Fuel 119 (2014) 27–37

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Optimization of hydraulic fracture geometry in gas condensate

reservoirs
H. Mahdiyar a, M. Jamiolahmady b,⇑
a
Shiraz University, Shiraz, Iran
b
Heriot-Watt University, Edinburgh, EH14 4AS, Scotland, United Kingdom

h i g h l i g h t s

 Proposed a new optimization method for hydraulic fracture dimensions.

 It is general and applicable to gas and gas condensate systems.
 Discussed limitations of available methodologies for non-Darcy flow systems.
 Demonstrated superiority of our proposed formulation through various examples.

a r t i c l e i n f o a b s t r a c t

Article history: An optimized design for hydraulic fracturing is of great importance especially with the growing demand
Received 12 December 2011 for this method as a means of production enhancement from unconventional gas reservoirs. The first
Received in revised form 6 November 2013 Optimum Fracture Design (OFD) approach, which maximizes well productivity for a given fracture vol-
Accepted 7 November 2013
ume, was introduced by Prats in 1960 for single-phase Darcy flow systems. This was then further devel-
Available online 23 November 2013
oped and presented in the form of Unified Fractured Design (UFD) charts by Valko et al. (1998), which is
applicable to Pseudo-steady state conditions. Later on, some methodologies have been proposed to make
Keywords:
UFD applicable to gas condensate systems assuming the distribution of the condensate phase around the
Hydraulic fracturing
Gas condensate
fracture as a rectangular damage zone with constant thickness and reduced permeability. These latter
Non-Darcy methods are generally oversimplified as they neglect different possible shapes of the two phase region
Coupling (Capillary number) effect around the fracture and the variation of relative permeability with interfacial tension (IFT) and velocity
Optimum design for these low IFT systems. They also require data that are not readily available, in particular the pressure
profile (required to identify the two-phase boundary) around the wellbore.
In this paper, we introduce an explicit formulation and a more general methodology for OFD that is
applicable to both Steady state and Pseudo-steady state single-phase gas and two-phase gas condensate
flow systems and includes the important flow parameters in both the matrix and fracture. The optimum
fracture dimensions are obtained by maximizing the effective wellbore radius, using the recently devel-
oped correlation by Mahdiyar et al. (2011). This formulation accounts for the mechanical and flow skins
based on quite readily available information at wellbore conditions.
The integrity of the introduced formulation has been verified for many different prevailing conditions,
whilst highlighting the errors of using conventional approaches with some important practical guide-
lines. In this exercise, the maximum productivity calculated using the proposed formulation is compared
with results of the literature or our in-house simulator. This program, using a fine grid approach, simu-
lates gas condensate flow around a hydraulically fractured well for various fracture length–width ratios
and identifies the optimum fracture dimensions, for a given fracture volume, providing maximum mass
flow rate.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Optimization of a hydraulic fracture geometry provides the

⇑ Corresponding author. Address: Heriot-Watt University, Institute of Petroleum
Engineering, Riccarton, Edinburgh, EH14 4AS, United Kingdom. Tel.: +44 131
4513122; fax +44 131 4513127.
E-mail address: jami.ahmady@pet.hw.ac.uk (M. Jamiolahmady).
maximum productivity/injectivity of a hydraulically fractured well
(HFW) for a fixed fracture volume.
Prats [18] was the first to introduce the concept of the optimum
fracture geometry. According to his results, optimum fracture

0016-2361/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.fuel.2013.11.015
28 H. Mahdiyar, M. Jamiolahmady / Fuel 119 (2014) 27–37

