SPE 64398

Screen Selection for Sand Control Based on Laboratory Tests

George Gillespie, SPE, USF/Johnson Screens and Calvin K. Deem, SPE, Calvin Deem Consulting and Christophe Malbrel,
SPE, USF/Johnson Screens
Copyright 2000, Society of Petroleum Engineers Inc.
This paper was prepared for presentation at the SPE Asia Pacific Oil and Gas Conference and
Exhibition held in Brisbane, Australia, 1618 October 2000.
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.

Where possible, the results from the experimental tests are

compared to prior literature screen selection methods. Graphs
of the normalized data allow for a logical selection of the
appropriate screen for a given formation sand.
Technical Contribution:
1. A new procedure for the selection of production
screens applicable to any formation
2. A comparison of the new method with prior literature
methods
3. A comparison of performance of wire wrapped
screens and premium screens.

Abstract
Introduction
Oil and gas production from unconsolidated/weaklyconsolidated sands requires a production screen in the hole to
inhibit the movement of formation sand and keep the hole
open. For open hole completions where the sand control
screen directly retains the formation sand, proper screen
selection is necessary to provide optimum life and minimum
sand passing. Several screen sizing selection criteria are
available in the literature such as Saucier, Coberly, Schwartz
and others. However, these rules may not always be
applicable, especially when premium screens, with metal
meshes are used. These screens may use multi layers of wire
mesh and its complex shaped pore opening may result in
retention performance quite different from wire wrap screen
slots.
Experimental work has been done on a very fine (d50 = 115130 m) uniform (uniformity coefficient = d40/d90 = 3) and a
non-uniform (uniformity coefficient = ca 7) sands on a series
of commercial screens segments. These screens include
standard wire wrap and also premium grade screens. The
testing consisted of pressure drop measurements and sand
retention while the screen is subjected to a fluidized stream of
sand. Analysis of the data provides a method of measuring
screen performance during and after filter cake
bridging/building. Utilizing this data, screen performance, i.e.
comparative lifetime and sand passing can be projected based
on a maximum a pressure drop across the screen assembly.
This analysis method is applicable for any type of screen and
formation.

Open hole completions in oil/gas wells have been common

practice for the past several years. These wells have generally
been long horizontal wells. The many problems and
difficulties to get successful completions has been well
documented, as many operators have been disappointed with
the results.
It is now recognized that a successful horizontal open hole
completion requires careful analysis of many factors. These
are often dependent variables since a selection of a technology
or method for either the drilling or completion may limit the
use to only a few other interrelated products and services.
One of the key factors is to evaluate what type of screen
technology may be necessary for sand control. The reservoir
conditions usually dictate whether sand control is needed and
greatly influence whether the applied technology can be
successful.
When direct retention of reservoir sands is the likely
completion scenario, a good understanding of the capabilities
of the various types of screen technology is required. This
includes the screen's capabilities to retain the formation sands,
and its sensitivity to drill-in-solids based on the optimum
drilling fluids, and well completion cleaning additives and
methods that may be utilized.

G.GILLESPIE, C. K. DEEM, C. MABREL

Introduction Review of Published Guidelines

A considerable amount of published information and
guidelines exist for gravel pack to formation sand size ratios.
For example, Saucier1 indicates the median gravel pack size
should be about 5 to 6 times the mean formation sand size,
when there is severe flow disturbance. Schwartz2
recommends the ratio of 6 be based on the 10% coarse size
(d10) for uniform sands, and a ratio of 6 be based on the 40%
coarse size (d40) for non-uniform sands.
Schwarz also indicates that different ratios have been applied
for gravel packed formation sands. Ratios much below 6
should achieve a stable, but likely impaired pack, while a
ratio of 8 would be considered a maximum for effective
retention, and ratios in the 10 to 13 range will promote fines
invasion of the gravel pack. This will reduce the gravel pack's
permeability over time.
Published guidelines for retention of the formation sand
directly by the screen is basically limited to wire wrap screen
technology. Coberly3 indicates spherical particles (sand) will
not flow through rectangular slots as large as twice as the
particle size, or through circular holes that are up to 3 times
larger than the particle. The theory governing this
phenomenon is based on the flow contacting the screen with
the particles in mass, and the particles within the flow having
reasonable grain-to-grain contact. Under these conditions,
bridges or arches form on the screen surface, the effective size
of the opening decreases, and the finer grains are retained
despite being much smaller than the clean screen opening.
Penberthy4 also indicates for a rectangular slot; a factor
approaching 2.5, may form unstable bridges that collapse due
to applied pressure, resulting in excessive sand production.
In the water well industry, direct retention of the formation
sand has been successfully practiced for many years. For wire
wrap screens, the normal practice is to base the nominal screen
slot size on the 40% retained fraction (d40) of the formation
sand.5 Driscoll also indicates under certain conditions such as
corrosive waters, unreliable samples, well sorted, or
calcareous formations (shell fragments), more conservative
retention, or tighter screens, could be required. If coarse, nonuniform sands are encountered, the screen slots may be sized
to retain only 25-30% of the formation material. This criteria
is also applied to fine sand and larger formations of both
uniform and non-uniform classifications for water wells.
Table 1 the Wentworth Grain Size classification, is a good
reference for a sand's description and its size range. For
example, wire wrap screens are commonly used to retain fine
sand formations of the 125 to 150 micron range

SPE 64398

Table 1: Wentworth Grain Size Classification

DESCRIPTION

SAND RANGE

Very Coarse Sand

0.04 to 0.08 in (1000 to 2000 Micron)

Coarse Sand

0.02 to 0.04 in (500 to 1000 Micron)

Medium Sand

0.01 to 0.02 in (250 to 500 Micron)

Fine Sand

0.005 to 0.01 in (125 to 250 Micron)

Very Fine Sand

0.002 to 0.005 in (50 to 125 Micron)

Silt

0.0002 to 0.002 in (5 to 50 Micron)

Clay

Below 0.0002 in (<5 Micron)

