Journal of Petroleum Science and Engineering 156 (2017) 835–844

Contents lists available at ScienceDirect

Journal of Petroleum Science and Engineering

journal homepage: www.elsevier.com/locate/petrol

Experimental investigation of factors affecting gravel pack efficiency for

thermal recovery wells in Bohai Bay, China
Fucheng Deng a, *, Yongcun Feng b, *, Chuanliang Yan c, *, Hai Lin d, Ning Gong d, Jie Wang a
a
Yangtze University, Jingzhou, PR China
b
The University of Texas at Austin, United States
c
China University of Petroleum, Qingdao, PR China
d
CNOOC China Limited, Tianjin Branch, Tianjin, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: Sand production from unconsolidated formations is a common problem of thermal recovery wells in Bohai Bay,
Sand control China. In order to select and optimize the method of sand control, statistics of the formation properties of the
Gravel packing major sand production zones is first reported in this paper. Next, laboratory experiments are carried out to
Sand production
investigate the effects of several controllable production parameters on the performance of the gravel packing
Thermal recovery wells
sand control method, which is the most widely used sand control technique in Bohai Bay. Both oil and sand
production rates for different gravel packing parameters are investigated. Based on the experimental results, an
empirical formula of Oil Productivity Index Per Meter (OPIPM), a parameter used to quantify the oil production
rate with sand control treatments, is obtained for this area. Finally, implications of this study on sand control
design and optimization of oil production rate are discussed.

1. Introduction
The existing studies on the selection and design of mechanical sand
control methods mainly focus on the analyses of the sand control screens
(Dees, 1992, 1993; Talaghat et al., 2009; Hugh and Ramos, 1995). Based
on experimental studies, Coberly (1937), Schwartz (1969), Saucier,
(1974), Tiffin et al. (1998), Gillespie et al. (2000), Abass et al. (2002),
and McCarthy and Mickelburgh (2010), proposed a series of methods for
the selection of mechanical sand control techniques, including slotted
liner, wire wrapped liner, metal mesh screen, metal cotton screen, pre-
pack screen, and gravel packing. Johnson's (Gillespie et al. 2000)
methods consider both the grain size distribution and the heterogeneity
of formation. Johnson's study is useful for selecting sand control methods
and optimizing the precision of the screens of the corresponding
methods. However, for gravel packing specifically, which is the major
sand control method for heavy oil wells in Bohai Bay, the effects of
several factors on the performance of this technique are still not fully
understood. These factors include in-situ viscosity of the production
fluid, thickness of the gravel pack, differential pressure of production,
and the size of screen and gravel.
Four thermal recovery wells were tested in the Nanbu oilfield in Bohai
Bay. The results show that thermal recovery treatment can remarkably
improve the production of heavy oil. However, three of the four wells
experienced serious sanding problems which finally caused the cease of
the productions. Two of the three sanding wells did not have sanding
problems during the cold production period, but significant sanding
occurred after thermal stimulation. This phenomenon implies that shear
failure occurs in the formation. To mitigate sand production, sand control
measures must be taken during well completions. Gravel packing is the
most popular sand control method in the Bohai Bay area. The influence of
various factors on the performance of gravel packing method needs to be
investigated for guiding sand control design for the thermal stimulation
wells in this area.
This paper presents laboratory tests of gravel packing treatments with
sand samples similar to the formation sands and test conditions similar to
the in-situ down hole conditions. Based on the test results, several pa-
rameters’ effects on gravel packing are investigated, including produc-
tion parameters (e.g. differential production pressure) and sand control
parameters (e.g. screen type, gravel size, sand retention precision,
packing thickness gravel). Furthermore, a guideline is proposed for the
design of gravel packing treatments in the unconsolidated formations in
Bohai Bay. The ultimate goal of this work is to use an optimized gravel
packing method to effectively mitigate sand production without
compromising the productivity of the wells.

* Corresponding authors.
E-mail addresses: dengfucheng128@163.com (F. Deng), yongcun.f@gmail.com (Y. Feng), yanchuanliang@163.com (C. Yan).

http://dx.doi.org/10.1016/j.petrol.2017.06.054
Received 27 March 2017; Received in revised form 30 May 2017; Accepted 20 June 2017
Available online 22 June 2017
0920-4105/© 2017 Elsevier B.V. All rights reserved.
F. Deng et al. Journal of Petroleum Science and Engineering 156 (2017) 835–844

Table 1
Characteristics of Minghuazhen heavy oil formation in the oilfields in Bohai Bay.

