Twenty-one of the wells (6.2%) are shut in due to some
issue. The primary reasons as far as it can be determined for
the shut-ins are shown in Figure 6. There may have been
multiple problems in a well but only the primary cause is
noted.
The most common is sand production with grains greater
than the weave aperture. Nine wells are shut-in for this
reason. Three wells were shut-in due to restrictions at a
collapsed section of the ESS. Three wells were considered
erosive failures; these were cased hole applications.
Two other wells also produced sand, but were not thought
to be ESS failures. One of these wells had a failed frac-pack,
and in the other, the sand is thought to have been produced
from damage in the completion above the ESS.
Two more wells were shut-in due to poor productivity.
Both of these were cased hole applications and the poor
productivity was thought to be due to perforating issues.
Another two wells were shut-in for non-ESS related
issues; one due to a failed subsurface safety valve and the
other due to an underground blowout in a nearby well.
So of the original 21 shutins, 15 are directly due to the ESS
and 2 others involved sand produced from nonESS
component of the well. Whether the last 2 failures should be
considered as ESS failures or system failures is arguable.
These 15-17 failures equate to a value of 0.021-0.023
failures/well.year depending on whether the non-ESS sand
production is included. This value is slightly higher than the
value quoted in the King paper, but very similar to the OHGP
figure.

SPE/IADC 97282

Hydraulic Collapse
Formation Collapse

3
2

Low Productivity other


0

Low Productivity ESS

High Sand non ESS

High Sand Erosion

High Sand > w eave


2

No ESS Issue
0

10

Figure 6 Reasons for Well Shut-ins

One other well is under investigation, for a non-ESS


problem, this has not been included in the analysis.
With the benefit of hindsight, improved design and
applications screening could have prevented most of these
failures.
Most of the wells with high sand production employed
filtration weave sizes that were based on non-fully
representative sand sieve data. This is a common problem
because accurate data on the sand encountered in a well is
only really available after the well has been drilled. High sand
production through the weave could have caused erosion and a
complete loss of sand control. High solids production could
also stop the formation of a stable arch around the wellbore.
This could cause transfers of large loads to the ESS, causing it
to deform, sometimes severely.
Staying below critical velocities by changing production
rates, hole angle or perforating strategies could easily have
prevented the three erosive failures.
With recent geomechanical experiments results and
predictive software, weak and collapsing formations can be
effectively predicted and the offending sections isolated
behind blank pipe.
The design and application screening process is discussed
later.
Figure 6 also lends credence to one of the principal
benefits of an expandable sand screen, namely the elimination
of plugging or loss of productivity tendencies associated with
conventional sand control. It appears that eliminating the
annulus and stabilizing the formation reduces the tendency of
the formation to produce fines, and even when fines are
produced, they flow through the metal weave and hence do not
plug the screen. It is perhaps too early to claim this as a
proven benefit of compliance, but the indications are good.
Long Term Reliability and Failure Rates
Gravel packs and screen only completions have long case
histories, which allows an evaluation of the long-term
reliability of these completion options. ESS has less of a track
record due to a relatively short six year run history. The
failure rate of the ESS is 0.021 per well.year, this is very
similar to the rates reported in the King et al paper for
OHGP17.

One of the parameters that must be known to have a


reasonably accurate field development plan is the probability
of a well extending the design life of the field. This can be
estimated by plotting the failure rate over time. Sand control
completions tend to have a relatively rapid failure rate to begin
with then the rate falls off with time. Although there is sparse
data, this can be seen in the failure rate of the ESS wells.
There is more in the first year of production than in the second
year for example.
Figure 7 shows the fraction of the wells still producing as a
function of time on a semi-log scale. The plot was generated
using the failure times for each of the failed wells. The graph
shows that the data is falling on a straight line, which suggest
an exponential decay in failure rates.
The data could be used to project the survival rate of the
wells out to say 20 years. This predicts the survival rate after
20 years is still greater than 90%.

1.05

1
Fraction Still Producing

Re as ons for Shutins

0.95

0.9

0.85

0.8
0.10

1.00
time (Years)

Figure 7 Times to Shut-in for the Failed Wells

10.00

SPE/IADC 97282

ESS Application Qualification Process.