Ref is Driscoll, FD "Groundwater and Wells", 2nd Ed., pg 410

On the other hand, Penberthy indicates direct retention of

formation sand by a screen is only suitable in clean, coarse
uniform sand. The difference in these two viewpoints can be
explained by how the well is completed and developed for
production. The major difference between the two
applications is the extensive use of mechanical well
development in water wells, whereby jetting, isolation tools
and other means are used to extensively work the formation
opposite the screen.6 This moves the finer sand fractions
through the screens and develops a coarse permeable filter
pack, essentially a gravel pack, around it as shown in
Figure 1.
Using this technology to develop oil/gas wells is not very
feasible, and would be too costly for use in deep, long
horizontal completions. Consequently, the well is generally
brought on stream by breaking down or removing the filter
cake without significant plugging of the screen. Generally,
this leaves fines in the near well bore as shown in Figure 1.
Considerable work has been done to test different types of
screen surfaces for their resistance to plugging by drill-insolids. Marken7 indicates that the screen opening had to be at
least 2.5 times larger than the mud particle d50 size or at least
partial plugging of a wire wrap would occur. In another
investigation, Lau8 based his evaluation on the one-seventh
rule for the mud solids d50 size versus the opening of wire
wrap and premium screen technology. For the different types
of screen tested, he found that wire wrap screens began to plug
when the ratio was about 3, while both prepacks and premium
screens began to plug significantly when the ratio was about 6.
Consequently, the successful application of screen technology
in an open hole oil/gas completion depends on the screen's
ability to meet two conflicting objectives. These are to
prevent sand production with a slow pressure build up, and to
pass any drill-in-solids fines without extensive plugging.

SPE 64398

SCREEN SELECTION FOR SAND CONTROL BASED ON LABORATORY TESTS

These demanding requirements have led to some new

rationales for successful applications for direct retention of
formation sand in open hole completions. Makestad9,
describes an evaluation method that defines a range of slots
for a wire wrap screen to retain a formation sand whereby the
smallest slot would have minimal plugging of the screen, and
the largest slot would not have excessive sand production.
The result is a range of slots that are effective for that
formation sand. In general, the criteria results in a larger
maximum opening then the Gulf Coast criteria that the slot
should be equal to the d10 size, but the opening is usually less
than Coberly criteria that the slot is equal to twice the d10
size. The report also indicates that wire wrap screens have a
low risk of plugging in coarse, well sorted sands, but the risk
is high if the sand is coarse and poorly sorted. In general, one
can conclude a wire wrap screen provides effective sand
retention in quite uniform reservoir sands that vary moderately
in their d50 sizes.

Test Method/ Procedures

The initial effort was to determine that it was feasible to test
screen assemblies that simulate how the screen is used in a
well, as tests have already been conducted without considering
the effects of protective covers, any support screens or the
perforated base pipe15.
A test apparatus was developed as shown in Figure 2. It is
similar to that used by Ballard12. A stable viscosified sand
slurry is pumped into a water stream where it was dispersed by
a pressure drop through 1/8 inch tubing. The resultant slurry
was then passed to the bare surface of the test screen at a
1.0L/min constant flow rate. The viscosity of the fluid at the
screen surface was about 1.0 centipoise similar to water. The
sand concentration was at 0.19 g/L. The effluent from the
screen was collected in quantities up to 2 L . The water was
decanted/evaporated and the passed sand measured to
determine the % passing or retained of the screen.

Tiffin10 suggests some new applications criteria be used with

different screen technologies and/or well completion strategies
based on the characteristics of the formation sand, The key
parameters of the sand are its sorting coefficient (d10/d95),
uniformity coefficient (d40/d90), and the amount of 325
mesh (44 micron) fines. The sorting coefficient looks at the
entire range of sand sizes within distribution, while the
uniformity coefficient only considers the finer fraction of the
sand.

To ensure a stable system was being utilized, tests were run

without a screen present . The effluent was collected in
samples during the test and the sand content measured. These
results are shown in Figure 3. As can be observed from the
diagram, the concentration is uniform over the duration of the
test. The concentration varied less than 10% between the
samples of a test run.

Considerable interest is paid to the presence of 325 mesh (44

micron fines) within a sand's distribution, as they may
promote plugging of the near well bore. Tiffin also addresses
some limitations in formation sand uniformity for successful
application of premium screens.

Three different sands were used to test the screens. The

criteria for selecting these sands were based on its average
particle size (d50 ) and uniformity coefficient (d40/d90).
Since no sands were readily available with the desired
characteristics, various sands were blended to give the
properties desired. A particle size distribution of the various
sands is shown in Figure 4.

Malbrel11 suggests some general guidelines on the application

of different premium screen pore size ratings using the
formation sand d50 size and the sand's uniformity coefficient.
For fine, highly, non-uniform (UC is over 5) and very fine,
non-uniform (UC is 3 to 5) sands, gravel packing the well is
advocated instead of an open hole completion.
These guidelines, or different forms of them, are currently
being applied for oil/gas open hole completions. Often,
operators are requiring different forms of testing to verify a
screen's retention capabilities, resistance to plugging by drillin-solids or some type of mechanical test.
As a provider of different types of screen technologies used in
oil/gas wells, it is necessary to understand the capabilities of
each product so effective application recommendations could
be made. In the current paper, a testing program was
established with an independent laboratory to conduct sand
retention tests on different types of screen technologies.

Composite test screens of 1.87" diameter disc samples were

assembled as individual layers in a holder and sealed along the
edges using RTV sealer. Spacer rings were also used to
accomplish a standoff between the screen and its base support.
Figure 5 shows a typical layout of a premium screen assembly
ready for the test fixture.
The screens are identified by a code number to conceal the
identity. The code numbers starting with D are premium
screens with drainage, those with P are premium grade
screens with no drainage and those with W are wire
wrapped screens. The various products were offset from the
flat base support by the use of ring spacers. Wire Wrap
screens and the premium screens without enhanced drainage
were offset from the base support by 0.050". Premium screens
with drainage were offset by 0.10".