Oilfield Median grain size (d50), Uniformity coefficient Total clay content, Clay minerals Relative content of montmorillonite,
μm (UC) % %

Penglai 19-3 160–230 1–3 19.1–45 Illite, kaolinite 5

Nanbao 35-2 140–200 3–20 10–45 Montmorillonite, illite smectite mixed layer 60
Kenli 3-2 50–210 5–10 13.1 Montmorillonite, illite smectite mixed layer, 58
kaolinite
Bozhong 29- 70–180 4–11 12.6 Montmorillonite, kaolinite 68.7
4S
Bozhong 29-4 40–80 5–14 16.4 Montmorillonite, Kaolinite 55
Bozhong 34-1 85–253 3–15 10.2–20.3 Kaolinite, illite smectite mixed layer 10
Bozhong 28- 73–213 3–5 13.2 illite smectite mixed layer 65
2S

300

A1 well
250
A2 well

200
Viscosity(mpa·s)

150

100

50

0
50 100 150 200 250 300
Temperature T (℃)
Fig. 2. Gravel packing combination for 52 mm gravel thickness (tests 3, 4 and 8).

Fig. 1. Viscosity of heavy oil from Minghuazhen formation at different temperature.

Table 2
Designs of the orthogonal tests.

Test Influential factors

No.
Production Sizes of screen and gravel Oil viscosity, Gravel
pressure, MPa (SSG), μm mPa s thickness, mm

1 1 120 (20/40) 50 19
2 1 150 (16/30) 100 38
3 1 200 (10/30) 200 52
4 2 120 (20/40) 100 52
5 2 150 (16/30) 200 19
6 2 200 (10/30) 50 38
7 4 120 (20/40) 200 38
8 4 150 (16/30) 50 52
9 4 200 (10/30) 100 19

2. Characteristics of minghuazhen heavy oil formation in Bohai
Bay
In this work, information from several hundreds of thermal recovery
wells in Bohai Bay is investigated to characterize the sanding formations.
These wells spread over several oilfields, including Qinhuangdao 33-1S
oilfield, Penglai 9-1 oilfield, Nvda 16-1 oilfield, and Nanbao 35-2 oil
field. The major target heavy oil formation of these oilfields is the Min-
ghuazhen formation. The keycharacteristics of this formation are re-
ported in Table 1.
Several main aspects of the Minghuazhen formation can be obtained
from the statistical data in Table 1 as follows:
(1) The median grain size of the formation is in the range of
70–250 μm, indicating the formations are sand to fine-sand
formations.
(2) The uniformity coefficient is in the range of 3–15, indicating the
formations are relatively homogenous.
Fig. 3. Gravel packing combination for 38 mm gravel thickness (tests 2, 6 and 7).
(3) The total content of clay minerals ranges between 5% and 30%,
with the majority around 15%, indicating the formations are clay-
rich. The main clay types are montmorillonite and illite.
3. Experiment description
As aforementioned, the major sand control method used in the oil-
fields in Bohai Bay is gravel packing, with only a few wells using inde-
pendent quality screens. This is because the service life of independent
quality screen is relatively short compared with the gravel packing.
Therefore, the experimental study in this paper focuses on
gravel packing.
The prerequisite of a successful gravel packing operation is a proper
design of the accuracy of sand screen, the corresponding number, and the
corresponding thickness of the gravel. These 3 parameters must reach
appropriate match with each other to maximize the production of heavy
oil reservoirs. To this end, a laboratory sand control experiment is carried

836
F. Deng et al. Journal of Petroleum Science and Engineering 156 (2017) 835–844

200 mPa⋅s), as shown in Fig. 1, are analyzed in the experiments. The

influences of four controllable treatment factors in gravel packing prac-
tices on production capacity are investigated, which are production

Table 3
Experimental results.