Nomenclature

A a parameter, showing the effect of fracture penetration b single-phase inertial factor

ratio on effective wellbore radius at Steady state l viscosity
B a parameter, showing the effect of fracture penetration q density
ratio on effective wellbore radius at Pseudo-steady state W pseudo pressure
(a,b,c)SS parameters of FSS l average viscosity based on GTR
(a,b,c)PSS parameters of FPSS q average density based on GTR
C a constant in JD expression, Eq. (2), which is 1/2 for Stea-
dy state systems and 3/4 for Pseudo-steady state Subscripts
CfD bsolute fracture conductivity b base
CfD-eff effective fracture conductivity c condensate
fA function varying with parameter A. D dimensionless
fB function varying with parameter B. e external as in re
FSS Steady state function eff effective
FPSS Pseudo-steady state function f fracture
h formation thickness g gas
IX fracture penetration ratio, xf/xe m matrix
JD dimensionless productivity index opt optimum
k absolute reservoir permeability r relative
kf absolute fracture permeability w wellbore
kr relative permeability
m mass flow rate Abbreviations
P pressure
C1 methane
q volumetric flow rate at bottom-hole conditions C4 normal butane
r radius C10 normal decane
r 0w effective well bore radius EOH equivalent Open-hole
Re Reynolds number
EOS equation of state
S0f pseudo fracture skin factor GTR gas to total (gas plus condensate) flow rate ratio.
Vf facture volume per unit height of the fracture IFT interfacial tension
Ve drainage volume per unit height of the fracture
HFW hydraulically fractured well
wf fracture width HF hydraulic fracture
xe half length of the square drainage area HFWS hydraulically fractured well system
xf half length of the fracture OH open-hole
M mass mobility
PR3 3 parameter Peng Robinson EOS
MR mass mobility ratio SS Steady state
PSS Pseudo-steady state
Symbols
d a parameter defined in Eqs. (4)–(6)

design for a HFW in a square drainage area under the Darcy flow
regime is obtained when dimensionless fracture conductivity, that
is the ratio of flow ability of the fracture (the permeability-width
product) over that of the matrix (the permeability-fracture-half
length product), is 1.26.
Valko et al. [20] presented an optimization approach called Uni-
fied Fracture Design (UFD). They emphasize that ‘‘the key to formu-
lating a meaningful technical optimization problem is to realize that
penetration and dimensionless fracture conductivity are competing
for the same source: the propped volume’’. In the UFD method, the
propped number is introduced as two times of the ratio of the
propped volume to the reservoir volume, weighted by their perme-
ability contrast. Fracture conductivity is defined as the ratio of ability
of the fracture to pass flow to the wellbore to that of the matrix to
pass it to the fracture, i.e. the ratio of fracture permeability, width
product to matrix permeability, fracture half length product. Their
charts present the dimensionless productivity index (a measure of
well deliverability) of Hydraulically Fractured Wells (HFWs) at Pseu-
do-steady state (PSS) as a function of dimensionless propped num-
ber and fracture conductivity. In these graphs, it is clearly shown
that for each propped number there is an optimum fracture conduc-
tivity at which the productivity index has the maximum value.
Economides et al. [1] in discussing the optimal design stated
that ‘‘. . .what is good for maximizing PSS flow is also good for max-
imizing transient flow’’.
A semi-analytical formula for estimation of effective wellbore
radius of a HFW in a rectangular closed drainage area for single-
phase Darcy flow at PSS was developed by Meyer and Jacot [16].
They also presented a chart, which correlates the optimum fracture
conductivity, for drainage area with rectangular aspect ratio, with a
restriction for fracture penetration ratio; i.e. the chart is applicable
for HFWs with fracture penetration ratio less than 0.2. According to
the results of Meyer and Jacot [16] and also Valko et al. [20], opti-
mum fracture conductivity in square drainage areas (with Ix less
than 0.2) is about 1.57, which is a little higher than the Prats sug-
gested value of 1.26.
The optimization of hydraulic fracture in a non-Darcy flow sys-
tem has also been the subject of study by some investigators. Lo-
pez-Hernandez et al. [13] introduced the concept of using the
effective propped number in the UFD method for estimation of
optimum fracture length for non-Darcy flow systems. In this con-
cept, as inertia reduces absolute fracture permeability, effective
permeability should be used in the calculation of propped number
and fracture conductivity.
There are also publications available in the literatures [21,8,17]
on optimization of fracture design for gas condensate systems.
However, in all these studies gas condensate flow has been treated
as a single phase flow with a fracture face damage related to a con-
stant thickness two phase region around the whole fracture. For in-
stance, Wang et al. [21] used the Cinco-Ley skin equation,
 H. Mahdiyar, M. Jamiolahmady / Fuel 119 (2014) 27–37 29