G.GILLESPIE, C. K. DEEM, C. MABREL

The different types of screens tested were:

Premium Screens: A sintered laminate consisting of several
layers of special weave and/or square mesh. It utilized a
perforated cover and base support. The product was tested
with and without enhanced drainage.
Premium Screens: A special weave mesh supported directly
by a support screen of a slightly larger nominal opening,
which was supported by the base support. A protective cover
consisting of a series of slots in an offset pattern was also
utilized.
Wire Wrap Screens: 90 Wire Wrap screens of both keystone
shapes and heavy duty designs were supported directly by its
base support.
High flow wire wrap screens, which provide about 70% more
flow area than 90 wire were also tested. Both Keystone and
heavy duty designs were tested. The products were supported
by the base support, and a perforated protective cover was
used above the high flow wire wrap screens.
Premium screens of several different openings were tested,
while the gauge opening in the wire wrap screens was selected
to be 2 to 2.5 times the sand's d50 size to determine whether
wire wrap screens are effective at such large ratios in uniform
or non-uniform sands.
This paper reports on the retention results for three sands:

#2VFNU is a Very Fine, Highly Non-Uniform Sand,

with d50 = 114, UC = 7
#2VFU is a Very Fine Uniform Sand, with d50 = 124,
U.C. = 3
#1VFNU is a Very Fine, Highly Non-Uniform Sand, with
d50 = 49, UC = 12

There are plans to test fine to medium, uniform and nonuniform sands, and the very fine highly non-uniform sands
modified to have lower uniformity coefficients.
Figures 7, 8, 9 are microphotos of the feed sands. The photos
show the differences in the size range of the sands, and that
the sands are quite irregular in shape. The sand's appearance
is considered representative of what could be encountered in
an oil well.
Results
The retention of sand directly by a screen is very similar to the
theory of filtration as given in Barkman14. Sand sizes smaller
than the screen opening will form a bridge that will allow the
deposition of a filter cake on these bridges. The bridges may
be stable or unstable where the bridges collapse and reform
(see Markensen et al.). Sand of the size 1/6 to 1/2 the
opening size will generally bridge a larger gap although this

SPE 64398

ratio may change somewhat based on the sand concentration.

The pressure drop across the screen will then be composed of
two elements:
Pt = Ps + Pc
Where Pt = total pressure
Ps = Pressure drop across the screen
Pc = Pressure drop across the cake
Figure 10 shows a typical curve for pressure buildup and sand
removal for a screen. Three distinct phases were observed.
These are:

Cake Formation, 0 to about 0.3 psi: Initially, the cake

forms a very thin layer on the screen surface. Particles
are bridging and some retention occurs. During this
phase, most of the pressure drop is across the screen (Pt).
Cake Development, 0.3 to 5 psi: During this phase, the
arches are becoming stabilized as the filter cake thickens.
The % passing is being reduced as the range of formation
sand builds the filter cake. As the pressure builds to about
5 psi, the total pressure (Pt) becomes more dependent on
the pressure in the cake (Pc).
The slope, or rate of pressure build up, during this phase
is not linear.

Pressure Build Up, 5 psi and Higher: During this

phase, the pressure drop is almost entirely due to the
interaction of the range of sand particles in the filter cake.
This pressure build up is at a constant rate, as regression
analysis of the pressure build up curves of each screen
tested had 0.96 or higher values. This indicates the slope,
or rate of pressure build up, is constant with more
formation sand contacting the filter cake and screen
assembly.
The rate of change was found to be linear through about
250 psi pressure. At about 250 psi, premium screens with
the special weave retention surface had fluctuating
pressure build up and % passing curves. This indicates
the weave was flexing due to the applied pressure,
resulting in unstable arches. This phenomenon was not
observed with the sintered mesh premium screens.
During the pressure build up phase, the % sand passing is
normally a low figure, and does not change significantly
as the pressure increases, unless unstable arches occur.
For analysis purposes, the data for the cake development
and pressure build up phases were combined into one set.

SPE 64398

SCREEN SELECTION FOR SAND CONTROL BASED ON LABORATORY TESTS

Uniform Sand Results: Eight different screens were tested in

a very fine uniform sand (#2VFU). The test was stopped after
15 psi was reached because of the elapsed time averaged
almost an hour per test. Table 2 summarizes the data for %
passed and rate of pressure build up. The premium screens,
which had ratios of about 1.6 to about 2.0 for their nominal
opening vs the d50 sand size, all exhibited quick development
of the filer cake, and excellent retention. All % passing were
less than 1.2%. At the same time, the rate of pressure build up
is low, as the rates all were less than 2.5 psi per gram of feed.
The sintered mesh design (P1A4) with enhanced drainage had
the lowest rate (1.44 psi/g). This rate was almost 40% less
than the premium screen (P2A3) with the highest ratio (2.33
psi/g).
The wire wrap screens, which had their openings sized at
about twice the sand d50 size, were not as effective retaining
the sand. Considerably more sand passed during the cake
formation phase, and the % passing in the cake formation and
pressure build up phase was up to 1.26%. Despite the
somewhat lower retention efficiencies, the pressure build up
rates were higher on the wire wrap screens. The 90 wire
design's rate was about three times the premium screen's rate,
while the high flow designs were about 50% higher. The tests
indicate this large of slot to sand d50 ratio should not be used
for wire wrap screens unless up to about 1.5% fines
production are acceptable.
The tests also indicate that high flow wire wrap have much
lower pressure build up rates than 90 wire designs in very fine,
uniform sands. The 90 wire design should not be considered
for direction retention of very fine, uniform sand.
Non-Uniform Sand Results: Two different highly, nonuniform sands were tested. Ten different screens were tested
in #2VFNU sand with a d50 of 114 micron and a uniformity
coefficient of almost 7. Six different screens were tested in
#1VFNU sand with a d50 of 49 micron and a uniformity
coefficient slightly over 12.
Both tests we conducted to higher pressure drops. Values are
shown to 100 psi, which represents a likely point when
pressure build up across the filter cake and screen would be
unacceptable in a production well. The time to conduct the
tests were less than 12 minutes for the d50, 114 sand, and less
than 6 minutes for the d50, 49 sand.
For the six premium screens tested, Table 3 shows the %
passed and rate of pressure build up for the #2VFNU sand.
The screens generally had about 5% passing for the pressure
build up phase. A review of their % passing indicates values
as high as 3.9% for P3A3 to as low as 1.4% for D1B4 for the 5
to 100 psi range. The results indicate the larger pore opening
meshes have basically the same retention characteristics as the
finer mesh openings.