Test Test conditions (Production pressure- Results Comparison

No. SSG-viscosity-gravel thickness)
Flow rate OPIPM (m3/ Sand
(kg/s) (d
m
MPa)) content
(‰)

1 1-120 (20/40)-50-19 0.02826 11.627 0.087

2 1-150 (16/30)-100-38 0.00387 1.47 0.3542
3 1-200 (10/30)-200-52 0.0004579 0.13,321 0.021
Fig. 4. Gravel packing combination for 19 mm gravel thickness (tests 1, 5 and 9). 4 2-120 (20/40)-100-52 0.003283 0.6158 0.42
5 2-150 (16/30)-200-19 0.000787 0.13,486 0.125
6 2-200 (10/30)-50-38 0.05028 9.5263 0.01004
out. The experimental are designed specifically for the Minghuazhen 7 4-120 (20/40)-200-38 0.01489 0.1346 0.063
heavy oil formation described above, according to the characteristics of 8 4-150 (16/30)-50-52 0.05647 5.1073 0.07617
the formation. The key aspects of the experiment design are as follows: 9 4-200 (10/30)-100-19 0.009112 0.8133 0.669
10 2-120 (20/40)-50-19 0.0040,334 7.8083 0.1608
11 1-200 (10/30)-100-19 0.00489 1.9111 0.1907
(1) The experiment is performed at room temperature;
(2) The median grain size of the sands is 150 μm and the uniformity
coefficient is 5–10. The synthetic formation sand is quartz sand
with the industrial standard mesh sizes. According to the particle Test statistics F
size distribution characteristics (D10, D40, D50, D90, UC, etc) of Test
If the sand content
Minghuazhen heavy oil formation, quartz proportion with Data meet requirement,
different industrial mesh is adjusted to match with the reservoir Sand content
OPIPM is the main
characteristics; evaluation index
(3) The clay content is 19%; half of the clay is montmorillonite and
the other half is illite; Evaluated the relevancy
(4) The method of sand control is gravel packing; Production of various factors and the
OPIPM
(5) The gravel material is carbon ceramics; Pressure optimal selection
(6) The type of sand control screen is covered metal screen (CMS).
Thickness
Three combinations of screen size and gravel size, used extensively in Viscosity SSG
of gravel
the oilfields in Bohai Bay, are tested in the experiments. These combi-
nations are: 120-μmscreen with 20/40-meshgravel, 150-μmscreen with
16/30-mesh gravel, and 200-μmscreen with 10/30-meshegravel. The Determine the influence
viscosity of the produced heavy oil from the Minghuazhen formation is of various factors in the sand control
measured under different temperature and the results are reported in
Fig. 1. Three typical viscosities of the heavy oil (i.e. 50, 100, and Fig. 6. The roadmap for experimental results analysis.

Fig. 5. Sand control experimental setup.

837
F. Deng et al. Journal of Petroleum Science and Engineering 156 (2017) 835–844

4.5

3.5 Production Sand retention Thickness of

Viscosity
pressure accuracy gravel layer

Variance analysis(F values)

3

2.5

1.5

0.5

0
Flow in Sand content Flow in Sand content Flow in Sand content Flow in Sand content
steady state in steady steady state in steady steady state in steady steady state in steady
state state state state

Fig. 7. Variance analysis results.

9 0.6
OPIPM in steady state
8
Sand content in steady state 0.5
7
OPIPM in steady state(m /d×m×MPa)

Sand content in steady state（‰）

6 0.4

5
3

0.3
4

3 0.2

2
0.1
1

0 0
50mPa·s 100mPa·s 200mPa·s

Fig. 8. Effect of fluid viscosity on OPIPM and sand production content.