developed for estimation of fracture face damage, for calculation of

condensate build-up damage. In this approach, the absolute
permeability of the damaged layer is set equal to kkrg and the
thickness of the damaged layer is replaced with the extent of
the two-phase region, which-it is assumed-is a constant layer
around the fracture. For estimation of two-phase region thickness,
Wang et al. [21] recommend that: ‘‘For any fracture length and a
given flowing bottom-hole pressure that is known to be inside
the retrograde condensation zone of a two-phase envelope, the
pressure profile normal to the fracture face and into the reservoir
will delineate the points where the pressure is equal to the dew
point pressure. From this pressure profile the fracture face skin Fig. 1. An HFW located in the center of square or circular drainage area.
distribution along the fracture face is determined.’’ This oversim-
plified approach, which also requires information not easily
available, i.e. pressure profile within the reservoir, could result in For this geometry, fracture volume per unit height of the frac-
very rough estimations of gas condensate flow characteristics. ture is calculated by following equation:
Flow of gas and condensate has been shown to be affected by
V f ¼ 2xf wf ; ð1Þ
two different velocity effects. While the coupling effect increases
relative permeability with the increase in velocity (Danesh et al. where xf is the fracture half length and wf is the fracture width. Lar-
[6,9,10] and/or the decrease in interfacial tension [22], the iner- ger fractures can result in larger productions but the cost of hydrau-
tial effect decreases relative permeability [7,11]. These effects lic fracturing increases with the increase in the fracture volume.
have been missed in the Wang et al. [21] study. Another limita- Actual hydraulic fractures are very complex flow geometries to
tion of their approach is neglecting two-phase flow inside the replicate, and hence, to the best of authors’ knowledge, the above
fracture. assumptions are the basis of most similar studies in this area
To the best of the authors’ knowledge, a method for optimiza- [24,18,2,23,20,4,3,19,5].
tion of an HF design in gas condensate reservoirs, correctly Technically, for a certain fracture volume maximum productiv-
accounting for two-phase flow parameters including coupling ity can be achieved with an optimum fracture length/width ratio.
and inertial effects, has not been previously reported. In other words, optimum fracture design is defined as the geome-
Mahdiyar et al. [14] investigated the two-phase region around a try of a given fracture volume that provides the maximum dimen-
HFW in gas condensate reservoirs and demonstrated that it is ellip- sionless productivity (JD). For an hydraulically fractured well,
tical and becomes radial when the effective fracture conductivity located in the center of an undamaged cylindrical or square drain-
approaches zero. They showed that the extent of two-phase region age area. JD is defined by:
into the reservoir could be equal to, smaller than or larger than the
1 1 1
fracture length depending on the operating flow conditions. They JD ¼   ¼   ¼   ; ð2Þ
also developed a general formula for the effective wellbore radius ln re
rw
þ Sf  c ln re
xf
þ S0f  c ln re
r 0w
c
(r w 0) of an equivalent open-hole system, replicating flow around a
HFW. It was demonstrated that the value of r0w depends on the where Sf and S0f are the fracture and pseudo fracture skin, respec-
exterior boundary condition, i.e., it is different for Steady state tively. re is the exterior radius of the circular drainage area and r 0w
(SS) constant pressure and Pseudo-steady state (PSS) no flow outer is the effective wellbore radius. The constant c attains the value of
boundary conditions. 1/2 for Steady state (SS) systems and 3/4 for Pseudo-steady state
Here the formulae proposed by Mahdiyar et al. [14] for effec- (PSS) systems. For a hydraulic fracture that performs better than
tive wellbore radius have been used to develop two sets of for- the un-fractured open hole system, Sf is negative but S0f is always
mulae for estimation of the optimum fracture conductivity and positive. Based on this equation, it can be stated that maximum JD
the optimum fracture half-length for SS and PSS. These formulae, is achieved when S0f is at minimum or r0w is at maximum. That is,
which are applicable for oil, gas and gas condensate square reser- for the optimization of fracture design, the following equation
voirs, are developed mathematically, based on the maximization should be solved:

of an effective wellbore radius for a fixed fracture volume. The @r 0w
applicability of these formulae has been verified, for several dif- ¼ 0: ð3Þ
@xf V f ;P w
ferent HFW illustrations operating at different prevailing
conditions. According to this equation, we are looking for a HF design,
which, for a given Vf and at a known bottom-hole pressure, pro-
vides the maximum possible r 0w :
2. Optimum fracture design To obtain the conditions at which Eq. (3) is satisfied, we borrow
the corresponding equation for r 0w ; Eq. (4), developed by Mahdiyar
Fig. 1 shows an ideal hydraulically fractured well located in the et al. [14]:
center of a cylindrical reservoir. The basic assumptions of this xf
study are: r0w ¼ ; ð4Þ
2dþC p
fDeff

a) The fracture and matrix are two different porous media but where
each one is a uniform porous medium.
 
b) The width of the fracture is constant. 0:17
c) Flow to the wellbore directly from the matrix is negligible
for SS systems : d ¼ A ¼ ln e1  ð5Þ
reD  0:87
compared to that from the fracture.
!
d) The fracture has penetrated symmetrically in both directions 0:504
1
and vertically through the whole height of the reservoir. and for PSS systems : d ¼ B ¼ ln e þ : ð6Þ
e) Gravity force is neglected in this 2D system.
ðreD  0:783Þ1:255
30 H. Mahdiyar, M. Jamiolahmady / Fuel 119 (2014) 27–37

In Eq. (6), reD is the dimensionless outer radius. In the case of a As described in Appendix A, solving Eq. (3), using Eqs. (4)–(13),
square drainage area, this value can be calculated as the result of results in the following expression:
equating the circular area with the square area to give:
Fðxf Þ ¼ a  x3f þ b  x2f þ c: ð14Þ
re 2 1
r eD ¼ ¼ pffiffiffiffi ; ð7Þ
xf p IX Optimum fracture half-length can be estimated when F(xf) is equal
where IX is the fracture penetration ratio. In a square drainage area to zero. The values of a, b and c are included in Table 1. As shown in
with the half length equal to xe is defined as follows: this table a, b have different formulae at SS and PSS conditions be-
cause the value of d, in Eq. (4), has different expressions at SS (Eq.
xf
IX ¼ : ð8Þ (5)) and PSS (Eq. (6)) conditions. When none of the roots of Eq.
xe (14) is between zero and reservoir half length, the plot of F(xf) ver-
In Eq. (4) the formula for effective fracture conductivity (CfDeff) sus IX does not intersect the x-axis (i.e. xf), and optimum practical
is defined as follows: fracture design is achieved at IX = 1.
Eq. (14) can be solved to obtain optimum absolute and effective
1
C fDeff ¼ C fD  MRw  ; ð9Þ fracture conductivity as expressed by Eqs. (15) and (16),
1 þ 0:62Rew
respectively.
where CfD, MRw and Rew refer to (absolute) fracture conductivity, ra-
tio of relative mobility and Reynolds number, respectively, and the
p
C fD;opt ¼ ð1 þ 2  0:62Rew Þ ð15Þ
subscript w denotes the value of quantity for the wellbore 2f 1 MRw
conditions.  
Absolute fracture conductivity, which is defined as the absolute p 0:62Rew
C fDeff ;opt ¼ 1þ ð16Þ
flow conductivity in the fracture over that in the matrix, is calcu- 2f 1 1 þ 0:62Rew
lated as follows:
The value of f1, can be found in Table 1. As seen from Eq. (16)
kf wf optimum effective fracture conductivity is independent of relative
C fD ¼ ; ð10Þ
kxf mobility ratio.
Mahdiyar et al. [15] demonstrated that S0f at PSS is greater than
where kf and k refer to fracture and matrix permeability,
that at SS, and that the difference is more significant when IX ap-
respectively.
proaches one, especially for higher conductive fractures. Moreover,
The ratio of relative mobility at the wellbore conditions can be
the shape of variation of S0f with IX is different in these two systems.
estimated by the following equation:
That is, with an increase in IX: S0f jPSS increases while S0f jSS decreases.
 