The rate of pressure build up was much higher on the premium

screens in the non-uniform sands. The results varied from a
low of 72 psi/g for D1B4 (250 micron sintered mesh, with
drainage) to 441 psi/g for P2A2 (#110 micron special weave).
The premium screen pore size had a big impact on the rate of
pressure build up despite making little difference in its
retention capabilities.
The wire wrap screens had very poor retention of the #2
VFNU sand, as 50 to 70% of the sand passed during the cake
development and pressure build up phases, and these values
were only about 6 percentage points less if only the pressure
build up phase is considered.
The pressure build up rates of the wire wrap screens were
much higher than the premium screens. The 90 wire designs
were six times as high, while the high flow wire wrap designs
exceeded the best premium screen rate by about 350%
Table 4 summarizes the data for the #1VFNU sand. The
sand's 325 mesh (44 micron) fines were about 45% of the
material. None of the sintered mesh screens had an enhanced
drainage system.
Since a typical test lasted about 3.5 minutes, it was not
feasible to collect data in smaller increments than listed in the
table. Consequently, % passing is not available for only the
pressure build up phase (5 to 100 psi). The results for the
premium screens show the % passing is sensitive to the pore
size to d50 ratio as the relatively low ratio (1.22 for a 60
micron screen) had substantially less sand passing than the
200 micron screen, which had about a 4.0 ratio. The premium
screens with a ratio of about 2.30 had somewhat lower
retention rates than the 60 micron screen.
Only high flow wire wrap screens were tested with the #1
VFNU sand. Both screens had substantially higher build up
rates than the nominal 100 micron premium screens, while
their % passing values were more like the 200 micron
premium screen. These tests indicate the slot ratios in the 2.5
range for wire wrap screens are not effective for very fine,
highly non-uniform sands.
Uniform vs Non-Uniform: Figure 11 compares the rate of
pressure build up and percent passing for the sintered mesh
with enhanced drainage (D1B4). Its percent passing of the
uniform sand is somewhat lower during the cake formation
and development phases, but the % passing are of comparable
values after the cake is fully developed in the pressure build
up phase. At the same time, its rate of pressure build up is
substantially faster in the non-uniform sand.
Consequently, this premium screen (D1B4) provides effective
retention for both uniform and non-uniform sands when its
ratio of pore size to sand d50 is as high as 2.5.

G.GILLESPIE, C. K. DEEM, C. MABREL

Table 5 summarizes the % passing and pressure build up for

various screens in all of the sands tested. The values listed for
the sintered mesh products are the enhanced drainage designs.
While the screens show comparable % passing, the rate of
pressure build up of the special mesh is 30% or more higher
for the 200-250 micron series. It's apparent the sintered mesh
designs perform better than the special weave designs of
comparable micron ratings. The most likely explanation is
reduced effective open area of the special mesh since it is in
direct contact with a support screen of substantially lower flow
area. This reduces the mesh's effective flow area because as
increasing pressure is applied, the special weave conforms
tightly to the round wire support screen resulting in reduced
available flow area.
T able 5: C o m p arison of Prem ium Screen s for U n ifo rm and
N o n-U n ifo rm San ds
SPEC .
SIN T .
SIN T .
SPEC .
W EAVE
M ESH
M ESH
W EAVE
SIN T .
110
115
200
230
M ESH 60
SAN D C R IT ER IA

#2VFN U

% Passing

0.42%

1.20%

0.96%

1.69

2.33

1.44

Slope (psi/g)

#2VFN U

#1VFN U

SIN T .
M ESH
250

% Passing

8.38%

5.23%

5.79%

5.91%

5.26%

Slope (psi/g)

441.31

186.67

134.54

227.32

72.04

16.35%

28.79%

23.79%

41.27%

433.8

467.03

520.67

389.59

% Passing
Slope (psi/g)

Effect of Drainage: Table 6 compares the % sand passing

and rate of pressure build up for two identical sintered mesh
screens. The only difference in the screens was the use of a
drainage pattern under DB4 to insure almost full access of the
flow into the base support series of holes. For both types of
sand, the % passing was essentially the same, but the rate of
pressure build up was over 30% less for the enhanced drainage
design.
Table 6: Comparison of Drainage Effects
P1A4
250 um Sint.
Mesh
No Drainage

D1B4
250 um Sint.
Mesh with
Drainage

0.39%

0.96%

2.19

1.44

% Passing

5.86%

5.26%

Slope (psi/g)

172.41

72.04

SAND CRITERIA

#2VFU

% Passing
Slope (psi/g)

#2VFNU

SPE 64398

Wire Wrap vs Premium: The retention results for the wire

wrap screens confirm that the use of slot ratios of 2.0 or more
result in a slow development of a filter cake, and the cake may
be unstable as the pressure increases. Generally, this would
result in excessive sand production.
The test results also indicate the effects of non-uniform sand
on wire wrap screens in terms of both percent passing and rate
of pressure build up. The increased available open area of the
high flow wire wrap significantly improves both its retention
capabilities and the rate of pressure build up. The form of the
wire wrap opening did not appear to significantly affect its
retention capabilities, but the rate of pressure build up was
somewhat lower for the keystone design versus the heavy duty
design.
Evaluation Methods
Underdown15 recommends an evaluation method that
summarizes a screen's retention capabilities, called sand
control factor, and rate of pressure build up, called
performance factor, into one nomograph called a screen
efficiency plot. The method uses normalized values from the
tests, and the inverse ratios are plotted. Consequently, an ideal
screen would have values of 1 for both factors.
An alternative method is proposed as shown in Figure 12,
which contains Screen Efficiency Plots (SEP) for the three
sands tests. The nomographs are based on plotting directly the
% sand passed in the cake formation and pressure build up
phases versus the rate of pressure build up, or slope, for each
screen tested. Consequently, the "ideal" screen would have
values that are closest to zero for both performance criteria.
The proposed SEP analysis method is advantageous in that
direct comparisons of a screen's performance can be made;
and each screen's performance can be easily evaluated versus
acceptance criteria established for a particular type of well.
For example, the SEP plot could utilize one of the following
acceptance criteria for sand production:
a.
b.
c.

