This article was downloaded by: [Shokir, eissa Mohamed]

On: 21 August 2009

Access details: Access Details: [subscription number 914078040]
Publisher Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Petroleum Science and Technology

Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713597288

Experimental and Numerical Investigation of Proppant Placement in Hydraulic

Fractures
E.M. El-M. Shokir a; A. A. Al-Quraishi b
a
Faculty of Engineering, Mining, Petroleum and Metallurgical Engineering Department, Cairo University,
Giza, Egypt b King Abdulaziz City for Science & Technology, Riyadh, Saudi Arabia
Online Publication Date: 01 October 2009

To cite this Article El-M. Shokir, E.M. and Al-Quraishi, A. A.(2009)'Experimental and Numerical Investigation of Proppant Placement in

Hydraulic Fractures',Petroleum Science and Technology,27:15,1690 1703

To link to this Article: DOI: 10.1080/10916460802608768
URL: http://dx.doi.org/10.1080/10916460802608768

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.

Petroleum Science and Technology, 27:16901703, 2009

Copyright Taylor & Francis Group, LLC
ISSN: 1091-6466 print/1532-2459 online
DOI: 10.1080/10916460802608768

Experimental and Numerical Investigation of

Proppant Placement in Hydraulic Fractures
E. M. El-M. Shokir1 and A. A. Al-Quraishi2
1 Cairo

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

University, Faculty of Engineering, Mining, Petroleum and

Metallurgical Engineering Department, Giza, Egypt
2 King Abdulaziz City for Science & Technology, Riyadh, Saudi Arabia

Abstract: In hydraulic fracturing treatments, a fracture is initiated by rupturing

the formation at high pressure by means of a fracturing fluid. Slurry, composed
of propping material carried by the fracturing fluid, is pumped into the induced
fracture channel to prevent fracture closure when fluid pressure is released. Productivity improvement is mainly determined by the propped dimension of the fracture,
which is controlled by proppant transport and proper proppant placement. Settling
and convection (density driven flow) are the controlling mechanisms of proppant
placement. In this study, proppant transport and placement efficiency of four nonNewtonian fluids with controlled density differences was experimentally investigated
and numerically simulated. Small glass model was used to simulate hydraulic fracture
and parameters such as slurry volumetric injection rate, proppant concentration, and
polymer type (rheological properties) were investigated.
It has been observed that small glass models easily and inexpensively simulated
flow patterns in hydraulic fractures and the flow patterns observed are strikingly
similar to those obtained by very large flow models used by previous investigators.
Convection was observed to be significant flow mechanism even with small density
contrast. As viscous to gravity ratio increases, due to increasing slurry injection
rate or decreasing proppant concentration, convection settling decreases and proppant
placement efficiency increases. Increasing non-Newtonian flow behavior index (n) by
using different types of polymers shows more gravity underrunning and less proppant
placement efficiency. Therefore, larger slurry volumes are needed to be injected to
prop the entire fracture height. Experiments conducted were simulated and some of
the simulated experiments were presented. The simulator quantitatively replicates the
experimentally observed.

Keywords: fracture, fluids, hydraulic, numerical, proppant, placement

Address correspondence to E. M. El-M. Shokir, King Saud University, Petroleum

& Natural Gas Engineering Department, P.O. Box 800, 11421 Riyadh, Saudi Arabia.
E-mail: shokir@ksu.edu.sa
1690

Proppant Placement in Hydraulic Fractures

1691

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

1. INTRODUCTION
Bulk density gradient between the injected proppant carrying slurry and
fracture pad fluid can cause strong gravity driven motion (convective settling),
which in turn result in proppant misplacement at the bottom of a fracture.
Cleary and Fonseca (1992) indicated that convection is scaled to fracture
width while settling is scaled to proppant diameter. The ratio of convective
velocity to settling velocity may be of the order of 100 to 1000 even under
ideal conditions and it may even be much larger due to the encapsulation
effects or the suspension of solid particles in almost solid like viscoelastic
gels. Clark and Courington (1994) supported experimentally the findings of
Cleary and Fonseca (1992) using a point source model and they indicated
that convection is less important in thick viscous fluids. Clark and Zhu
(1994) expanded the range of density differences used in the previous work
of Clark and Courington (1994) and concluded that more comprehensive
studies are needed to understand the convection process. Barree and Conway
(1994) developed a computer model that yields evidence that convection is
dominant under certain conditions of viscosity and fracture nonuniformity.
Unwin and Hammond (1995) presented and proved numerically that settling
and convection can occur under field conditions and convection rates are
slightly greater in sheet flow than in homogeneous flow. Mobbas and Hammond (2001) numerically estimated the amount of convection during proppant
placement. They indicated that convection and settling rates are greater in
sheet flow than in homogeneous flow. Clark and Zhu (1995) developed and
validated experimentally dimensionless groups to predict the importance of
convective flow. Both dimensionless groups are ratios of vertical to horizontal
forces driving the flow. The non-Newtonian fluids viscous to gravity ratio was
expressed as follows:
Nvg D 2.4 C 2=n/n .kq n =ghw .1C2n/ /