 qg qc It should be noted that HFW performance under SS and PSS condi-
Mf  w lg krgbf þ lc krcbf w
MRw ¼ ¼ q  : ð11Þ tions is similar if IX is less than 0.2. They also showed that when the
M m jw g qc
l krgm þ l krcm g c w
fracture is penetrated in a reservoir with a constant exterior pres-
sure boundary the performance approaches SS conditions and the
In Eq. (11) kr, q and l are the relative permeability, density and well productivity, even at the late transient state, can be estimated
viscosity and the subscripts g, c, w, f and m refer to the gas, conden- with the appropriate SS effective wellbore radius (r0w ) formula.
sate, wellbore, fracture and matrix, respectively. krgb and krcb refer Likewise, when the fracture is penetrated in a reservoir with no
to the base gas and condensate relative permeabilities (conven- flow boundary conditions the performance approaches PSS condi-
tional measurements), which are not a function of velocity and tions and the well productivity, even at the late transient state, can
interfacial tension [12]. In the case of a single-phase flow of gas, be estimated with the appropriate PSS effective wellbore radius
krc = 0 and krg = 1.0. Hence, MRw = 1.0 and CfDeff only accounts for formula.
the inertial (non-Darcy) effect based on Reynolds number (Re) Figs. 2 and 3 show the plot of optimum absolute fracture con-
for single-phase flow of gas. ductivity (CfD,opt) of a HFW at SS and PSS conditions, respectively,
In Eq. (9), Reynolds number at the wellbore conditions is de- for single phase systems (MRw = 1.0) as a function of IX and Rey-
fined as follows: nolds number. The variations of optimum effective fracture con-
q w v w kf bf ðkrgbf þ krcbf Þw ductivity (CfDeff,opt) with fracture penetration ratio, at SS and PSS,
Rew ¼ ð12Þ are shown in Figs. 4 and 5, respectively.
l w
Here b refers to the single-phase inertial factor. The average
density and viscosity values are estimated based on the gas frac- 100
tional flow (GTR = qg/(qg + qc)) as follows:
Ix=0.9
q ¼ GTR  qg þ ð1  GTRÞqc
: Ix=1
l ¼ GTR  lg þ ð1  GTRÞlc Ix=0.7
C fD,opt

10 Ix=.5
Ix=0.0001
Table 1
Formulae of a, b, c and f1 for the use in Eqs. (14)–(16).

Steady state (SS) Pseudo steady state (PSS)

0:62pC 2 qtw pC 2 qtw
a aSS ¼ fA MRw
aPSS ¼ 0:62
fB MRw
p p 1
b bSS ¼ 2fA MRw bPSS ¼ 2fB MRw 0.1 1 10 100
c c¼
kf V f
 2k c¼
kf V f
 2k
Re
f1 fA = 14.592A2  29.166A + 15.573 fB = 2.9905B2 + 5.3765B  1.3789
kf bf Fig. 2. Optimum fracture conductivity, for single phase flow systems, as a function
C 2 ¼ ðkrgbf þ krcbf Þw lq ww

hV f
.
of Reynolds number and fracture penetration ratio, Steady state conditions.
 H. Mahdiyar, M. Jamiolahmady / Fuel 119 (2014) 27–37 31

1000 25
Ix=0.95

Ix=0.9 20
Ix=0.95
100 Ix=0.8
15

C fD-eff,opt
C fD,opt

Ix=0.7
Ix=0.9
10
Ix=0.5
10
Ix=0.8
5
Ix=0.0001 Ix=0.7
Ix=0.5
Ix=0.0001
0
1 0.1 1 10 100
0.1 1 10 100
Re
Re
Fig. 5. Optimum effective fracture conductivity as a function of Reynolds number
Fig. 3. Optimum fracture conductivity, for single phase systems, as a function of and fracture penetration ratio, Pseudo-steady state conditions.
Reynolds number and fracture penetration ratio, Pseudo-steady state conditions.