A gas well may be based on the overall % passing results.

An oil well may be based on a maximum pressure build
up rate and % passing (5 to 100 psi).
If an oil well, such as a deep, offshore application, can
have only limited sand production, then combining the %
passing of cake development and pressure build up phases
(0.3 to 100 psi) could be used.

Based on evaluating a screen in terms of acceptable % sand

passing and the lowest rate of pressure build up. The best
screens to use for the sands tested are:

#2VFU: D1B4
#2VFNU: D1B4
#1VFNU: P1A2

SPE 64398

SCREEN SELECTION FOR SAND CONTROL BASED ON LABORATORY TESTS

Based on the results of these laboratory tests, a selection guide

is offered for using screen technology for direct retention of
formation sand. Figure 13 shows three different sand sizes
and the suggested use for different types of screen technology
based on the uniformity coefficient of the formation sands and
the variation of the d50 size of the sand in a reservoir. Direct
retention is not effective when the sand is highly non-uniform,
and/or there is too much variability in the reservoir sand sizes.
The sensitivity to these variables increases with smaller
formation sands. It is apparent from these tests that open hole
completions are not always the best strategy to complete an oil
or gas well. Gravel packing the bore hole using shunt tube
technology16, or other means, may be the best strategy to
pursue when the formation sand is highly non-uniform or
varies significantly throughout the reservoir.
The test results also indicate that criteria for premium screens
can be established based on the nominal screen opening and
the sand size and its uniformity. Figure 14 is offered for the
sintered mesh screens. The special weave premium screen
would have similar ranges, except the minimum sand sizes
would be a little larger because of the likelihood of the sand
arches becoming unstable at pressures exceeding 200 psi.

Acknowledgements
The authors of this paper thank US Filter for their support for this work, and
permission to publish this paper.
We offer thanks to numerous industry leaders, such as Clive Bennett, David
Bryant, Bob Burton, Richard Hodge, George King, Steve Svedeman, Dave
Underdown and George Wong for their thoughtful and timely information on
open hole completions for oil/gas wells.
References
1.

R. S. Saucier, "Considerations in Gravel Pack Design" SPE Reprint No.

43, 1996, 205-212 .

2.

D. H. Schwartz,"Successful Sand Control Design for High Rate Oil and

Water Wells", J.Petr. Tech (Sept 1968), 1193-1198.

3.

C. J. Coberly, SPE Reprint No. 5 "Well Completions", 156-175.

4.

W.L. Penberthy and C.M. Shaughnessy, SPE Series on Special Topics

Volume 1 "Sand control", 1-5.

5.

F.G. Driscoll,"Groundwater and Wells", 2nd Ed., 434-437

6.

F.G. Driscoll, "Groundwater and Wells", 2nd Ed., 497-522

7.

C. Marken, P. Markestad, and O. Rorvik, "Criteria for Back Production

of Drilling Fluids Through Sand Control Screens", SPE 38187,
prepared for 1997 SPE European Formation Damage Conference,
Hague, Netherlands 2-3 June.

8.

H. C. Lau, and C. L. Davis, "Laboratory Studies of Plugging and CleanUp of Production Screens in Horizontal Wellbores", SPE 38638,
presented at 1997 SPE Annual Technical Conference, San Antonio, TX,
5-8 October.

9.

P. Makestad, {?} Christie, A. Espedal and O. Rovik, "Selection of

Screen Slot Width to Prevent plugging and Sand production", SPE
31087, presented at 1996 SPE Formation Damage Symposium,
Lafayette, LA, 14-15 February.

Conclusions
1.

Based on the formation sand size and its uniformity

coefficient, % passing and the rate of pressure build up
can vary significantly for a screen.

2.

For premium screens, effective retention of very fine and

fine sands can be achieved when the screen opening is up
to 2.5 times the d50 sand size, provided the formation
sand's uniformity coefficient does not exceed 6.0.

3.

Enhanced drainage significantly reduces the rate of

pressure build up on a premium screen without
significantly affecting its retention capabilities.

10. D. L. Tiffin, G. E. King, R. E. Larese, and L. K. Britt, "New Criteria

for Gravel and Screen Selection for Sand Control", SPE 39437,
presented at 1998 SPE Formation Damage Control, Lafayette, LA, 1819 February.

4.

The ratio of a wire wrap screen's opening should not

exceed twice the d50 size of the sand, and the use of wire
wrap screens to directly retain formation sands is limited
to sands which are somewhat uniform, and the reservoir
d50 sand size variability is quite small.

11. C. Malbrel, , A. Proycyk, and J. Cameron, "Screen Sizing Rules and

Running Guidelines to Maximize Horizontal Well Productivity", SPE
54743 presented at 1999 SPE European Formation Damage Conference,
Hague, Netherlands, 31 May-1 June.
12. T. Ballard, N. Kageson-Loe, and A. Mathisen, The Development and
Application of a Method for the Evaluation of Sand Screens, SPE
54745, presented at the 1999 European Formation Damage Conference,
The Hague, The Netherlands, 31 May-1 June 1999.
13. R. E. Larse, Private Correspendence
14. J. H. Barkman and D. D. Davidson, Measuring Water quality and
Predicting Well Imparement, Journal of Petroleum Technology, July,
1972, p 865-873.
15. D.R. Underdown and R.C. Dickerson, "The Nominal Sand Control
Screen: A Critical Evaluation of Screen Peformance", SPE 56591, 1999
SPE Annual Conference and Exhibition, Houston, TX, 3-6 October.
16. L.C. Jones, R.J. Tibbles, L. Myers, D. Bryant, J. Hardin, and G. Hurst,
"Gravel Packing Horizontal Well Bores with Leak-Off Shunt Tubes",
SPE 38640, 1997 SPE Annual Technical Conference, San Antonio, TX
5-8 October.