(1)

The aim of this study is to investigate experimentally and to simulate

numerically the proppant transport and proppant placement in hydraulic fractures. The experiments were conducted using small glass model where easily
and inexpensively different parameters such as, slurry volumetric injection
rate, fracturing fluid viscosity, proppant concentration, and fracturing polymer
type were investigated.

2. EXPERIMENTAL SETUP AND PROCEDURE

Figure 1 shows a schematic of the used experimental setup. It consists of
a small rectangular glass model, syringe pump, and digital video camera
connected to a laptop computer for video image analysis. The glass model
consists of two glass plates, 8 inches long and 4 inches high. An entrance

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

1692

E. M. El-M. Shokir and A. A. Al-Quraishi

Figure 1. Experimental setup.

port about 0.125 inches diameter centered from top to bottom was drilled to
qualify the entrance to act as a point source. The same size hole was drilled
at the outlet end. Strips of shim stock 0.0236 inch thick were cut and placed
around the edges between the two glass plates. The outside edges of the two
plates were sealed and bonded together with a layer of epoxy and supported
mechanically with binder clips.
All experiments were conducted by displacing pad fluid by proppant
carrying slurry. Experiments were recorded and video frames of the experiments were captured with image processing software capable of manipulating,
displaying, and analyzing captured images. White board was placed behind
the glass flow model to ensure even distribution of light.
Four different non-Newtonian fracturing fluids were investigated in this
study. These fluids are 20 and 30 lb/Mgal guar solutions, 5 lb/Mgal polyacrylamide solution, and 10 lb/Mgal Xanvis solution. The concentrations were
chosen so that the viscosities of the fluids were close to each other at the
shear rate range of the experiments conducted, and to minimize inflation of
the slot due to pressure caused by injecting viscous fluids through the model.
Silica flour was chosen in place of proppant to provide density contrast
between the displaced and displacing fluids, to simulate fracture slurry opaque
to light with particles small compared to the flow model gap width, and to
minimize particulate settling during the experiments. Sieved silica particles
with diameter range of 0.00250.0035 inches were added to the polymer
solutions to simulate proppant-carrying slurries. The used silica flour concentrations are 2, 4, and 6 lb/gal added for the guar solution and 4 lb/gal
added for both the polyacrylamide and Xanvis solutions.
Slurry viscosity was calculated using the Generalized Equation for Fracturing Fluid Actual Rheology correlation (GEFFAR). Slurry viscosity is de-

Proppant Placement in Hydraulic Fractures

1693

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

termined as a function of solids loading and shear rate as follows (Stim-Lab,

1995):
 n0 1
P
tan 1

L
 D 0 .1 Cn / a
(2)
 n0 1

P
1
tan
H
Figures 2 and 3 are a plot of the measured viscosity data of the 20 and
30 lb/Mgal guar solution respectively obtained using the Brookfield and
FAN-35 viscometers matched with the GEFFAR correlation using 0 lb/gal
silica concentration and power law model. GEFFAR predictions for the viscosity profile of the 0, 2, 4, and 6 lb/gal silica flour added slurries are given
in the same figures. Similarly, Figure 4 and 5 present viscosity data measured
for clear 5 lb/Mgal polyacrylamide and 10 lb/Mgal Xanvis solution and the
GEFFAR viscosity prediction of these polymers with 0 and 4 lb/gal silica
flour added. Table 1 lists fluids physical properties. Coefficients a D 1.8,
H D 10000000, and Cvmax D 0.56 were used to obtain slurry viscosities. To
validate the use of GEFFAR for modelling slurry viscosities, the viscosity
of 30 lb/Mgal guar with low 0.5 lb/gal added silica flour was measured at
different shear rates using Brookfield and FAN-35 viscometers. A good match
was obtained between the measurements and the GEFFAR prediction using
the previously mentioned a, Cvmax, H parameters.

Figure 2. Viscosity vs. shear rate of 20 lb/Mgal guar solution with different silica
flour concentrations.

