SPE 115526

Fracture Ballooning in Naturally Fractured Formations: Mechanism and

Controlling Factors
R. Majidi, SPE, S.Z. Miska, SPE, M. Yu, SPE, and L.G. Thompson, SPE, University of Tulsa
Copyright 2008, Society of Petroleum Engineers

This paper was prepared for presentation at the 2008 SPE Annual Technical Conference and Exhibition held in Denver, Colorado, USA, 2124 September 2008.

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 have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper 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 SPE copyright.


Abstract
Fracture ballooning observed while drilling naturally fractured formations has often been mistakenly interpreted as influx of
formation fluid or the loss of drilling fluids. This misinterpretation leads to costly well control procedures that may make the
situation even worse. The main mechanisms and factors controlling the ballooning phenomenon must be well understood to
avoid confusing this phenomenon with conventional losses or formation kick. Amongst several mechanisms that are quoted
for borehole ballooning, the opening/ closing of natural fractures plays a major role in naturally fractured formations.
In this work, a mathematical model describing the fracture ballooning process is developed and solved numerically using
finite difference approximation. The governing equation is derived using principles of conservation of mass and linear
momentum for transient radial flow in a single fracture. The effects of fracture parameters (aperture, extension and
deformability) have been studied as well as fluid properties and operational conditions. Describing drilling fluid rheology
with Yield-Power-Law (Herschel-Bulkley/YPL) allows for the investigation of the effect of drilling fluid rheology on
borehole ballooning.
Results show how the rheological properties of drilling fluid such as yield stress and shear-thinning/thickening effect,
influence ballooning or mud losses in fractured formations. We conclude that the fluid loss in the fractures could be stopped
either because of high yield stress of drilling fluid or limited extension of the fracture. The proposed model is also helpful for
detecting and treating ballooning as well as evaluating fracture characteristics. The field potential application of the model is
described.

Introduction
Fracture deformation, or ballooning observed while drilling fractured formations is the result of loss/gain due to fractures
being opened and closed. Fracture opening/closing is caused by the annular pressure fluctuation at the wellbore resulting
from the change in circulation rate. Mud losses take place when drilling fluid at the well flows into the fracture because the
fracture is pressurized and opened. However, when the pumps are turned off such as during a connection the pressure at the
well will fall and the mud in the fracture will return to the well due to fracture closing. Usually any flow during drilling is
interpreted as an influx of the formation fluid and the common cure is to increase the mud weight and ensure an adequate
overbalance. But if the mud weight is increased and the influx is only mud return, the situation will get progressively worse
with a rise in equivalent circulating density (ECD). Mud losses will continue and the fracture propagation pressure may even
be exceeded, resulting in total losses. Therefore, it is very crucial to understand the major mechanisms and factors controlling
the ballooning phenomenon to avoid confusion with conventional losses or formation kick. Ward et al.
1
suggested that
diagnosis of downhole pressure response with a real-time PWD tool is very helpful to distinguish an influx from mud return.
Several examples of ballooning modeling are found in the literature. Lavrov and Tronvoll
2-5
considered mud loss into a
deformable fracture of finite length. Two different flow geometries, linear and radial flow were modeled separately. Possible
leakoff through the fracture wall caused by the pressure difference between the incoming mud and the formation fluid was
accounted for in the linear system. They also justified and used a linear fracture deformation. According to the theory of
Lavrov and Tronvoll, the eventual stop of mud losses is due to finite fracture extension. They later studied the incorporation
of non-Newtonian mud rheology, ballooning and associated mud loss into a single deformable, horizontal and circular
fracture intercepted by a borehole at its center. The effect of different types of fluid rheology, i.e. Newtonian, Power-Law
and bi-viscous fluid and various formation properties and operational conditions e.g., formation pressure, borehole pressure
and fracture dimensions, were discussed.
2 SPE 115526
Ozdemirtas et al.
6
has recently developed a two-dimensional flow model with an emphasis on the effect of fracture
roughness. They also performed experiments on several artificially fractured rock samples to study this effect. From the
results of experimental observations and numerical models they reported that there are critical situations in which the
roughness can cause the losses to diminish.
We have developed a model for fracture ballooning of YPL type of drilling fluids as a general rheological behavior case.