Absolutely fundamental to a reliable completion performance
is the application qualification process. This process is
enshrined in Weatherfords Operations Project Management
System (OPMS) software. OPMS is used to manage ESS
installations from the first meetings with clients to final
invoicing. OPMS has gates which cannot be passed unless a
component of the qualification process has been satisfactorily
completed.
In the context of ensuring fit-for-purpose design the OPMS
addresses eight key aspects, which are listed below. The
important point is to be aware of these factors so they can be
allowed for at the early design stage.
1.
2.
3.
4.
5.
6.
7.
8.

Hole quality
Well trajectory for deployment and expansion
Weave selection
Mud selection
Borehole/ESS stability
Metallurgy
Erosion
Final clean up and bean up

Long horizontal sections at the end of extended reach wells


can make deployment and expansion difficult. Rejigging the
bottom hole assembly and adding drill collars at key locations
can often resolve difficulties. Even if the T+D modeling
shows the ESS can be successfully installed, there are
sometimes problems during the actual expansion. These can
usually be resolved by altering the expansion string, or by
using friction reducing agents.
Filter Media Selection Process
Extensive work has been performed on the performance of the
filtering characteristics of the Dutch twill weave filtration
media used in the ESS21. Lab tests measuring retention and
pressure build-up on ESS weaves with various reservoir sands
have been in progress since January 2002. The results of these
tests indicate that there is a good correlation between the
largest particles in a distribution and the retention performance
of the weave, i.e. the size of the d10, d5 or d1 of the sand.
Figure 8 shows the correlations between sand passing through
and the size of sand in terms of d10 for the 150 micron weave.

4.0

Well Trajectory for Deployment and Expansion


In a given application it is essential the ESS can be safely
deployed and expanded. This process is a function of the well
trajectory and the weight of the string.
Prior to every application detailed torque and drag (T+D)
modeling is done to ensure that the ESS can be deployed and
expanded safely and without damage. This is usually a
straightforward process, but there are circumstances that can
prove difficult.

3.5
sand passed through (g)

Hole Quality
Hole quality is important in all types of sand control
completions. It includes several aspects such as how in-gauge,
and free from tight spots is the hole. Also cleanliness, with the
removal of cuttings beds and thick filter cake recommended.
The production and sand control performance of the ESS
tends to be higher in a gauge hole where compliant expansion
is possible or where a fixed cone can be sized to minimize the
annulus18. Drilling a gauge hole for a given application is
down to a combination of proper BHA equipment selection
and configuration, Drill In Fluid [DIF] selection and adoption
of good drilling practices. Once the hole has been drilled then
running a caliper, either LWD or wireline, can quantify hole
quality.
Hole cleanliness can be managed with proper attention to
detail with the drilling fluid to ensure proper cuttings removal.
Also prior to running the ESS the drilling fluid must be
conditioned at high circulating rates to allow it to pass the
filter weave and to leave the reservoir section with a thin filter
cake.
Final hole quality can be checked by a slide trip from the
previous shoe to TD. This will also give an up to date
measure of the friction factors for input into the torque and
drag simulations which are updated immediately prior to
running in hole and real time to verify a problem free ESS
deployment and expansion process as below.

3.0
2.5
2.0
1.5
1.0
0.5
0.0
0

50

100

150

200

250

300

350

400

450

500

550

600

d10 of sand

Figure 8, 150 micron weave: D10 correlation

Where the d10 of the sand is greater than the weave aperture
(150 micron in this case) very little sand passes through the
weave. As the sand gets smaller with d10s less than the weave
aperture the amount of sand passing through the weave
increases gradually at first then more dramatically and there
comes a point where the sand is too small to form a stable
filtercake on the screen. Defining the cut off point for
acceptable solids production is difficult, but Hodge et al22
proposed a limit of 0.12 lbs/ft2, which correlates to 0.4g sand
in these plots. Using 0.4g as the limit for sand production, it
can be seen from the graph that a 150 micron weave would
give adequate sand retention for sands with d10s as low as 130
micron. The plot of the d5 of the sands against sand passing
through the 150 micron weave shows that a d5 of 150 micron
is just within the maximum area of retention (Figure 9).