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

1694

E. M. El-M. Shokir and A. A. Al-Quraishi

Figure 3. Viscosity vs. shear rate of 30 lb/Mgal guar solution with different silica
flour concentration.

3. RESULTS AND DISCUSSION

As mentioned previously, four different fracturing fluids were used in this
study. Density contrast between the displacing and displaced fluids was
achieved by adding different concentrations of silica flour to the displacing

Figure 4. Viscosity vs. shear rate of 5 lb/Mgal polyacrylamide solution with different
silica flour concentrations.

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

Proppant Placement in Hydraulic Fractures

1695

Figure 5. Viscosity vs. shear rate of 10 lb/Mgal Xanvis solution with different silica
flour concentrations.

fluids. Different parameters were investigated and these are volumetric

injection rate, fluid viscosity, proppant concentration, and polymer type.
3.1. Effect of Volumetric Injection Rate
Experiments using 20 lb/Mgal guar slurry loaded with 4 lb/gal added silica
flour displacing 20 lb/Mgal base solution, at different injection rates, were
Table 1. Fluids physical properties at room temperature (70 F)

Fluid
Guar (20 /Mgal)
Guar (20 lb/Mgal) with 2 lb/gal
Guar (20 lb/Mgal) with 4 lb/gal
Guar (30 lb/Mgal)
Guar (30 lb/Mgal) with 2 lb/gal
Guar (30 lb/Mgal) with 4 lb/gal
Guar (30 lb/Mgal) with 6 lb/gal
Polyacrylamide (5 lb/Mgal)
Polyacrylamide (5 lb/Mgal) with
4 lb/gal
Xanvis (10 lb/Mgal)
Xanvis (10 lb/Mgal) with 4 lb/gal


(gm/cc)

K
(lbf secn /ft2 )

o
cp

L
sec 1

1.0007
1.1375
1.2512
1.0010
1.1378
1.2537
1.3531
1.0013
1.1378

0.68
0.68
0.68
0.51
0.51
0.51
0.51
0.50
0.50

0.0021
0.0032
0.0042
0.0116
0.0155
0.020
0.028
0.018
0.023

70.00
87.88
115.4
380.0
491.9
645.9
874.0
1800
1960.9

15
12
10
5
2.5
1.3
1.2
0.6
0.5

1.2534
1.3526

0.50
0.50

0.028
0.035

2368.45
2924.75

0.4
0.3

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

1696

E. M. El-M. Shokir and A. A. Al-Quraishi

Figure 6. Flow pattern of 20 lb/Mgal guar slurry loaded with 4 lb/gal silica flour
displacing base guar solution at (a) 0.5 cc/min, (b) 1 cc/min, (c) 2 cc/min injection
rates.

compared. Three injection rates of 0.5, 1, and 2 cc/min were tested and frames
captured show decreasing convective settling with increasing injection rate.
As injection rate increases viscous force overrides gravity force and proppant
placement efficiency improves. Figure 6 shows the video frames of some
conducted experiments at 5, 15 and 30% of the time required to inject one
pore volume. The extent of convection, in the experiments, is indicated by
the slope of the interface between the injected slurry and the resident pad
fluid at the middle of the fracture model. Similar conclusions can be drawn
for the 30 lb/Mgal guar solution displaced by a 4 lb/gal added slurry of the
same solution as shown in Figure 7 but with higher placement efficiency due
to the high viscosity and hence higher carrying capacity of the 30 lb/Mgal
guar solution.
Images indicated less proppant settling at the bottom of the fracture
model as injection rate increases. However, convection dominates the flow of
slurry in the model.
3.2. Effect of Proppant Concentration
Different proppant concentrations were added to 20 and 30 lb/Mgal guar
solutions. The concentrations were chosen to show the worse and best case
scenarios. First 30 lb/Mgal base guar solution was displaced out of the model

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

Proppant Placement in Hydraulic Fractures

1697

Figure 7. Flow pattern of 30 lb/Mgal guar slurry loaded with 6 lb/gal silica flour
displacing base guar solution at (a) 0.5 cc/min, (b) 1 cc/min, (c) 2 cc/min injection
rates.