Mathematical Modeling
A horizontal disk-shaped fracture of outer radius r
e
located in an impermeable rock is considered in this study. Initially, the
fracture is fully saturated with liquid under the static formation pressure p
i
. Drilling fluid propagates into the fracture because
of the overbalance pressure at the well. All properties of the drilling fluid are assumed to be identical to the fluid that is
already in the fracture. Cylindrical coordinate r denotes the distance from the borehole axis in the fracture plane.

Fluid rheology
The fluid rheology is described by Yield-Power-Law. For YPL fluids, the shear stress is related to shear rate as,
m
y
k ) (
.
+ = Eq.1
where
y
is the yield stress, k and m are the consistency factor and flow behavior index of the fluid, respectively. Describing
drilling fluid rheology, (as described by the Yield-Power-Law model), allows for investigation of the effect of drilling fluid
rheological properties such as yield stress and shear-thinning on ballooning.

Fracture deformation
The major mechanism under which fracture ballooning takes place is the fracture deformation. Therefore, the model behavior
is critically dependent on the way that the fracture deformation is described. Hydro-mechanical models for fracture
deformation as a function of fluid pressure and mechanical stresses are beyond the scope of this work. A linear relationship
between fracture aperture and the fluid pressure inside the fracture is considered as
n
i
i
K
p p
w w

+ =

Eq.2
wherew is the width of the fractureK
n
is the normal fracture deformation modulus or fracture stiffness, and w
i
is the initial
fracture aperture at initial pressure p
i
before it is intercepted by the well. The assumed local linear relation between fluid
pressure and fracture aperture is a large simplification. For further development of the model, more realistic nonlinear
deformation laws such as the exponential law should be used for describing fracture deformation.

Governing equation
Conservation of mass for radial flow of an incompressible fluid in a variable width fracture is given by
0
1 ) (
=

+ +

t
w
v w
r r
v w
Eq.3
where v is the average radial velocity of the fluid in the fracture which varies with position and time. Fluid flow inside the
fracture can be described locally by momentum conservation; the average radial velocity for a YPL fluid at a distance r from
the borehole axis can be expressed as
7
:
m
y
m m
m
w m
m
dr
dp
k
w
m
m
v
1
1 1
1
1
1
)
2
)(
1
1 2
(
1
2
)
1 2
(

+
+

+
=
+
+

Eq.4
Substituting the average velocity from Eq.4 into Eq.3 we have

0 )
2
)(
1
1 2
(
1
2
1
)
1 2
( )
2
)(
1
1 2
(
2
1
)
1 2
(
1
1
2
1 1
1
1
1
2
1 1
1
=

+
+

+
+

+
+

+
+
+
+
+ t
w
w m
m
dr
dp
r
w
k
m
m
w m
m
dr
dp
w
r
k
m
m
m
y
m
m m
m
y
m
m m

Eq.5
Although the considered fracture deformation law is quite simple, Eq.5 is still nonlinear and needs to be solved
numerically. The resulting second order Partial Differential Equation (Eq.5) requires an initial condition and two boundary
conditions to be solved. Initially, the pressure within the fracture is uniform and equal to formation pressure. At time t=0, a
vertical borehole of radius r
w
intersects the fracture at its center. After being penetrated by the borehole, fluid pressure at
borehole suddenly increases from initial pressure p
i
to p
w
and stays constant. The fracture tip is assumed to act like a no-flow
boundary (dp/dr=0 at r=r
e
).
Equation 5 was solved implicitly using Newtons method with a variable fracture grid size and time-step (Appendix A).
The general solutions of Eq.5 in dimensionless coordinates are presented in the form of mud loss curves for the ballooning
process. These curves express the dimensionless volume or rate of losses/gains as a function of time for various fracture and
fluid properties.
SPE 115526 3
Results
Figure1 demonstrates the dimensionless pressure profile in the fracture at various times after it is intercepted by the wellbore.
When the fracture is encountered, a pressure disturbance is applied to the fracture from the wellbore. The pressure
disturbance propagates throughout the fracture until it reaches the tip of the fracture. As long as the pressure disturbance has
not reached the fracture tip the behavior of ballooning for various fracture extensions remains the same. However, when the
pressure disturbance reaches the fracture tip the incoming fluid pressurizes the fracture until the pressure throughout the
fracture becomes identical to the borehole pressure and as a result the flow stops. The eventual stop of losses depends upon
the extension of the fracture r
e
, fracture stiffness and the amount of overbalance pressure applied at the well. In the
equilibrium state the fracture aperture increases by (p
w
-p
i
)/k
n
. The ultimate volume of loss/gain can be calculated in terms of
fracture extension as
( )