SPE/IADC 97282

3
2.5
2
1.5
1
0.5
0
0

50

100

150

200

250

300

350

400

450

500

550

600

d5 of sand

Figure 9, 150 micron weave: D5 correlation

These experimental results indicate the boundaries of


retention, knowing these is essential for weave sizing. The
weave is sized on the finest sand likely to fail rather than
average sand. Knowing that good retention is achieved when
5% of the sand is larger than the weave aperture means that
the coarsest weave able to retain the sand can be selected
enabling easier mud flow through when the screen is run in
mud. There are instances where more caution in weave sizing
is advisable: for instance when particle size data is limited and
if the sands have very high fines contents, or are bimodal.
Enough retention tests have been performed to be able
compare the distributions of potential applications to the sands
already tested, any that are unsual (for instance bi-modal) are
tested. Although sizing according to the d5 of the sand will
give the maximum retention, there can be some flexibility in
sizing which is helpful for reservoirs with wide variations in
sand sizes.
Mud Selection
One common concern with the ESS in open hole is the
potential for the mud filtercake to plug the screen when
production is first started. With the ESS fully expanded against
the borehole wall, the mud filtercake will have to be produced
directly through the weave of the screen. Laboratory tests and
field experience suggest that filtercake flow back is not a
common problem, and Weatherford generally advocates
designing the mud in the normal way to ensure hole stability
and minimal formation damage prior to performing filtercake
flow back tests on the recommended ESS weave.
The filtercake lift-off testing takes the form of a core flood,
which is done at reservoir temperature and the desired mud
overbalance pressure. The mud is placed dynamically, which
gives a more realistic filter cake and filtrate leak-off, and
permeabilities are measured to oil with the core conditioned to
residual brine. A shroud section and weave disc is pressed into
the mud cake and the permeability to oil is then re-measured
in the production direction. These tests are performed on an
outcrop sandstone since they are designed to identify problems
with the screen rather than formation damage within the rock.
A low return permeability will require a baseline test without
the screen to check that damage to the core is not responsible.
In addition, if the ESS is to be run in the drill-in fluid, mud
conditioning will be required to ensure the mud will pass
through the screen without plugging, and laboratory tests may
be necessary to identify the level of solids control required. In

Borehole/ESS Stability
New operators are sometimes concerned that the relatively
modest ESS strength will cause problems with hole collapse.
A simple geomechanical model was developed to better
understand the processes leading to excessive deformation and
to be used as a screening tool.
The model is based on concepts developed for
tunnelling23,24 and tunnel support. It calculates the depth of a
failed or yielded zone around a wellbore as a function of
wellbore support from a mud overbalance or an expandable
sand screen. The yielded zone grows as the mud support is
removed and the well is drawndown and depleted. Volume
changes in the yielded zone compress the ESS.
The EWBS model predictions have been calibrated against
experimental and field measurements of ESS deformation.
Figure 10 shows the input/output sheet for the model. The
model uses various inputs, depth, formations stresses,
reservoir pressure, well trajectory, rock properties, and
depletion/drawdown. The model calculates the maximum ESS
deformation during field life. A detailed description of the
model is available.18

Well Details
Client
Well
Field
Region
Comments:

6.0

2.5

5.5
2.0

Well Properties
Depth
Overburden
Max Horizontal Stress
Min Horizontal Stress
Initial Pore Pressure
Final Pore Pressure
Azimuth (SHmax)
Inclination
Hole size
ESS OD

7214
5100
3800
3800
3440
2000
0
90
6
6.00

ft
psi
psi
psi
psi
psi
degs
degs
inch
inch

5.0
1.5

Skin

sand passed through (g)

3.5

these tests the drilling mud is flowed through the ESS weave
at a constant rate and the pressure drop across the weave is
monitored with time. If 500 mls of mud can pass the weave
without significant rise in pressure the test is deemed
successful. The drill in fluid is conditioned through
increasingly fine sieves until the test is successful, and the
mesh rating required is recommended for use in the rig
shakers. Before the screens are run in the field rigsite mud
tests should be performed to ensure the mud is properly
conditioned.
ESS has been successfully run with both water and oil
based muds in open hole. The only problems have occurred
with sized salt systems, the particle size of which can be
difficult to control. As a result Weatherford do not generally
recommend the use of sized salt systems with ESS.