G.GILLESPIE, C. K. DEEM, C. MABREL

SPE 64398

T a b le 2 : T a b le o f R e s u lts fo r # 2 V F U S a n d ,
d50 = 124, U C = 3
C o n c e n t r a tio n = 0 .1 9 g /L
S a m p le
D1B3
0 to 0 .3 p s i
0 .3 to 1 5 p s i
O v e r a ll

Feed, g
0 .5 2
8 .8 7
9 .3 9

Passed,
g
0 .1 0
0 .0 4
0 .1 3

%
Pass
1 8 .4 1
0 .4 2
1 .4 2

S lo p e ,
p s i/g
0 .5 7 4 1
1 .6 9 0 7
1 .5 9 6 7

P1A4
0 to 0 .3 p s i
0 .3 to 1 5 p s i
O v e r a ll

Feed, g
0 .5 6
6 .8 5
7 .4 1

Passed,
g
0 .0 8
0 .0 3
0 .1 1

%
Pass
1 4 .2 9
0 .3 9
1 .4 4

S lo p e ,
p s i/g
0 .5 3 5 7
2 .1 8 9 2
2 .0 2 3 8

D1B4
0 to 0 .3 p s i
0 .3 to 1 5 p s i
O v e r a ll

Feed, g
0 .6 3
1 0 .4 0
1 1 .0 4

Passed,
g
0 .2 3
0 .1 0
0 .3 3

%
Pass
3 5 .9 4
0 .9 6
2 .9 6

S lo p e ,
p s i/g
0 .4 7 3 5
1 .4 4 1 9
1 .3 5 9 1

P2A3
0 to 0 .3 p s i
0 .3 to 1 5 p s i
O v e r a ll

Feed, g
0 .4 5
6 .4 4
6 .8 9

Passed,
g
0 .1 0
0 .0 8
0 .1 8

%
Pass
2 2 .5 6
1 .2 0
2 .5 8

S lo p e ,
p s i/g
0 .6 7 2 8
2 .3 2 8 6
2 .1 7 7 8

W 1A6
0 to 0 .3 p s i
0 .3 to 1 5 p s i
O v e r a ll

Feed, g
1 .7 2
2 .1 6
3 .8 9

Passed,
g
1 .2 1
0 .5 4
1 .7 4

%
Pass
7 0 .0 1
2 4 .7 6
4 4 .8 2

S lo p e ,
p s i/g
0 .1 7 4 1
6 .9 2 9 7
3 .8 5 8 0

W 1B6
0 to 0 .3 p s i
0 .3 to 1 5 p s i
O v e r a ll

Feed, g
2 .7 1
2 .1 4
4 .8 5

Passed,
g
2 .2 5
1 .2 6
3 .5 1

%
Pass
8 3 .0 3
5 8 .9 0
7 2 .3 7

S lo p e ,
p s i/g
0 .1 1 0 7
6 .9 9 5 8
3 .0 9 0 0

W 1C 6
0 to 0 .3 p s i
0 .3 to 1 5 p s i
O v e r a ll

Feed, g
4 .2 0
4 .4 9
8 .6 9

Passed,
g
3 .5 0
1 .0 6
4 .5 6

%
Pass
8 3 .3 3
2 3 .5 2
5 2 .4 3

S lo p e ,
p s i/g
0 .0 7 1 4
3 .3 4 0 4
1 .7 2 6 0

W 1D 6
0 to 0 .3 p s i
0 .3 to 1 5 p s i
O v e r a ll

Feed, g
4 .5 0
4 .1 1
8 .6 1

Passed,
g
3 .5 3
1 .2 3
4 .7 6

%
Pass
7 8 .4 4
2 9 .9 7
5 5 .3 1

S lo p e ,
p s i/g
0 .0 6 6 7
3 .6 5 0 7
1 .7 4 2 4

M ic r o n
O p e n in g
200

R a tio t o
d50
1 .6 1

250

2 .0 2

250

2 .0 2

230

1 .8 5

235

1 .9 0

275

2 .2 2

243

1 .9 6

257

2 .0 7

SPE 64398

SCREEN SELECTION FOR SAND CONTROL BASED ON LABORATORY TESTS

T a b le 3 : T a b le o f R e s u lts fo r # 2 V F N U S a n d ,
d 5 0 = 1 1 4 , U C = 7 C o n c e n tr a tio n = 0 .1 9 g /L
M ic r o n
O p e n in g
S a m p le
F e e d , P a sse d ,
%
S lo p e ,
g
g
P a ss
p s i/ g
D 1 B 2
1 1 5
0 to 0 .3 p s i
0 .3 5
0 .1 0 2 9 .0 7
0 .8 7
0 .3 to 1 0 0 p s i
0 .5 4
0 .0 3
1 8 6 .6 7
5 .2 3
O v e r a ll
0 .8 8
0 .1 3 1 4 .6 6
1 1 3 .6 4