=
n
i w
w u
K
P P
r r V
e
2 2

Eq.6
( )
nD
eD
uD
k
r
V
2
1
2

=

Eq.7
where V
u
and V
uD
are the ultimate volume of losses in real and dimensionless coordinates respectively. The ultimate volume
of losses is directly proportional to the pressure difference at the well and inversely proportional to the fracture stiffness. This
is the result of the linear relationship assumed to describe the fracture deformation. The ultimate volume of losses also
quadratically changes with the extension of the fracture. Figure 2 shows the dimensionless volume of loss/gain due to
fracture deformation of various fracture extensions. The smaller the fracture the sooner the mud losses will stop.

Effect of drilling fluid rheology
In addition to fracture properties such as fracture extension and fracture stiffness, drilling fluid properties also influence the
rate and volume of losses. Figure 3 shows the effect of yield stress of drilling fluid on fracture ballooning. Dimensionless
mud invasion factor is a measure of the yield stress of the drilling fluid. Fluid flow may stop merely because of high yield
stress of the fluid before it could be stopped due to the limited length of the fracture or even before the pressure disturbance
could be felt at the fracture tip. The higher the yield stress the sooner the losses will stop. It seems that the high yield stress
fluid flow in the fracture may act like a limited extension fracture. However, the cessation of mud losses due to the high yield
stress of the fluid is not as dramatic as it is due to limited fracture extension. Depending upon the relative value of fracture
and fluid parameters one may precede the other.
A typical example in real coordinates is given to investigate shear-thinning effect. It seems misleading to judge about
the shear-thinning effect of the fluid directly from the dimensionless type-curves since the dimensionless time t
D
is a function
of flow behavior index m. Figure 4 demonstrates the effect of shear-thinning of the drilling fluid on mud losses due to
ballooning. In this example, we used Power-Law fluids (
y
=0 or =0) with consistency factor k=0.5 pas.s
m
. A 1000
overpressure is considered across a fracture with stiffness of 8.910
3
MPa/m and aperture of 1 mm. According to Fig. 4, the
rate of losses for shear-thinning fluids is much greater than for thick fluids. The ultimate volume of fluid losses is still
controlled by facture extension (r
eD
=1000) in this case (no yield stress). We know that in general the losses could stop either
due to yield stress of the fluid or limited fracture extension.

Effect of fracture properties
Figures 5 and 6 show the effect of fracture stiffness (deformability) on ballooning process. The rate of loss/gain due to
ballooning declines when the fracture stiffness increases. The ultimate volume of losses is proportional to the reciprocal of
fracture stiffness. Knowing the effect of fracture stiffness, the effect of overbalance pressure could be investigated in the
same manner. According to the definition of dimensionless fracture stiffness (A-10) the higher overbalance pressure
increases the volume of losses exactly in the same manner that the fracture stiffness decreases. The rate and volume of losses
for various pressure differences at the well are shown in Figs. 7 and 8. The rate and volume of fluid losses for a YPL fluid
with m=0.8, k=0.5 lb/100ft
2
.s
m
and
y
=16 lb/100ft
2
in a very long (r
eD
=10,000) fracture of 1 mm aperture and stiffness of
3.4510
4
MPa/m is considered. Losses highly depend upon the applied pressure difference at the well. The ultimate volume
of losses linearly increases with increasing the overbalance pressure.
Figures 9 and 10 demonstrate how the fracture aperture influences the losses. The model predicts much higher losses in
wider fractures. According to Fig. 10 fracture aperture can change the ultimate volume of losses in situations that the losses
are controlled by yield stress of the fluid. However, if the fracture is of limited extension or the yield stress of the fluid is not
high enough to stop the losses before the disturbance reaches the fracture tip, the width of the fracture can not influence the
ultimate volume of losses. In this case, the ultimate loss volume is controlled by fracture extension while the fracture aperture
only influence initial rate of losses.

4 SPE 115526
Conclusions
Flow of a non-Newtonian drilling fluid from a borehole into a non-propagating fracture was modeled under the
assumptions of linear fracture deformation and Yield-Power-Law fluid rheology.