pressure, screen size and gravel size, oil viscosity, and gravel thickness. The operation steps of sand control experiment are as follows:
In order to perform a comprehensive investigation of the four factors
to aid the selection of the best combination for effective sand control (1) Firstly after the equipment installation and commissioning, inject
without compromising the productivity of the well, a series of orthogonal oil and ensure that the oil flows through every shunt nets aperture
tests (Wu and Yang, 2006) are carried out in this work. The designs of which guarantees the formation of uniform radial flow, and
these tests are reported in Table 2. proofread the pressure measurement system.
To obtain the required gravel thicknesses as shown in Table 2, 500 - (2) Put the gravel packing combination in the autoclave and adjust it
diameter standard screen sand casings of different sizes are used. Com- to make sure that the axis of the screen overlaps the axis of the
binations of the screen and casing used in the experiments are shown autoclave, fill the gravel, put the rubber bearing on it, and finally
in Figs. 2–4. install the autoclave header. According to the particle size distri-
The experimental setup is shown in Fig. 5. During the tests, the fluid bution, the simulated formation sand is mixed by quartz sand with
(oil) is injected into the well. The fluid flow rate, sand production rate, the industrial standard mesh sizes. Then add the oil, sand and clay
and pressure difference across the gravel pack are measured in the ex- in mixing system.
periments. The oil-production index per meter (OPIPM), which is used as (3) Link the line pipe of the circulation system, the import and export
an indicator to evaluate the performance of the sand control treatment, is pressure sensor, the sand filter, the oil flow meter and data
determined using the experimental data as: acquisition card, and then set pump pressure according to the
formation pressure.
Q (4) Open computer data detection and collection system to record the
OPIPM ¼ (1)
ΔР*m pressure and flow data.
(5) Open air pump to test. The oil entered through high pressure
where OPIPM is the oil-production index per meter, m3/(m
d
MPa);
simulator entrance, flowed through simulation formation, annulus
Q is the fluid flow rate, m3/d; m is the length of the sand control screen,
(or gravel layer), screen. At last oil flowed out the autoclave from
m; ΔР is the production pressure difference, MPa.

838
F. Deng et al. Journal of Petroleum Science and Engineering 156 (2017) 835–844

0.7

0.6 2
y = -6E-05x + 0.0154x - 0.5044
2
R = 0.668

0.5
Fluid Fluid
Sand content(‰) flow carrying
0.4 ability capacity

0.3

0.2

0.1

0
20 40 60 80 100 120 140 160 180 200 220
Viscosityυ

Fig. 9. Sand content with fluid viscosity (solid line—fluid flow ability, dash line— fluid carrying capacity).

4.2. Overall analysis of orthogonal experiments

12
Variance analysis can be used to assess the relative impacts of mul-
10 tiple factors that influence the results of an experiment (Neville and
IP = 23.75e-0.027x
R² = 0.9263 Jensen, 2008). For the experiments in this work, variance analysis can
OPIPM(m /d×m×MPa)

8 directly reflect the effects of each experimental factor on OPIPM. In

addition, orthogonal analysis models can be established to evaluate the
6 mutual influence between production pressure difference, sand retention
3

accuracy (i.e. screen and gravel sizes), oil viscosity, and thickness of
4 gravel pack (Xiao et al., 2010; Peng et al., 2007). The results of variance
analysis are shown in Fig. 7.
2 In variance analysis, an indicator F (i.e. the vertical axis of Fig. 7) is
used to represent the relative impacts of various factors on the experi-
0 mental results. The higher the F value of a factor, the greater impact of
20 70 120 170 220 the factor on the experimental results. As aforementioned, OPIPM and
Viscosity(mPa·s)
sand production rate are used to evaluate the influences of different
Fig. 10. The OPIPM compared in different viscosity of fluid. experimental factors, including production pressure, sand retention ac-
curacy, oil viscosity, and gravel pack thickness. Fig. 7 shows the relative
the exhaust aperture. Sand filter separate the sand from the influence of various factors as follows:
mixture.
(6) Repeat the above steps of this experiment, at last clean the (1) Evaluation based on OPIPM: oil viscosity > production
equipment. pressure > gravel pack thickness > sand retention accuracy.
(2) Evaluation based on sand production rate: oil
Based on the experiment described in this section design, the in- viscosity > production pressure > gravel packing thickness > sand
fluences of various factors on OPIPM and sand production rate are retention accuracy.
evaluated. The experimental results and analysis are reported in the
section below.
4.3. Impacts of individual factors

4.. Experimental results and analysis

(1) Oil viscosity
According to the variance analysis above, it can be seen that oil vis-
4.1.. Results and analysis strategy
cosity is the most influential factor among the four factors investigated.
In thermal recovery, the viscosity of crude oil is highly dependent on the
The fluid flow rate, produced sand content, and the OPIPM in the
temperature of the reservoir. Therefore, we measured the OPIPM and
experiments with different test conditions are summarized in Table 3.
sand production content for different oil viscosities, as shown in Fig. 8.
Because the produced sand content and OPIPM directly illustrate the
The results indicate that OPIPM decreases with the increase of oil vis-
sand control result and the production capacity of the formation with
cosity. However, for sand content, the 100-mPa∙s viscosity leads to the
gravel packing, they are used as the main indexes for evaluating the
highest sand production, while the cases with 50- and 200- mPa∙s vis-
success of the gravel packing treatments in this paper. The analysis
cosity have very similar results.
procedure of the experimental results is given in Fig. 6, which will be
Figs. 9 and 10 further illustrate the influence oil viscosity on sand
further illustrated in the following sections.
content and OPIPM, respectively. Fig. 9 shows that the sand content first