ESS ID (inches)

4.5

1.0
4.0

Formation Properties
Unconfined Compressive Strength
Triaxial Stress Factor
Friction angle
Cohesion
Young's modulus
Poisson's ratio
Expansion
Yielded Material Cohesion
Initial Reservoir Permeability
Yielded Reservoir Permeability
Mud Calculations
Over Balance
Wellbore Pressure

200
2.15
21.4
68.2
493000
0.43
0.1
15
3000
1500

psi
mD
mD

250
3690

psi
psi

psi
psi/psi
degs
psi
psi

0.5
3.5

3.0
0

500

1000

1500
2000
2500
Wellbore Pressure (psi)

250psi Overbalanced
20% Deformation Limit
ESS ID

3000

3500

Reservoir Pressure
Expected Final Reservoir Pressure
Yield Zone Diameter

Figure 10 Input/Output sheet of the Geomechanical Model

0.0
4000

Metallurgy
The selection of materials for the constituent components of
ESS encompasses an evaluation of four main attributes;
suitability for expansion without tearing at achievable loads,
mechanical strength, fabricability and resistance to corrosion
& environmental cracking.
A material with a high strain-hardening exponent will
promote uniform plastic deformation to high strain levels and
will be suitable for slotted expandable applications. Fracture
toughness is important to prevent tearing at stress
concentration features, i.e. the slot ends. Solution annealed
UNS S31603 (316L) austenitic stainless steel is utilised for the
standard ESS basepipe, weave and shroud and offers high
ductility, strain-hardening rate and fracture toughness. Other
higher specification materials such as UNS S32760 (25 Cr),
and UNS N08825 are also suitable from an expansion
perspective.
A secondary aspect of suitability for expansion influenced
by material characteristics is the force necessary to generate
expansion. To enhance the strength of the system, it is
possible to increase the basic cross-sectional area, the system
design or the yield strength of the material. However, as a
consequence of improvements in mechanical performance,
e.g. tensile load bearing capacity or collapse resistance, there
is a concomitant increase in expansion force requirements.
This must be assessed against expansion tool loading
constraints for the length of a specific installation or the
generic tool design life.
The two main mechanical service requirements of the ESS
system are 1) the capacity to support both the string weight
and the axial force component of expansion and 2) to resist
collapse from formation stresses. The weight of the string is
carried by the basepipe through the connections. The
connection, being more restricted on cross-section by design,
has a higher strength requirement than the basepipe.
Therefore, solution annealed UNS S32760 or S32750 super
duplex stainless steels, possessing a specified minimum yield
strength more than double that of the austenitic basepipe, has
been adopted in the standard ESS product. This material
maintains good ductility and high fracture toughness at this
strength level.
The original selection of stainless alloys for standard ESS
stemmed from a requirement to provide resistance to
moderately-corrosive CO2-bearing reservoirs. Use of a 316L
filtration medium is common in traditional screen
construction.
An assessment of well corrosivity and a specific material
recommendation is typically conducted for all ESS
applications. For production wells the fluid, temperature and
pressure data is firstly assessed to determine the partial
pressures of acid gases and estimate the potential in-situ pH of
produced waters. The in-situ pH, pH2S, chloride content and
temperature levels are then assessed to evaluate the risk of
localized corrosion or environmental cracking. Particular
attention is paid to the weave material, given the nature of the
critical dimensions in a filter component, and stricter selection
criteria applied.
In cases where well conditions are deemed to present too
great a risk of weave corrosion failure, an upgrade from
UNS S31603 to UNS N08825 is recommended to enhance the