The presented model is very helpful for understanding the major mechanisms and factors controlling fracture
ballooning and to distinguish between this phenomenon and conventional losses or formation kick. This can avoid
unnecessary expensive well control procedures.

The presented model helps to investigate the effect of fracture parameters (extension, aperture and deformability)
and fluid properties such as yield stress and shear-thinning/thickening effect as well as operational condition
(overbalance pressure).

The volume of the fluid loss or gain due to ballooning into/out of the fracture is shown to be an almost linearly
decreasing function of the fracture stiffness. Shear-thinning effect of drilling fluids influence the rate of losses
particularly at the initial stages of losses, whereas, the ultimate volume of losses is controlled by fracture extension
and/or yield stress of the fluid. Mud losses may stop either because of limited extension of the fracture or high yield
stress of drilling fluid.

The model can be used to optimize the drilling regime in naturally fractured formations. It can also provide an aid
when using mud loss data for fracture characterization purposes when the ballooning is the major mechanism.

Acknowledgment
The authors wish to thank Tulsa University Drilling Research Projects (TUDRP) member companies for supporting this
work.


SI Metric Conversion Factors

bbl x 1.589 8 E01 =m
3

bpm x 2.649 667 E03 =m
3
s
-1

lb/100ft
2
x 0.478 803 E 00 =Pa
psi x 6.894 757 E+03 =Pa

Nomenclature
dr
D
= dimensionless radius differential
dt
D
= dimensionless time differential
m = flow behavior index
k = consistency factor, Pa.sec
m

p
w
= wellbore pressure, Pa
p
i
= initial formation pressure, Pa
q
D
= dimensionless mud loss flow rate
r
D
= dimensionless radius
r
w
= wellbore radius, m
t = time, s
t
D
= dimensionless time
v = radial velocity, ms
-1

V
D
= dimensionless mud loss volume
V
u
= ultimate mud loss volume, m
3

V
uD
= dimensionless ultimate mud loss volume
w = fracture aperture, m
= dimensionless mud invasion coefficient
=time factor, s
-1

.
=shear rate, s
-1

p = drilling overpressure, Pa

p
= drilling mud plastic viscosity, Pa.s
=shear stress, Pa

y
= drilling mud yield stress, Pa

SPE 115526 5
References
[1] Ward, C., Clark, R., Anatomy of ballooning borehole using PWD
TM
tool Presented at The Overpressure in Petroleum
Exploration workshop, Pau, France, 7-8 April , 1998.
[2] Lavrov, A., and Tronvoll, J .: Mud loss into a single fracture during drilling of petroleum wells: Modeling approach, in
Proceedings of the 6
th
International Conference on Analysis of Discontinuous Deformation, October 2003, 189.
[3] Lavrov, A., and Tronvoll, J ., Modeling Mud Loss in Fractured Formations, Paper SPE 88700 presented at the 11th Abu
Dhabi International Petroleum Exhibition and Conference, Abu Dhabi, U.A.E., 1013 October, 2004.
[4] Lavrov, A., and Tronvoll, J ., Mechanics of Borehole Ballooning in NaturallyFractured Formations Paper SPE 93747
presented at the 14
th
SPE Middle East Oil & Gas Show and Conference, Bahrain, 1215 Murch, 2005.
[5] Lavrov, Modeling Flow of a Biviscous Fluid from Borehole into Rock Fracture. SINTEF Petroleum Research, Journal of
Applied Mechanics, Vol.73 J an-2006.
[6] Ozdemirtas M., Babadagli T., and Kuru E., "Experimental and Numerical Modeling of Borehole Ballooning - Effect of
Fractures Roughness" Paper SPE 110121 presented at the 2000 SPE Annual Technical Conference and Exhibition,
Anaheim, California, 11-14 November 2007.
[7] R. Majidi, S.Z. Miska, M. Yu, and L.G. Thompson, Quantitative Analysis of Mud Losses in Naturally Fractured
Reservoirs: The Effect of Rheology, SPE 114130 the 2008 SPE Western Regional and Pacific Section AAPG J oint
Meeting held in Bakersfield, California, U.S.A., 31 March2 April 2008.
[8] Litard O., Unwin T., Guillot D., and Hodder M. H.: Fracture Width Logging While Drilling and Drilling Mud / Loss-
Circulation-Material Selection Guidelines in Naturally Fractured Reservoirs, SPEDC (December 2002) 237