839
F. Deng et al. Journal of Petroleum Science and Engineering 156 (2017) 835–844

5 0.6
OPIPM in steady state
4.5
Sand content in steady state
0.5
4

OPIPM in steady state (m /d×m×MPa)

Sand content in steady state (‰)

3.5
0.4
3

3
2.5 0.3

2
0.2
1.5

1
0.1
0.5

0 0
1MPa 2MPa 4MPa

Fig. 11. Effect of production pressure on OPIPM and sand content.

Table 4
Test results with different sand control designs and production pressure.

Test conditions 1 MPa 2 MPa 4 MPa

OPIPM (m3/(d
m
MPa)) Sand content (‰) OPIPM (m3/(d
m
MPa)) Sand content (‰) OPIPM (m3/(d
m
MPa)) Sand content (‰)

100,150 (16/30),38 1.47 0.354 0.7644 0.465 0.455 0.563

100,120 (20/40),52 0.5092 0.32 0.6158 0.42 0.298 0.544
50,120 (20/40),19 11.627 0.087 7.8083 0.1608 3.46 0.243
50,200 (10/30),38 16.194 0.08 9.5263 0.11 4.433 0.351

0.6

y = 0.0762x + 0.1246
0.5 2
R = 0.298

0.4
Sand content (‰)

0.3

0.2

0.1

0
0.5 1 1.5 2 2.5 3 3.5 4 4.5
Production pressure (MPa)

Fig. 12. Sand content versus production pressure with the same sand control design.

increases and then decreases with the increase of oil viscosity. The in-
flection point is at about 120 mPa s where the sand content reaches the
maximum value of 0.43‰. It means that both the maximum fluid car-
rying capacity and fluid flow ability are obtained at this critical viscosity.
This phenomenon implies that, in thermal recovery of heavy oil, we
should not only focus on reducing the viscosity of the oil, especially when
the initial oil viscosity is higher than the critical value; instead, we should
have a good knowledge of the critical viscosity value and keep the oil
viscosity far from it to avoid excessive sand production. Fig. 10 shows
that OPIPM decreases rapidly with the increase of oil viscosity. So, in
heavy oil thermal recovery, reducing the viscosity of reservoir fluid is
always beneficial for increasing of OPIPM. During the production of a
heavy-oil well, maintaining a low enough oil viscosity can not only in-
crease the production rate but also reduce the sand production.
(2) Production pressure
The design of sand control treatments for heavyoilwells depends not
only on OPIPM but also on the sand content in the produced fluid. If the
content of sand is high, it is likely to erode the sand screen and damage
the pump and sucker rod. Moreover, severe sand production can block up
the well and lead to the cease of production. All these undesirable phe-
nomena require sand cleaning work over and, consequently, lead to
high cost.
Fig. 11 shows the OPIPM and sand content with different production
pressures. It can be seen that a larger production pressure will lead to a
higher sand content. However, the largest production pressure of 4 MPa
results in the smallest OPIPM due to the influence of sanding, while a
moderate production pressure of 2 MPa yields the maximum OPIPM.
To examine the influence of production pressure on OPIPM and sand
content, four sets of experiments with different sand control designs and
different production pressures are established. The results of the

840
F. Deng et al. Journal of Petroleum Science and Engineering 156 (2017) 835–844

5 0.6

4.5 OPIPM in steady state

Sand content in steady state 0.5
4

OPIPM in steady state (m /d×m×MPa)

Sand content in steady state (‰)

3.5
0.4
3

3
2.5 0.3

2
0.2
1.5

1
0.1
0.5

0 0
19mm 38mm 52mm

Fig. 13. Effect of gravel pack thickness on OPIPM and sand content.