by slurry of 2, 4, and 6 lb/gal added silica flour suspended on 30 lb/Mgal guar

solution. Figure 8 shows some captured frames for the conducted experiments
at injection rate of 0.5 cc/min. The frames show the frontal flow pattern at
5, 15, and 30% of the time needed to inject one pore volume. The captured
frames indicate an increased lateral transport at the bottom of the model as a
result of downward convection due to the increasing density contrast between
the displacing slurry and displaced base fluid as proppant concentration
increases. This in turn increases the consistency index (K). Hence, The net
effect was a decrease in the viscous to gravity ratio and therefore less proppant
placement efficiency. Evidence of downward flow pattern exists at late time
even with low proppant concentration.
Similarly, Figure 9 presents the captured frames of flow pattern of
20 lb/Mgal guar solution loaded with 2 and 4 lb/gal silica flour displacing 20 lb/Mgal base guar solution at injection rate of 0.5 cc/min.
The captured frames indicate more downward convection even at early
injection time for both proppant concentrations. This behavior is due to
the low viscosity characteristic and hence the low carrying capacity of the
20 lb/Mgal guar fluid compared to the 30 lb/Mgal. The results of both
fluids illustrate that convective settling controls the flow pattern whenever
bulk fluid density gradient exist. However, convective settling decreases as

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

1698

E. M. El-M. Shokir and A. A. Al-Quraishi

Figure 8. Flow pattern of 30 lb/Mgal guar slurry loaded with (a) 2 lb/gal silica flour,
(b) 4 lb/gal silica flour, (c) 6 lb/gal silica flour displacing base guar solution.

viscosity force override the gravity force due to density gradients. This is
clearly noticed when comparing the frames captured when using 4 lb/gal
added slurries in both guar solutions (Figures 8b and 9b) where more
convection occurs for 20 lb/Mgal than that observed with 30 lb/Mgal
solution.

Figure 9. Flow pattern of 20 lb/Mgal guar slurry loaded with (a) 2 lb/gal silica flour,
(b) 4 lb/gal silica flour displacing base guar solution.

Proppant Placement in Hydraulic Fractures

1699

Figure 10. Flow pattern of 4 lb/gal silica flour slurries displacing base solutions of
(a) 30 lb/Mgal guar, (b) 5 lb/Mgal polyacrylamide, (c) 10 lb/Mgal xanvis fluids.

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

3.3. Effect of Polymer Type

Solutions of the three polymers (30 lb/Mgal guar, 5 lb/Mgal polyacrylamide
and 10 lb/Mgal Xanvis solutions) were displaced with proppant carrying
slurries of the same solutions loaded with 4 lb/gal silica flour. Figure 10 shows
the captured frames of the three polymers at 30% of the time required to
inject one pore volume. The frames show better proppant placement efficiency
of guar solution compared to that obtained using polyacrylamide or Xanvis
solutions. More gravity underrunning was observed for the displacements of
the last two polymers. The increased underrunning is attributed to the lower
flow behavior index (n) values of these solutions compared to that of guar,
which corresponds to greater shear thinning behavior. Therefore, the dense
slurry at the bottom of the slot flows easily while the less dense fluid at the top
become static and develops flow resistance. This underrunning phenomenon
continues as the displacing fluid progress in the slot. In field operations, this
observation indicates the need for larger injected volumes required to prop
the entire fracture height with fluids that have lower (n) values than that of
guar.

4. NUMERICAL SIMULATION
Selective experiments were simulated using a constrained configuration of
Marathon Oil Companys fully three-dimensional hydraulic fracture simulator
GOHFER. This configuration allowed for simulation of displacements in
fixed fracture geometry, appropriate for comparisons to displacements in
a laboratory setting. It incorporates a series of sequential finite difference
solutions based on a fixed spatial grid. This version is named GOHFLAB.
Figure 11 shows the experimental and simulated frames for slurries of 6 lb/gal
silica flour added to 30 lb/Mgal guar solution displacing 30 lb/Mgal guar
solution. This figure show the video frames of the experiments and their
corresponding numerical simulation predictions at 5 and 30% of the time
required to inject one pore volume at three conducted injection rates. The
amount of particle settling in the simulation frames is indicated by the color

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

1700

E. M. El-M. Shokir and A. A. Al-Quraishi

(a)

(b)

(c)
Figure 11. Experimental and simulated frames of 4 lb/gal guar slurry displacing
30 lb/Mgal base guar solution at (a) 0.5 cc/min, (b) 1 cc/min and (c) 2 cc/min injection
rates.

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

(a)

(b)

(c)
Figure 12. Flow pattern of 4 lb/gal silica flour slurries displacing base solutions of
(a) 30 lb/Mgal guar, (b) 5 lb/Mgal polyacrylamide, (c) 10 lb/Mgal Xanvis fluids.
1701

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

1702

E. M. El-M. Shokir and A. A. Al-Quraishi

change at the bottom of the frame where particle concentration is highest.