Appendix A
Mathematical Modeling

To obtain a general solution for Eq.5 we use the following dimensionless variables.
Dimension distance from borehole axis:
w
D
r
r
r =

(A-1)
Dimensionless width of the fracture:
i
D
w
w
w =

(A-2)
where w
i
is the initial fracture width. Dimensionless pressure is defined as the pressure difference at any point in the fracture
divided by the total pressure difference at the well (over/under balance pressure).
i w
i
D
p p
p p
p

=

(A-3)

Therefore, the dimensionless form of the inner boundary condition at the well will be P
D
=1 at r
D
=1 and the outer boundary
condition will be
0 =

=
eD D
r r
D
D
r
p

(A-4)

The dimensionless mud invasion factor is defined as:

p w
r
m
m y
i
w

+
+
=

2
)
1
1 2
(

(A-5)
where is a measure of the ratio of the yield stress of the drilling fluid to the applied overbalance pressure. It represents the
relative value of the resistance to flow due to yield stress of drilling fluid to the driving force (overpressure).
Dimensionless time factor is defined as:
m
m
w
i
m
k
p
r
w
m
m
1
1
1
1
1
)
1 2
(
2
1

+
=
+
+

(A-6)
So dimensionless time is calculated as:
t t
D
=

(A-7)



6 SPE 115526
Using the dimensionless variables defined from Eq.A-1 to A-7 the dimensionless form of Eq.5 obtained as:
0
1 1
2
1
1
2
=

+
+
D
D
m
D D
D
D
m
D
m
D D
D m
D
D
t
w
w r
p
r
w
w r
p
w
r


(A-8)

where the dimensionless fracture aperture is a function of dimensionless pressure as shown below:
nD
D
D
K
p
w + =1

(A-9)
p
K w
K
n i
nD

=

(A-10)

K
nD
is the dimensionless fracture stiffness. Substituting w
D
from Eq.A-9 into Eq.A-7

0
1
1
1
1
1
1 1
2
1
1
2
=

+
+

+
+
D
D
nD
m
nD
D D
D
D
m
nD
D
m
nD
D D
D
m
nD
D
D
t
p
K
K
p r
p
r
K
p
K
p r
p
K
p
r


(A-11)

The dimensionless flow rate is obtained as
m
D D
D
D
m
D D
w r
p
r w q
1
1
2

=
+


(A-12)
where the dimensionless and actual flow rate are related as below.
q
p
k
w
r
m
m
r w
q
m m
i
w
w i
m
D
1 1
1
2
1
1
)
1 2
(
2
2

+
=
+ +

(A-13)
Dimensionless flow rate at the wellbore is
m
r
D D
D
m
D Dw
D
w r
p
w q
1
1
1
2
=
+

=


(A-14)

Dimensionless volume of losses is obtained from cumulative rate of losses at the well as shown below.

= =
D
t
D Dw
w i
D
dt q
r w
V
V
0
2
2

(A-15)
In order to numerically solve the resulting Partial Differential Equation (eq.A-7), one may discretise the fracture into a
number of intervals and solve implicitly in time. The discritised form of the Eq.A-7 is

( )
0
1
2
1
1
1
1
1
1
1
1
2 1 1
1
1
2
2
1
1
1
1
1
1
2
2
1
1
=

+ +
+
+

+ + +

+
+
+
+
+
D
n
Dj
n
Dj
nD
m
D Dm
n
Dj
n
Dj
Dj
m
Dj
D Db
n
Dj
n
Dj
m
Dj
m
D Df
n
Dj
n
Dj
m
Dj
D
t
p p
K w r
p p
r
w
w r
p p
w
w r
p p
w
r
m


(A-16)

where
nD
Dj Dj
Di
Dj
K
p p
w w
2
1
2
1
+
+
+
+ =

(A-17)
nD
Dj Dj
Di
Dj
K
p p
w w
2
1 1
2
1
+
+ =

(A-18)
Equation A-16 leads to a system of nonlinear equations that requires an iterative method to solve for pressure profile.
n n
f
f P J
D
=
+
) (
1