optimum fracture design, production rate and fracture volume. It

From Figs. 2 and 4, it is seen that for HFW systems with
constant exterior pressure (SS conditions), both CfD,opt and CfDeff,opt
decrease with increasing IX. This decrease becomes more signifi-
cant when IX approaches one. It is also seen that for SS conditions
CfDeff,opt varies between 0.785 (for IX = 1.0 and very small Reynolds
numbers) and p, at infinite Reynolds number and small IX.
Figs. 3 and 5 show that for HFW systems with a closed exterior
boundary (PSS conditions), both CfD,opt and CfDeff,opt increase with
the increase in IX especially when IX approaches one. This trend
is opposite to that observed in Figs. 2 and 4 for SS systems. In fact,
the top curves in Figs. 2 and 4 are almost the same as the bottom
curves in Figs. 3 and 5, respectively. According to Fig. 5, CfDeff,opt
for PSS could be in the range of p/2 (for Darcy flow with small frac-
ture penetration ratio), and 22.6 (for a very high Reynolds number
and IX = 0.95).
should be noted that fluid properties are obtained from available
literature data, correlations or a tuned equation of State EOS mod-
el. Owing to the implicit dependency of the variables, the proce-
dure for solving the design problem is an iterative method and
depends on which variables are known. In Appendix B, the corre-
sponding procedures for solving three possible types of optimum
fracture design problems are described.
In here to verify the applicability of the newly developed meth-
od and to evaluate its accuracy, the optimized geometries of differ-
ent Hydraulic Fractures (HF) are discussed. The results are
compared with the results available in the literature or those ob-
tained using our in-house simulator. The specifications of the cases
under study in this section (referred to as HFWS-O1 to O6) are de-
scribed in Tables 2 and 3. As mentioned above the solution proce-
dure for different optimization problems are given in Appendix B.

3. Illustrations
3.1. Illustration 1: single-phase Darcy flow, small HF
The method introduced in this paper helps production/reservoir
engineers to design the optimum fracture geometry for the desired For this HF (referred to as HFWS-O1 in Table 2), the well pro-
operating conditions. In a fracture design problem, having identi- duction rate is 106 m3/s, which does not result in the non-Darcy
fied the physical properties of the fracture and matrix (k, b, krb),
fluid properties (l, q, GTR, . . .) and average (or external) pressure
Table 2
of the drainage area, there are five variables remaining: fracture
Parameters of hydraulic fractured wells, optimized in the illustrations 1–3, single-
length (xf), fracture thickness (wf), fracture volume (Vf), production phase flow systems.
_ and drawdown ðDP ¼ P
rate (m)   PW Þ. Among these five variables
HFWS-O1 HFWS-O2 HFWS-O3
just two are independent variables and the other variables are cal-
culated using the available equations. These are the equations of Matrix core Texas cream Texas cream Texas cream
k = 9.1 mD k = 9.1 mD k = 9.1 mD
b = 3.927E+9 b = 3.927E+9 b = 3.927E+9
Fluid kinematic l 2 l 2 l 2
3.5 q ¼ 1E  7 m =s q ¼ 1E  7 m =s q ¼ 1E  7 m =s
viscosity
kf (D) 146 146 146
3
bf (m1) 3.511E+5 3.511E+5 3.511E+5
Ix=0.5 Ix=0.7 Reservoir half length 500 500 500
2.5 Ix=0.0001
Formation thickness 30 30 30
Ix=0.9
(m)
C fD-eff,opt

2
Propped volume 0.625 62.5 62.5
Ix=1 (m3/m)
1.5
Absolute propped 0.02 2.00 2.00
number
1
Wellbore production 1E06 1E06 0.05
rate (m3/s)
0.5
PSS optimum Ix 0.1128 0.819 0.478
PSS optimum CfD 1.57 2.99 8.8
0
0.1 1 10 100 PSS optimum 1.57 2.99 2.9
CfDeff
Re SS optimum IX 0.1128 1.0 0.504
SS optimum CfD 1.57 2.0 7.9
Fig. 4. Optimum effective fracture conductivity as a function of Reynolds number SS optimum CfDeff 1.57 2.0 2.5
and fracture penetration ratio, Steady state conditions.
32 H. Mahdiyar, M. Jamiolahmady / Fuel 119 (2014) 27–37