D 1 B 3
0 to 0 .3 p s i
0 .3 to 1 0 0 p s i
O v e r a ll

F e e d ,
g
0 .5 1
0 .7 4
1 .2 6

P a sse d ,
g
0 .1 4
0 .0 4
0 .1 8

%
P a ss
2 7 .5 0
5 .7 9
1 4 .4 4

S lo p e ,
p s i/ g
0 .5 9
1 3 4 .5 4
7 9 .3 7

P 1 A 4
0 to 0 .3 p s i
0 .3 to 1 0 0 p s i
O v e r a ll

F e e d ,
g
0 .8 4
0 .5 8
1 .4 2

P a sse d ,
g
0 .1 6
0 .0 3
0 .2 0

%
P a ss
1 9 .2 7
5 .8 6
1 3 .7 3

S lo p e ,
p s i/ g
0 .3 6
1 7 2 .4 1
7 0 .4 2

D 1 B 4
0 to 0 .3 p s i
0 .3 to 1 0 0 p s i
O v e r a ll

F e e d ,
g
0 .7 7
1 .3 9
2 .1 6

P a sse d ,
g
0 .2 7
0 .0 7
0 .3 4

%
P a ss
3 4 .6 5
5 .2 6
1 5 .7 4

S lo p e ,
p s i/ g
0 .3 9
7 2 .0 4
4 6 .3 0

P 2 A 2
0 to 0 .3 p s i
0 .3 to 1 0 0 p s i
O v e r a ll

F e e d ,
g
0 .3 3
0 .2 3
0 .5 6

P a sse d ,
g
0 .0 9
0 .0 2
0 .1 1

%
P a ss
2 6 .7 2
8 .3 8
1 9 .2 9

S lo p e ,
p s i/ g
0 .9 0
4 4 1 .3 1
1 7 8 .5 7

P 3 A 3
0 to 0 .3 p s i
0 .3 to 1 0 0 p s i
O v e r a ll

F e e d ,
g
0 .5 4
0 .4 4
0 .9 8

P a sse d ,
g
0 .1 8
0 .0 3
0 .2 1

%
P a ss
3 3 .7 5
5 .9 1
2 1 .4 3

S lo p e ,
p s i/ g
0 .5 5
2 2 7 .3 2
1 0 2 .0 4

W 1 A 6
0 to 0 .3 p s i
0 .3 to 1 0 0 p s i
O v e r a ll

F e e d ,
g
0 .8 7
0 .2 1
1 .0 8

P a sse d ,
g
0 .5 5
0 .1 3
0 .6 8

%
P a ss
6 3 .7 5
6 0 .9 5
6 2 .9 6

S lo p e ,
p s i/ g
0 .3 4
4 7 6 .1 9
9 2 .5 9

W 1 B 6
0 to 0 .3 p s i
0 .3 to 1 0 0 p s i
O v e r a ll

F e e d ,
g
0 .8 8
0 .2 2
1 .1 0

P a sse d ,
g
0 .5 8
0 .1 6
0 .7 4

%
P a ss
6 6 .3 4
7 2 .7 3
6 7 .6 4

S lo p e ,
p s i/ g
0 .3 4
4 5 4 .5 5
9 0 .9 1

W 1 C 6
0 to 0 .3 p s i
0 .3 to 1 0 0 p s i
O v e r a ll

F e e d ,
g
1 .9 3
0 .4 0
2 .3 3

P a sse d ,
g
1 .2 3
0 .1 9
1 .4 2

%
P a ss
6 3 .7 4
4 6 .2 5
6 0 .7 7

S lo p e ,
p s i/ g
0 .1 6
2 5 0 .0 0
4 2 .9 2

W 1 D 6
0 to 0 .3 p s i
0 .3 to 1 0 0 p s i
O v e r a ll

F e e d ,
g
1 .6 7
0 .3 2
1 .9 9

P a sse d ,
g
1 .2 0
0 .1 7
1 .3 6

%
P a ss
7 1 .5 3
5 1 .8 8
6 8 .4 4

S lo p e ,
p s i/ g
0 .1 8
3 1 2 .5 0
5 0 .2 5

R a tio to
d 5 0
1 .0 1

2 0 0

1 .7 5

2 5 0

2 .1 9

2 5 0

2 .1 9

1 1 0

0 .9 6

2 3 0

2 .0 2

2 3 5

2 .0 6

2 7 5

2 .4 1

2 4 3

2 .1 3

2 5 7

2 .2 5

10

G.GILLESPIE, C. K. DEEM, C. MABREL

SPE 64398

Table 4: Table of Results for #1VFNU

d50 = 49, UC = 12 Concentration = 0.19 G/L
Micron
Opening
Sam ple
Feed, Passed,
Slope,
g
g
% Pass
psi/g
P1A1
60
0 to 0.3 psi
0.25
0.06
22.38
1.22
0.3 to 100 psi
0.23
0.04
16.35
433.86
Overall
0.48
0.09
19.47
209.76

P1A2
0 to 0.3 psi
0.3 to 100 psi
Overall

Feed, Passed,
g
g
% Pass
0.46
0.25
53.55
0.19
0.05
23.79
0.65
0.29
44.80

Slope,
psi/g
0.65
520.67
153.05

P1A3
0 to 0.3 psi
0.3 to 100 psi
Overall

Feed, Passed,
g
g
% Pass
0.85
0.53
62.57
0.26
0.11
41.27
1.11
0.64
57.64

Slope,
psi/g
0.35
389.59
90.21

P2A2
0 to 0.3 psi
0.3 to 100 psi
Overall

Feed, Passed,
g
g
% Pass
0.37
0.18
49.93
0.21
0.06
28.79
0.58
0.25
42.18

Slope,
psi/g
0.81
467.03
171.30

W 1C3
0 to 0.3 psi
0.3 to 100 psi
Overall

Feed, Passed,
g
g
% Pass
0.31
0.15
49.06
0.08
0.04
49.06
0.38
0.19
49.06

Slope,
psi/g
0.98
1318.39
260.86

W 1D3
0 to 0.3 psi
0.3 to 100 psi
Overall

Feed, Passed,
g
g
% Pass
0.23
0.12
50.29
0.11
0.04
33.06
0.35
0.16
44.65

Slope,
psi/g
1.28
877.19
287.36

Ratio to
d50
1.22

115

2.35

200

4.08

110

2.24

120

2.45

110

2.24

SPE 64398

SCREEN SELECTION FOR SAND CONTROL BASED ON LABORATORY TESTS

Figure 1

11

Figure 2 Sand Retention Test Apparatus

Water
Well
Screen

Analysis Procedure

Water Well: Fluids are moved across screen to

remove fines near well bore and create stab le
interface with the c oarse formation sand against
screen.

Wire Wrap
Screen

Install the screen holder into test facility. Purge all the air
from the screen holder with water.
Measure the amount of water contained by the screen
holder.
Set up the data acquisition system to log the pressure drop
on the screen and the fluid flow rate.
Inject the slurry into the transfer cylinder. Eliminate any
air in the connecting lines.
When slurry enters the transfer cylinder, start the test to
limit any of the sand from falling out of the slurry.
Start recording screen pressure and flow rate.
With the flow bypassing the test screen, turn on the water
pump to 1 L/min. Turn on the HPLC pump at 10 ml/min.
Flow for a couple of minutes to purge the flow lines.
Collect a sample from the bypass line to document the
feed sand concentration.
Start the test by closing the screen bypass line and
opening the test cell outlet valve.
Catch the fluid exiting the test cell in about 2 L batches
until the screen pressure drop exceeds 300 psi.
Collect about 1000 L of fluid from the bypass line at the
end of the run.
Turn off sand slurry flow, and water flow after flushing
flow lines.
Slowly drain fluid from screen holder. Carefully remove
screen to collect and weigh the amount of sand on the
shroud and screen face.
Weigh the 2 L fluid samples then decant its water, dry and
weigh sand to determine the sand concentration in each
sample.