0.8

0.7

0.6
Sand content (‰)

0.5

0.4

0.3

0.2

0.1

0
10 15 20 25 30 35 40 45 50 55
Gravel pack thickness (mm)

Fig. 14. Sand content versus gravel pack thickness.

experiment are shown in Table 4.
Fig. 12 shows that with the same sand control design, sand content
increases with the increase of production pressure. The average sand
contents with 2- and 4-MPa production pressure are 1.37 and 2.02 times
of that of the 1-MPa case, respectively. The result indicates that main-
taining a low production pressure during heavy oil thermal recovery is
important not only for preventing sanding in the formation but also for
controlling sand movement through the gravel pack. Therefore, main-
taining an appropriate production pressure is a critical element to meet
the requirements for sand control.
(3) Thickness of gravel pack
Whether for vertical or horizontal wells, it is important to form a
continuously compacted gravel pack for sand control. The thickness of
gravel pack is one of the key factors that affect the effectiveness of the
sand control treatments. However, the thickness of gravel pack may be
restricted by the sizes of the production pipe or screen pipe in real situ-
ations. In the following, effects of the thickness of gravel pack on the
OPIPM and sand content described.
Fig. 13 shows OPIPM and sand content for different gravel pack
thicknesses. The results indicate that gravel pack thickness has a strong
impact on the OPIPM: the larger the thickness, the smaller the OPIPM. A
smaller gravel pack thickness of 19 mm results in a larger sand content.
However, for the cases with a gravel pack thickness of 38 and 52 mm, the
sand contents do not show very large difference. Although OPIPM has the
highest value with a gravel pack thickness of 19 mm, the sand content is
also very high and beyond the limit for safe production. After an integrate
devaluation of the effects of gravel pack thickness on OPIPM and sand
content, the thickness of 38 mm is considered to be the optimum value
for this case.
Fig. 14 shows the relationship between sand content and gravel pack
thickness. The results show that, in general, sand content decreases with
the increase of the gravel pack thickness. A thick enough gravel pack is,
therefore, required for blocking sand movements from the formation into
the well during production. However, for the experiments with a large
gravel pack thickness of 52 mm, there is still a particular case (the one in
the red circle in Fig. 14) exhibiting a sand content close to the allowable
upper limit of sand content (the red line in Fig. 14) for safe production.
Although gravel pack thickness is an important designing factor for sand
control, the effect of gravel pack thickness on sand production content is
less than the effects of oil viscosity and production pressure for the
particular heavy oil formation investigated in this study.
Fig. 15 shows the influence of gravel pack thickness on OPIPM with
different oil viscosities. The results indicate that OPIPM decreases with
the increase of gravel pack thickness. Moreover, when the oil viscosity is
relatively lower (e.g. 50 mPa s), the effect of gravel pack thickness on the
OPIPM is larger compared with the cases with a high oil viscosity (e.g.
100 mPa s). For the formation investigated in this study, the effect of
gravel pack thickness on OPIPM is very small when the fluid viscosity is
more than 100 mPa s.
(4) Sizes of screen and gravel (SSG)
The Saucier's criterion is commonly used for the design of the size of
gravel. This criterion suggests that D50 of the gravels should be equal to

841
F. Deng et al. Journal of Petroleum Science and Engineering 156 (2017) 835–844

14

12 Viscosity 50mPa·s

y = -0.1925x + 15.749 Viscosity 100mPa·s

R² = 0.9184
10 Viscosity 200mPa·s
OPIPM (m /d×m×MPa)

8
3

y = -0.0383x + 2.7248
4 R² = 0.9292

2 y = -5E-05x + 0.136
R² = 0.7999

0
15 20 25 30 35 40 45 50 55 60
Gravel pack thickness (mm)

Fig. 15. Effect of gravel pack thickness on OPIPM with different oil viscosities.

6 0.6

OPIPM in steady state

5 0.5
Sand content in steady state
OPIPM in steady state (m /d×m×MPa)

Sand content in steady state (‰)

4 0.4
3

3 0.3

2 0.2

1 0.1

0 0
120（20/40） 150（16/30） 200（10/30）

Fig. 16. Effects of sand control precision on OPIPM and sand content.