Table 3 180
Parameters of hydraulic fractured wells, optimized in the illustrations 4–6, gas
condensate flow systems.
(rw-eff)SS
140
HFWS-O4 HFWS-O5 HFWS-O6

rw-eff or F(xf)/1000
Matrix core Texas cream Texas cream Texas cream
k = 9.1 mD k = 9.1 mD k = 9.1 mD 100 UFD Method
b = 3.927E+9 b = 3.927E+9 b = 3.927E+9 (rw-eff)PSS
(F(xf)/1000)PSS
Fluid C1–C10 C1–C10 C1–C10
(ZC1 = 80%) (ZC1 = 80%) (ZC1 = 80%) 60
T = 500 K, T = 500 K, T = 500 K,
PD = 3541.5 psia PD = 3541.5 psia PD = 3541.5 psia
kf (D) 146 146 146 20
bf (m1) 3.511E+5 3.511E+5 3.511E+5
Reservoir half length 250 250 250 0 0.2 0.4 0.6 0.8 1
Formation thickness 30 30 30 -20
(m) IX
Propped volume 6.25 6.25 6.25
(m3/m) Fig. 7. Variation of F(xf) and effective wellbore radius with IX for the case referred to
Absolute propped 0.8 0.8 0.8 as HFWS-O2 in Table 2. For Pseudo-steady state (PSS), optimum fracture penetra-
number tion ratio is 0.819, where FPSS(xf) becomes zero, but for Steady state (SS) it is 1.0 and
Optimum 15.97 80.9 133.16 FSS(xf) does not intersect the x-axis, Single-phase Darcy flow.
production rate
(m3/h) reported in the literature [16] for a single-phase Darcy flow system.
Pw/ (psi) 3486.5 3486.5 3486.5
The effective optimum fracture conductivity (CfDeff,opt), is also the
Pe (psi) 3536.5 3641.5 3741.5
GTRw 0.885 0.885 0.803
same as CfD,opt (1.569), confirming that inertia is negligible.
MRw 1.25 0.75 0.803
SS optimum IX 0.585 0.32 0.27 3.2. Illustration 2: single-phase Darcy flow, big HF
SS optimum CfD 2.35 7.71 11
SS optimum CfDeff 1.98 2.46 3
In this system (referred to as HFWS-O2 in Table 2), there is still
SS optimum 0.78 2.17 3
Reynolds number Darcy flow inside the fracture; however, the fracture volume, and
consequently the propped number, have been increased 100 times:
Np = 2.0. Fig. 7 shows the optimization curves of this system. Obvi-
ously, where IX is greater than 0.2, the performances of the HFW at
inertial effect inside the fracture. The propped number of this sys-
SS and PSS are different [15]. For SS, optimum fracture length is
tem, which is defined by the following equation [20], is 0.02:
equal to the reservoir length, while for PSS, optimum fracture pen-
etration ratio is 0.82.
kf wf xf kf V f
NP ¼ I2X C fD ¼ ¼2 ; ð17Þ The importance of using the correct formulation in optimization
k x2e k Ve
is clearly observed from Fig. 7. For instance, if for a constant exte-
rior pressure (i.e. SS conditions), PSS optimum point is incorrectly
Mahdiyar et al. [15] showed that for small propped numbers the
selected, the effective wellbore radius (r0w ) will be 17% less than the
difference between the performance of HFWs at SS and that at
maximum possible value.
PSS is not significant. Fig. 6 shows the result of optimization of this
The optimum point estimated using the UFD method [20],
HF (referred to as HFWS-O1 in Table 2). In this figure, effective well-
which has been developed for PSS conditions, is also shown in
bore radii (Eq. (4)) and F(xf) (Eq. (14)) versus fracture penetration
Fig. 7. According to this method for PSS, optimum IX is 0.73. This
ratio have been plotted for both Steady state (SS) and Pseudo-
is about 10.6% less than the optimum IX predicted by our new
steady state (PSS) systems. Optimum fracture penetration ratio for
method. As the variation of r0w with IX around the optimum point
both cases is 0.113 and, because of the small propped number
is not very considerable, the difference between the optimum
(Np = 0.02), the curves of both systems overlap. For both systems
PSS wellbore radii calculated by these two methods is not as signif-
at optimum point, Reynolds number is 0.0015 and optimum
icant (AD% = 6%).
fracture conductivity (CfD,opt) is 1.57, which is similar to that
3.3. Illustration 3: single-phase non-Darcy inertial flow, big HF
20
The characteristics of this system (HFWS-O3 referred to as in
Table 2), are almost the same as the last system (HFWS-O2), but
15
the flow rate is much greater (0.05 m3/s), enabling us to observe
(F(xf)/1000)SS
rw-eff or F(xf)/1000