SPE/IADC 97282

resistance to corrosion and cracking in more aggressive


conditions.
Life cycle fluid exposure also plays a role in materials
selection for ESS. Brine and acidizing fluid exposure tests
have been performed to evaluate the effectiveness of inhibiting
agents and the overall resistance of materials utilized.
In one case, sour and corrosive production conditions
combined with a severe acidising program necessitated the
development and installation of a nickel alloy ESS system.
The filtration medium was first upgraded to UNS N08825. To
maintain expansion characteristics, UNS N08825 was also
used for the basepipe and shroud materials. This grade
exhibits mechanical behaviour similar to UNS S31603. The
equipment was successfully installed with no significant
differences from a standard installation.
In water injection applications, whether seawater,
produced or commingled waters, a risk of corrosion exists.
This is largely dependent on the control of oxygen levels by
stripping and chemical additions. Other factors include
exposure temperature, chloride content, CO2 in gas-stripped
systems, residual chlorine levels and microbiological species.
The quality of injected fluid is therefore controlled by the
operator and ESS material selection is based upon reported
fluid control capabilities and control measures.
In general terms, the minimum metallurgy recommended
for permanent completion jewellery in water injection service
is super duplex stainless steel, particularly where seal surfaces
are concerned. In applications with a high level of control,
e.g. maintaining <20ppb residual oxygen, there is a potential
for use of standard ESS materials. However, given the nature
of the ESS construction and the abundance of natural crevices,
an upgrade of weave material to UNS N08825 or
superaustenitic material is typically recommended for such
applications.
In more critical or severe water injection applications,
other materials have been utilized to enhance resistance to
localized corrosion. In one instance, the weave material was
upgraded to UNS N06625, with the remainder of the
components subject to PREn (Pitting Resistance Equivalent
number) based selection. This allowed the use of standard
connection material subject to a higher minimum PREn
requirement. To maintain the suitability of basepipe and
shroud materials for expansion, UNS S31254 and
UNS N08367 superaustenitic stainless steels were selected to
provide enhanced corrosion resistance whilst preserving
mechanical performance.
The selection of appropriate materials for mechanical and
environmental application requirements is an important
contributory factor in the prevention of ESS failures. The
selection philosophy adopted by Weatherford has been
demonstrated by the high success rates enjoyed by the ESS
product line. To date, no failures relating to corrosion or
mechanical performance within the design envelope have been
attributed to the materials of construction.

SPE/IADC 97282

Erosion
Erosion is usually only a problem is cased hole applications.
In a cased hole ESS application, a high flux from the
perforations can lead to erosion and a loss of sand control.
The ESS has been extensively tested to understand the
controlling factors and limits on erosion.
Over thirty erosion tests have been performed at the South
West Research Institute in San Antonio. These tests have
enabled the development of a model which can predict the
specific mass loss as a function of particle size, concentration,
type and rate. The specific mass loss is loss in grams from the
ESS per gram of solids impinging on the weave. The sand
integrity of ESS weave with various percentages of weight
removed was also measured. This allows the approximate
time to failure to be calculated for a given application.
For a given production rate the failure times are most
sensitive to the area open to flow. This is due to the specific
erosion being proportional to velocity to the power of between
two and four. The sand loading and the weight loss at loss of
sand control have a much lesser effect on the time to failure.
If in a given application erosion is becoming a concern,
then perforating with more shots per foot or with bigger entry
holes can dramatically extended the time to failure.
Bean up
It is of crucial importance to the productivity, reliability and
functionality of the ESS that it is brought on to initial
production in a controlled manner. After installation if the
well is rapidly opened then the filter cake will quickly fall off,
possibly mix with failed formation material and plug the ESS.
A high-pressure drop could then be applied across the ESS,
which could lead to severe deformation or collapse failure of
the screen. Deformation due to mud plugging takes the screen
out of compliance with the formation.
Once the ESS begins to deform, it can start to restrict
access. To mitigate against this, generic bean up procedures
have been developed25. These recommend that nodal analysis
be used to determine the FTHP and FBHP as a function of
choke setting and rate. If a downhole pressure gauge is used
then the FBHP can be directly measured. If not or if the gauge
fails then choke settings and FTHP must be relied upon.
The first choke setting must be small enough (e.g. 8 to 16
64ths) so that the total drawdown expected is less than 150psi.
This means that a drawdown larger than 150psi is unlikely to
be applied to the ESS. The first choke setting should be held
for fours hours to allow the filter cake to clean up. The 2nd
choke setting is designed such that the increment in drawdown
is less than 150psi and so on up to the maximum rate desired.
The total drawdown can exceed 150psi after the 1st choke
setting, only the increment in drawdown must be kept below
150psi.
After the well has been cleaned up, normal bean-up
procedures can be used.