five to six times of the D50 of the formation grains. For different gravel
sizes, screens with different precisions are needed. The combined influ-
ence of gravel size and screen precision determines the final precision of
sand control. For the SSG combinations of 120 (20/40), 150 (16/30), 200
(10/30), the final precisions of sand control are 120,150, and 175 μm,
respectively. Fig. 16 shows the OPIPM and sand content for different SSG.
The results indicated that sand content decreases with the increase of
sand control precision. OPIPM has a maximum value with a high sand
control precision (SSG ¼ 120 (20/40)) and a minimum value with a
moderate sand control precision (SSG ¼ 150 (16/30)).
Fig. 17 shows the sand content for different sand control precisions. In
general, the sand content decreases with the increase of sand control
precision (lower value on the x-axis), which indicates that improving the
precision of sand control is an effective measure to reduce sand
production. However, for the sand control precision of 120 μm, there is a
set of experimental results with sand content approaching the allowable
upper limit as shown in Fig. 17. This result indicates that sand control
precision is not the only factor controlling sand production. By
comparing the effects of oil viscosity, production pressure and sand
control precision on sand content, it is found that the effect of sand
control precision is smaller than that the effects of oil viscosity and
production pressure.
Fig. 18 shows the OPIPM for different sand control precision with
different oil viscosities. The results show that OPIPM increases with the
decrease of precision of sand control (larger value on the-axis). When the
viscosity of the oil is 50 mPa s, the change of OPIPM with sand control
precision is most significant, followed by the case with a viscosity
100 mPa s.In the case of high viscosity, e.g. 200 mPa s, the effect of sand

842
F. Deng et al. Journal of Petroleum Science and Engineering 156 (2017) 835–844

0.8

0.7

0.6

Sand content (‰)

0.5

0.4

0.3

0.2

0.1

0
100 110 120 130 140 150 160 170 180 190 200
SSG (um)

Fig. 17. Sand contents for different sand control precisions.

10

7 y = 0.0268x + 3.5111
R2 = 0.1061
OPIPM (m /d×m×MPa)

5 Viscosity 50mPa·s
3

Viscosity 100mPa·s
4
Viscosity 200mPa·s
y = 0.0242x - 2.245
3
R2 = 0.9861

y = -2E-05x + 0.1378
1
R2 = 0.5506
0
110 130 150 170
SSG (um)

Fig. 18. Effect of sand control precision on OPIPM with different oil viscosities.

control precision on OPIPM is very small. This result indicates that in the
case of sand content meeting the production requirements, sand control
precision can be reduced appropriately to improve the production ca-
pacity for reservoir fluid with low viscosity.

4.4. Correlation model

Fig. 19. Schematic of gravel pack and screen.
Assuming the fluid flow across the gravel pack and the screen obeys
Darcy law (Fig. 19), SSG and thickness of the gravel pack can be cor-
rected to the initial permeability of the gravel pack under the corre-
sponding reservoir conditions (Kooijman et al., 1996). Next, by
combining the seepage lengths of the gravel pack and screen and the
experimental results obtained in this study, a correlation equation for
the relationship between OPIPM and oil viscosity, dimensions of the
gravel pack, and initial permeability of the screen and gravel pack can
be obtained as given below. Fig. 20compares the results calculated using
the correlation model with corresponding experimental results. It can be
seen there is a good match between them with a maximum error less
than 15%.