The extent of convection, in the experiments or the simulation, is indicated by
the slope of the interface between the slurry and the pad fluid at the middle of
the fracture model. In addition, this figure illustrates the similarity between
the simulator and physical experiments results.
Similarly, Figure 12 shows the experimental and simulation frames for
slurries of 4 lb/gal silica flour added to 30 lb/Mgal guar, 10 lb/Mgal Xanvis
and 5 lb/Mgal polyacrylamide solutions displacing base solutions. A good
match was seen except for density-driven fingering at the bottom of the cell
and the bypassing of the pad fluid at the lower left-hand corner in some of
the experiments that could not be matched with the simulator. The fingering
occurs in the experiments due to the bypassing of the pad fluid in the lower
left-hand corner of the glass model during initial injection. Since pad fluid
is less dense than the injected slurry, density-driven fingering can occur. The
extent of bypassing may increase for non-Newtonian fluids because of their
shear-thinning and gelling behavior. To reduce the effect of gelling, the shut
off time between injecting the pad and switching to the slurry was minimized.

5. CONCLUSIONS
1. Small glass models can simulate flow patterns in hydraulic fractures and
the flow patterns observed are strikingly similar to those obtained by
very large flow models used by previous investigators. With such small
model, wide variety of flow conditions can be investigated easily and
inexpensively.
2. Proppant transport and placement efficiency of different fracturing fluids
can be predicted on the basis of fluid physical properties, fracture dimensions, and injection rate.
3. As injection rates increases, convective settling decreases and proppant
placement efficiency increases, which is the perfect engineering requirement for proppant placement.
4. Convective settling increases as proppant concentration increases however,
proppant ramping could decreases this effect, as a field practice.
5. Larger injected volumes are required to prop the entire fracture height
with fluids that have lower (n) values than that of guar.
6. GOHFER simulator nearly quantitatively replicates the experimentally
observed proppant placement.

ACKNOWLEDGMENTS
The authors would like to express their deep appreciation to Saudi Arabian
Basic Industries Company (SABIC) for their financial support for this research work.

Proppant Placement in Hydraulic Fractures

1703

Downloaded By: [Shokir, eissa Mohamed] At: 10:43 21 August 2009

REFERENCES
Barree, R. D., and Conway, M. W. (1994). Experimental and numerical
modelling of convective proppant transport. SPE Paper 28564, 1994 SPE
Annual Technical Conference and Exhibition, September 2628, New
Orleans, LA.
Clark, P. E., and Courington, T. J. (1994). Visualization of flow into a
vertical fracture. SPE Paper 27692, 1994 SPE Permian Basin Oil and
Gas Recovery Conference, March 1618, Midland, TX.
Clark P. E., and Zhu, Q. (1994). Fluid flow into vertical fractures from a
point source. Paper SPE 28509, 1994 SPE Annual Technical Conference
and Exhibition, September 2527, New Orleans, LA.
Clark P. E., and Zhu, Q. (1995). Flow of Newtonian and non-Newtonian
fluids into a slot from a point source, developments in non-Newtonian
flows. J. Fed. 231:129134.
Cleary, M. P., and Fonseca, A., Jr. (1992). Proppant convection and encapsulation in hydraulic fracturing: Practical implications of computer and
laboratory simulation. SPE Paper 24825, 1992 SPE Annual Technical
Conference and Exhibition, October 47, Washington, DC.
Mobbas, A. T., and Hammond, P. S. (2001). Computer simulations of proppant transport in hydraulic fractures. SPE Prod. Facil. 19:112121.
Stim-Lab. (1995). Investigation of the rheology and proppant carrying capacity of common fracturing fluids. Stim-Lab (STIMLAB), Inc., a Core
Lab Production Enhancement Division, Technical Report. Duncan, Oklahoma.
Unwin, A. T., and Hammond, P. S. (1995). Computer simulation of proppant
transport in hydraulic fracture. SPE Paper 29649, 1995 SPE Western
Regional Meeting, March 810, Bakersfield, Canada.

NOMENCLATURE
no
L
H
Cn
Cv
Cvmax
a
Nvg
o
q

w
h
g

clean fluid power law flow behavior index

adjustable parameter to match onset of deviation of low shear
viscosity from the calculated power law viscosity
adjustable parameter to match onset of deviation of high shear
viscosity from the calculated power law viscosity
Cv /Cvmax
volume fraction of solids
maximum volume fraction of solids
slurry viscosity increase exponent
viscous to gravity ratio
clean fluid viscosity
flow rate
displaced and displacing fluids density difference
fracture width
fracture height, and
gravitational acceleration