Hole Quality
Well Trajectory
Weave Selection
Mud Selection
Borehole/ESS Stab.
Metallurgy
Erosion
Beanup

No Issue
Slight Issue
Major Issue

Figure 11 OPMS Checklist

OPMS Checklist and Failure Avoidance


The OPMS process can be summarised as a checklist where
the components are represented as a traffic light display
(Figure 11). The application should proceed only if all the
boxes are green.
An amber box signifies a slight issue and a red box a major
issue. Effort must be spent converting all the boxes to green.
Many of the ESS failures discussed could have been
avoided if the applications could have been more thoroughly
screened under OPMS. However there are always limitations
in the validity and extent of data made available. Of the initial
failures, eight were due to parted connectors, which have been
fully addressed with a new connector design. Five more
failures were due to sand size distributions being finer than
expected.
One ESS completion was damaged during
deployment, another failed due to possible collapse at the ETC
and one more through erosion failure at the ETC. All of these
could have been prevented.
Of the subsequent ESS failures, nine were due to high sand
production with grains larger than the weave. Some of these
had weaves sized in retrospect on inappropriate data. Three
wells had restrictions, which were related to formation
collapse, these could have been screened out. Better well
management could also have avoided the three erosive
failures. All the other failures were not really related to the
ESS, but they could also have been avoided.
Conclusions
ESS is a rapidly gaining a track record. The track record
shows that from a production performance viewpoint the ESS
has a very low skin and out performs other sand control
completion types in similar applications.

Average compliant openhole skin = 0.3


Average openhole skin = 0.9
Compliant expansion performance is better than
non compliant
No instances of poor or declining productivity
attributed to ESS

ESS has comparable initial failure rates and production failure


rates to OHGP and CHGP.

Combined initial failure rate 4.7%


Initial failure rate is decreasing with time due to
revised designs and better practices, currently
<2%
Production failure are 0.021 failures/well year

10

SPE/IADC 97282

Projecting this gives greater than 90% survival


rate after 20 years although this statistic must be
treated with caution.
The vast majority of failures are now avoidable if
the proper data is available and all procedures are
followed.
The single most important aspect is to have good sand size
data on which to base weave sizing. Resources spent
acquiring this data, or having two strings of different weave
size available could lead to a reduction in failures.
Acknowledgements
Many thanks to the operators who took the time to supply
information and to Weatherford for permission to publish the
paper.
Nomenclature
BHA = Bottom Hole Assembly
CHGP= Cased Hole Gravel Pack
CHFP= Cased Hole Frac Pack
EBC = ESS Bottom Connector
ESP = Electric Submersible Pump
ESS= Expandable sand screen (Weatherford Trademark)
ETC = ESS Top Connector
EWBS = ESS Well Bore Stability
FBHP= Flowing Bottom Hole Pressure
FTHP = Flowing Tubing Head Pressure
MMscf/d=Million standard cubic feet per day
NPT=Non Productive Time
OPMS=Operations Process Management System
OH = Open Hole
OHGP=Open Hole Gravel Pack
ppb = parts per billion
psi = pounds per square inch
SAS=Stand Alone Screen
T+D = Torque & Drag simulation
References
1

Metcalfe P. and Whitelaw C.: The Development of the First


Expandable Sand Screen, paper OTC 11032, presented at the 1999
Offshore Technology Conference, Houston, Texas, May 1999.
2
Heiland, J. Cook, J. Johnson, A. Jeffryes, B.:The role of the
annular gap in expandable sand screen completions SPE 86463
3
Weekse, A., Grant, S. and Urselmann, R.: Expandable Sand
Screen: Three New World Records in the Brigantine Field, paper
SPE 74549 IADC/SPE Drilling Conference, Dallas, February 2002.
4
Al-Lamki, A., Twycross, J. and Clarke, G.: Maximising
Productivity in the Scoter Gas Condensate Field, UKCS: A
Multidisciplinary Approach to Expandable Sand Screen Design and
Subsea Installation, paper SPE 91004 presented at the SPE Annual
Technical Conference and Exhibition, Houston, September 2004.
5
Wang, J.L., Aitken, S., Ren, G. and Yuan, M.: An Alternative
Wellbore Stabilisation and Sand Control Technology Application
of Expandable Sand Control System in multi-lateral Wells, paper
SPE 80445 presented at the SPE Asia Pacific Oil and Gas Conference
and Exhibition, Jakarta, April 2003.
6
Oluwatosin, A.O., Adegborioye, A.M. and Anya, S.O.: Reducing
Well Delivery Unit Technical Cost Using Expandable Sand Screens
(ESS) With Medium Radius Technology On Sidetrack Wells, paper
SPE 84160 presented at the SPE Annual Technical Conference and
Exhibition, Denver, October 2003.
7
Owoeye, O.O., Aihevba, L.O., Hartmann, R.A. and Ogoke, V.C.:

Optimisation of Well Economics By Application of Expandable


Tubular Technology, paper SPE 59142 presented at the 2000
IADC/SPE Drilling Conference, New Orleans, February 2000.
8
Erivwo, E., Owoeye, O.O. and Ogoke, V.C.: A Novel Approach to
Slim Well Delivery, paper SPE 67725 presented at the SPE/IADC
Drilling Conference, The Netherlands, March 2001.
9
Lau, H.C., van Vliet, J., Ward, M., Morin, D., Djamil, A., Kuhnert,
P. and Aldaz, W., Shanks, S.: Openhole Expandable-Sand-Screen
Completions in Brunei, paper SPE 87896, January 2004.
10
Jones, C., Cameron, J., Owoeye, O., & Baaijens, M.: Production
Issues Influencing the Choice of Completion Type in Horizontal
Wells With Sanding Problems, paper OTC15131 presentated at the
2003 Offshore Technology Conference, Houston 5-8 May 2003.
11
Metcalfe, P., Urselmann, R. and Saeby, J.: The Global Impact of
Expandable Sand Screens on Reservoir drilling and Completion,
paper SPE 67726 presented at the SPE / IADC Drilling Conference,
Amsterdam, March 2001.
12
Saeby, J., de Lange, F., Aitken, S.H., & Aldaz, W.: The Use of
Expandable Sand-Control Technology as a Step Change for MultipleZone SMART Well Completion - A Case Study, SPE 68634
13
Al-Shandoodi, M. and van Noort, R.: Expandable Tubular
Technology in PDO An Expandable Sand Screen Case History,
paper SPE 72287 presented at the IADC / SPE Middle East Drilling
Technology, Bahrain, October 2001.
14
Sanford, B., D., Terry, C., Bednarz, M., J., Palmer, C. and
Mauldin, D., B.: Case History: First Installation of an Expandable
Sand-Screen Completion in the Gulf of Mexico, paper SPE 13282
presented at the 2001 Offshore Technology Conference, Houston,
May 2001.
15
Jones, C. Cameron, J., & MacKay, E.: Productivity, Water
Breakthrough and Ultimate Recovery in Horizontal Wells With a
Range of Completion Types, paper OTC 15150 presentated at the
2003 Offshore Technology Conference, Houston 5-8 May 2003.
16
Gee, N., Jones, C., & Ferguson, S.: Towards the Expandable
Reservoir Completion: The Case for Open-Hole Completions, paper
OTC 16714, presented at Offshore Technology Conference,
Houston, May 2004.
17
Van Everdinging, A.F. and Hurst, W.:The Application of the
Laplace Transformation to Flow Problems is Reservoirs, Trans.,
AIME, vol. 186, pp 305-324, 1949
18
Jones, C., Tollefsen M, Somerville J.M., & Hutcheon R.: The
Pediction of Skin in Opnehole Sand Control Completions, paper
SPE 94527 presentation at the SPE 6th European Formation Damage
Conference, Scheveningen, The Netherlands, 25-27 May 2005.
19
Mason, D., Ramos, M., J., Pena, C., M., Cameron, J. and Jones, C.:
A Comparison of the Performance of Recent Sand Control
Completions in the Mokoko Abana Field Offshore Cameroon, paper
SPE 94651 presented at the 6th SPE European Formation Damage
Conference, The Netherlands, May 2005
20
King G.E., Wildt P.J. & OConnell, E.: Sand Control Completion
Reliability and Failure Rate Comparison With a Multi-Thousand
Well Database, SPE 84262
21
Ballard, T. and Beare, S.: Media Sizing for Premium Sand
Screens: Dutch Twill Weaves SPE 82244 presented at SPE 5th
European Formation Damage Conference, May 13-14 2003
22
Hodge, R.M., Burton, R.C., Constien, V. and Skidmore, V.: An
Evaluation Method for Screen-Only and Gravel-Pack Completions,
SPE 73772, presented at the 2002 SPE International Symposium on
Formation Damage Control.
23
Wilson, A.H, 1980: The stability of underground workings in the
soft rocks of the coal measures, PhD thesis, University of
Nottingham.
24
Somerville, J.M. & Smart B.G.D.: (1991) The prediction of well
stability using the yield zone concept. SPE 23127
25
PLB-ESS-023 Weatherford Product line bulletin on well start up.