(A-19)
where J denotes the J acobean matrix of pressure distribution vector and f is the pressure function from Eq.A.16 which is
( )
0
1
1
1
1
1
2 1 1
1
1
2
2
1
1
1
1
1
1
2
2
1
1
1
=

+ =
+
+
+

+ + +

+
+
+
+
+
+ n
Dj
D Dm
n
Dj
n
Dj
Dj
m
Dj
D Db
n
Dj
n
Dj
m
Dj
m
D Df
n
Dj
n
Dj
m
Dj
D
nD n
Dj j
p
w r
p p
r
w
w r
p p
w
w r
p p
w
r
t K
p f
m


(A-20)
Solving the system of nonlinear equations by Newtons method, the pressure profile at any time can be obtained. Having
pressure profile, we can calculate mud flow rate at the well and also volume of fluid losses from eq.A-14 and eq.A-15.
SPE 115526 7

0
0.2
0.4
0.6
0.8
1
0 2000 4000 6000 8000 10000
r
D
P
D
tD=10^6
tD=5*10^6
tD=10^7
tD=2*10^7
r
eD
=10,000

Figure 1Pressure Profile in the fracture



1.E+0
1.E+1
1.E+2
1.E+3
1.E+4
1.E+5
1.E+6
1.E+7
1.E+8
1.E+0 1.E+1 1.E+2 1.E+3 1.E+4 1.E+5 1.E+6 1.E+7 1.E+8 1.E+9
t
D
V
D
r
eD
=10
4
r
eD
=10
3
r
eD
=10
2
r
eD
=10

Figure 2Volume of losses due to ballooning for various fracture extensions




1.E+0
1.E+1
1.E+2
1.E+3
1.E+4
1.E+5
1.E+6
1.E+7
1.E+8
1.E+0 1.E+1 1.E+2 1.E+3 1.E+4 1.E+5 1.E+6 1.E+7 1.E+8 1.E+9
t
D
V
D
=0
=0.0001
=0.01
=0.001

Figure 3Volume of losses due to ballooning for various fluid yield stress


8 SPE 115526
r
eD
=1000
1
10
100
1000
0.1 1.0 10.0 100.0 1000.0
ti me (mi n)
V
o
l
u
m
e

o
f

l
o
s
s
e
s

(
b
b
l
)
m=0.6
m=1
m=0.8

Figure 4Shear-thinning effect on mud losses



1.E+0
1.E+2
1.E+4
1.E+6
1.E+8
1.E+0 1.E+2 1.E+4 1.E+6 1.E+8
t
D
V
D
knD=0.5
knD=1
knD=2
r
eD
=10,000

Figure 5 The effect of fracture stiffness on volume losses




0
1
2
3
4
5
6
7
1.E+0 1.E+2 1.E+4 1.E+6 1.E+8
t
D
q
D
r
eD
=10,000
KnD=0.5
KnD=0.5
KnD=1

Figure 6Effect of fracture stiffness on rate losses







SPE 115526 9

r
eD
=10,000
0.001
0.01
0.1
1
10
100
0.0 100.0 200.0 300.0 400.0 500.0 600.0
ti me (mi n)
l
o
s
s

r
a
t
e

(
b
b
l
/
m
i
n
)
Dp=2000 psi
Dp=1000 psi
Dp=500 psi

Figure 7Rate of losses for various pressure difference



r
eD
=10,000
1
10
100
1000
10000
0.1 1.0 10.0 100.0 1000.0
ti me (mi n)
V
o
l
u
m
e

o
f

l
o
s
s
e
s

(
b
b
l
)
Dp=2000 psi
Dp=1000 psi
Dp=500 psi

Figure 8Volume of losses for various pressure difference




















r
eD
=10,000
1
10
100
1000
0.1 1.0 10.0 100.0 1000.0
ti me (mi n)
V
o
l
u
m
e

o
f

l
o
s
s
e
s

(
b
b
l
)
w=0.5
w=1
w=2

Figure 9Rate of losses for various fracture apertures



r
eD
=10,000
0.01
0.1
1
10
100
0.0 100.0 200.0 300.0 400.0 500.0 600.0
ti me (mi n)
l
o
s
s

r
a
t
e

(
b
b
l
/
m
i
n
)
w=2
w=0.5
w=1

Figure 10Volume of losses for various fracture apertures