Data Collection
1.
Oil/Gas Well: One way flow has smaller particles
in the near well bore, and may be in contact with
the screen.

2.
3.

Record pressure drop vs amount of sand feed during test.

Extrapolate the initial, 0.3, 1.0, 5.0, 15, 50, and 100 psi values
from the data.
Determine total solids retained (TS) at 0.3, 1.0, 5.0, 15., 50, and
100 psi.
Retain the TS samples for possible further analysis.

12

G.GILLESPIE, C. K. DEEM, C. MABREL

Figure 3: Sand Concentration Variance No Screen

Present

SPE 64398

Figure 4: Formation Sand Curves

# 1 VF Non-Uniform d10 = 229.6
#2 VF Uniform
d10 = 164.0
#2 VF Non-Uniform d10 = 245.4

Sand Concentration, g/L

S an d C on te n t
Th e ore ti cal S an d C on te n t

d40 = 63.3
d40 = 141.0
d40 = 154.7

d50 = 49
d50 = 124
d50 = 114

d90 = 5.1
d90 = 48.8
d90 = 22.5

U.C. = 12
U.C. = 3
U.C. = 7

Percentage Larger, %

0.25

100
0.2

80

#2 VFU
#2 VFNU
#1 VFNU

60

0.15

40
0.1

20
0

0.05

10

0
0

10

15

20

25

30

35

40

45

1000

Particle Diameter, um

50

Time , min

Screen Assembly
Installed in
Housing

Base
Support

Spacer

1.87" OD
Assembly

100

Premium
Screen

Spacer
Protective Cover

Figure 5 Photos simulated Premium 200 Micron screen assembly with

spacer between protective cover and Premium Screen surface and 0.10"
spacer for drainage between Premium screen and perforated base
support.

10000

SPE 64398

SCREEN SELECTION FOR SAND CONTROL BASED ON LABORATORY TESTS

13

Figure 10: Typical Pressure Build Up and % Passing

Curves
Pressure, psi

% Passing, %
Press ure

% Removal

100

40

80

32

60

Figure 7: Feed Sand #2VFU, d50 = 124,

UC = 3

24

Pressure Build Up

16

40
20

Cake
Development

0
0

Cake
Formation

Sand F eed to Screen, g

Figure 11 Comparison of Pressure Build Up and % Passing of D1B4
for Very Fine Uniform and Non-Uniform Sands
Figure 8: Feed Sand #2VFNU,
d50 = 114, UC = 7

% Passing %

Pressure, psi

100
80
60
40
20
0

50
40
30
20
10
0
0

Sand Feed to Screen, g

Figure 9: Feed Sand, #1VFNU, d50=49,

UC = 12

VFNU Pressure

VFU Pressure

VFNU % Removal

VFU % Removal

14

G.GILLESPIE, C. K. DEEM, C. MABREL

SPE 64398

Figure 13 Selection Guide for Open Hole Completions

Screen Technology for Direct Retention

Figure 12 Screen Efficiency Plots (SEP)

#2VFNU Sand d50 = 114, UC = 7
Overall Performance (0.3 psi to 100 psi)

Very Fine Sands (d50 is 50 to 125 m)

60
40
20
0
0

100

200

300

400

500

Sand Uniformity (d40/d90)

% S a n d P a sse d P a ss

80

GravelPack*

6
5
4
3

Wire Wrap
(High Flow)

Premium
Screens

2
1

S lo p e , p si/g

50%

25%

D1B 4

D1B 3

P1A 4

P3A 3

W 1C 6

W 1D 6

W 1B 6

W 1A 6

D1B 2

P2A 2

100%

Reservoir Heterogeneity
(d50 variations)
* Alternate Path Technology is the recommended gravelpack method.

#2VFU Sand d50 = 124, UC = 3

60
50
40
30
20
10
0
0

W1C6

W1D6

W1A6

W1B6

Gravel Pack*

6
5
4

Wire Wrap
(High Flow)

Premium
Screens

3
2

Wire Wrap

Prepack
(Thin Annulus)

1
25%
50%
Reservoir Heterogeneity
(d50 variations)

S lo p e , p si/g

P1A 4

Fine Sands (75 m < d50 is 50 to 125 m )

Sand Uniformity (d40/d90)

% S a n d P a sse d P a ss

Filter Cake Growth(0.3 psi to 15 psi)

D1B4

D1B3

100%

* Alternate Path Technology is the recommended gravel pack method.

P2A 3

#1VFNU Sand d50 = 49, UC = 12

Medium Sands (d50 is 250 to 500 m)

Filter Cake Growth (0.3 psi to 100 psi)

Sand Uniformity (d40/d90)

50
% Sand Passed Passed

40
30
20
10
0
0

200

400

600

800

1000 1200 1400

Gravel Pack*

6
5

Wire Wrap
(High Flow)

Premium
Screens

4
3
2

Wire Wrap

Prepack
(Thin Annulus)

1
25%
50%
Reservoir Heterogeneity
(d50 variations)

100%

* Alternate Path Technology is the recommended gravel pack method.

Slope, psi/g
P1 A3

P1 A1

P2 A2

P1 A2

W1 D 3

W1 C 3

SPE 64398

SCREEN SELECTION FOR SAND CONTROL BASED ON LABORATORY TESTS

15

Figure 14: Sintered Mesh Premium Screen Opening vs Formation Sand Range

NON UNIFORM - 112 TO 550 MICRON

250 :m

UNIFORM 200 TO 450 MICRON

NON UNIFORM 90 TO 440 MICRON

200 :m

Very Fine
Sand
(50 to 125M)

UNIFORM 120 TO 360 MICRON

NON UNIFORM 45 TO 220M

115 :m
UNIFORM 60 TO
180M

NON UNIFORM 27
TO 132 MICRON

60 :m

UNIFORM 36
TO 108
MICRON

100

Medium Sand
250 to 500M

Fine Sand
(125 to 250M)

150

200

250

300

350

400

FORMATION SAND D50 SIZE

450

500