843
F. Deng et al.
12
OPIPM in steady state (m /d×m×MPa)
10
3
6
0
1 2
Fig. 20. Comparison of the results calculated using the correlation model with corresponding experimental results.
 1
πd L1 L2
analysis. The error of this correction model is within 15%.
IP ¼ ⋅eK2=K1 ⋅ 1:054 þ 0:0703 R2 ¼ 0:74
398:78μ K1 K2
where d is the diameter of screen pipe, mm; μ is the viscosity of the oil,
mPa⋅s; K1, K2 are the initial permeability of screen pipe and gravel, D; L1,
L2 are seepage length of screen pipe and gravel pack, mm; IP is the
OPIPM, m3/(d
m
MPa).
5. Conclusions
According to reservoir characteristics and fluid properties in Min-
ghuazhen heavy oil formation in Bohai Bay, China, laboratory sand
control experiments using gravel packing technique are designed and
conducted. The effects of oil viscosity, production pressure, SSG, and
gravel pack thickness on OPIPM and sand content in the produced fluid
are investigated.
The experimental results show that the relative impacts of the
investigated factors on OPIPM and sand content are: oil
viscosity > production pressure > gravel packing thickness > SSG. The
sand content increases with the increase of production pressure.
Considering the concurrent effects of production pressure on sand con-
tent and oil production, it is suggested that the production pressure for
Minghuazhen formation should be less than 4 MPa. If the SSG does not
have enough precision and/or the gravel pack does not have enough
thickness, the production pressure should be reduced accordingly.
Improving sand control precision and/or gravel pack thickness are
not the only considerations to solve the problem of sand production. Sand
production is also highly influenced by oil viscosity and production
pressure, which should be equally considered in sand control design.
Maintaining a low oil viscosity can effectively increase oil production and
reduces the sand production. Maintaining a low production pressure is
also an important mean to prevent sand production. When the produc-
tion pressure is low, sand control precision can be properly reduced (e.g.
to 200 μm) to maintain the production rate of oil. Sand production has
the maximum value when the fluid viscosity is around 120 mPa s; under
such situation, production pressure should not be greater than 4 MPa no
matter what SSG is used. With low fluid viscosity, the production pres-
sure can vary in a relatively large range.
Using the experimental data and assuming Darcy flow across the
gravel pack and screen, a correction formula for predicting OPIPM for the
Journal of Petroleum Science and Engineering 156 (2017) 835–844
Test data
Theoretical data
3 4 5 6 7 8 9 10 11
Test number
investigated formation is obtained based on multivariate regression
(2)
Acknowledgments
This paper was supported by The Yangtze Youth Fund (Grant No.
2015cqn43), The National Natural Science Foundation of China (Grant
No. 51504040), the Program for National Science and Technology
Major Project of the Ministry of Science and Technology of China
(2011ZX05024-003-02).
References
Abass, H.H., Nasr-El-Din, H.A., BaTaweel, M.H., Sand Control: Sand Characterization,
Failure Mechanisms, and Completion Methods. SPE Paper77686 Presented at the SPE
Annual Technical Conference and Exhibition San Antonio, Texas, 2002.
Coberly, C.J., 1937. Selection of Screen Openings for Unconsolidated Sands. API
Washington D.C, pp. 37–189.
Dees, J.M., Method of Sand Consolidation with Resin. United States Patent 5178218,
1993.
Dees, J.M., Sand Control in the Wells with Gas Generator and Resin. United States Patent
5101900, 1992.
Gillespie ,G., Deem, C.K., Malbrel, C., Screen Selection for Sand Control Based on
Laboratory Tests. SPE Asia Pacific Oil and Gas Conference and Exhibition held in
Brisbane, Australia, 16–18 October 2000.
Hugh, J.A., Ramos, J., 1995. Guidelines to Sand Control. Report. Halliburton Services
from Halliburton Co.
Kooijman, A.P., Van den Hoek, P.J., De Bree, Ph., et al., Horizontal Wellbore Stability and
Sand Production in Weakly Consolidated Sandstones. SPE Paper36419 Presented at
the SPE Annual Technical Conference and Exhibition Denver, Colorado, 1996.
McCarthy, T.A., Mickelburgh, I.J., 2010. The Evolution of Sand Control Completions
Practices for the Enfield Deep Water Development. SPE127394.
Neville, J., Jensen, D.B., 2008. Bias/variance analysis for relational domains. Lect. Notes
Comput. Sci. 4894 (17), 27–28.
Peng, Q., Zhou, F., Chen, Y., et al., 2007. Orthogonal experiment of mixture additives for
coal powder combustion enhancement. J. Wuhan Uni. Sci. Tech. Nat. Sci. Ed. 30 (05),
461–464.
Saucier, R.J., 1974. Considerations in Gravel Pack Design. SPE Paper 4030.
Schwartz, D., 1969. Successful Sand Control Design for High Rate Oil and Water Wells.
SPE Paper 2330.
Talaghat, M.R., Esmaeilzadeh, F., Mowla, D., 2009. Sand production control by chemical
consolidation. J. Petroleum Sci. Eng. 67, 34–40.
Tiffin, D.L., King, G.E., Larese, R.E., et al., 1998. New Criteria for Gravel and Screen
Selection for Sand Control. SPE Paper 39437.
Wu, M., Yang, K., 2006. Application of orthogonal design in the performances of
microwave absorbing material. J. Shanxi Univ. Sci. Ed. 29 (02), 162–164.
Xiao, Z., Yi, S., Fang, Z., 2010. Baotou rare earth ore concentrate roasting technological
parameters' optimized in medium baking temperature. J. Tianjin Univ. Sci. Technol.
25 (06), 26–29.
844