the effect of inertia on the optimization results. Fig. 8 shows the

10 (rw-eff)PSS optimization curves of this system. For PSS, optimum fracture pen-
(rw-eff)SS etration ratio, absolute fracture conductivity and effective fracture
5
conductivity are 0.48, 8.8 and 2.9, respectively, whilst for SS the
(F(xf)/1000)PSS corresponding values of these parameters are 0.50, 7.9 and 2.5,
respectively.
0 A comparison of the results of illustrations 2 and 3 clearly
0 0.2 0.4 0.6 0.8 1
shows how inertia decreases the optimum length of the fracture
-5 for a fixed fracture volume. This is consistent with our previous
IX finding [15] that for the same applied pressure drop, inertia is more
pronounced for longer fractures and it is reduced if fracture width
Fig. 6. Variation of F(xf) and effective wellbore radius with IX for the case referred to
as HFWS-O1 in Table 2. Optimum fracture penetration ratio is 0.113 where F(xf)
is increased. In this example the optimum fracture length in the
becomes zero and effective wellbore radius has its maximum value, Steady state presence of inertia is about 50% shorter than that discussed above
(SS) and Pseudo-steady state (PSS) conditions, Single-phase Darcy flow. without the inertial effect.
 H. Mahdiyar, M. Jamiolahmady / Fuel 119 (2014) 27–37
100
(rw-eff)SS
rw-eff or F(xf)/1000
(rw-eff)PSS
60
(F(xf)/1000)PSS
20 (F(xf)/1000)SS
0 0.2 0.4 0.6
-20
IX
Fig. 8. Variation of F(xf) and effective wellbore radius with IX for the case referred to
as HFWS-O3 in Table 2. For Pseudo-steady state (PSS) and Steady state (SS),
optimum fracture penetration ratio is 0.48 and 0.50, respectively, Single-phase
inertial flow.
3.4. Illustration 4: gas condensate flow, HF with two-phase region in
the whole drainage area
In order to verify the applicability of this newly developed opti-
mization method for gas condensate flow systems, the geometry of
an HF was optimized for HFWS-O4 (see Table 3). In this system, the
fluid is a mixture of C1–C10 ðZ C 1 ¼ 80%Þ. The fluid properties have
been estimated using PVTi software with Three-parameter Peng
Robinson (PR3) equation of state (EOS). The bottom-hole pressure
and exterior pressure are 3486.5 psi and 3536.5 psi, which are 55
and 5 psi below the dew point pressure, respectively; thus there is
a two-phase gas condensate flow within the whole system. Fig. 9
shows the optimization results of this HF by three approaches:
plotting the (i) calculated production rate, (ii) effective wellbore ra-
dius and iii) FSS(xf), all versus fracture penetration ratio. The use of
these three approaches that are used for the following cases, are
aimed at a better demonstration of the integrity of our proposed
formulation for the more complex case of two-phase flow of gas
and condensate.
It should be noted that the production rate varies depending on
the drawdown and the fracture dimensions. The production rate
curve versus IX curve, for a fixed fracture volume but different frac-
ture penetration ratios shown in this figure, has been estimated
using our in-house 2D HFW simulator. This computer program
simulates the SS flow of gas and condensate around a HFW, includ-
ing the phase change and the impact of velocity and IFT on relative
permeability. The generalized relative permeability correlation
expressing the combined effect of coupling and inertia developed
by Jamiolahmady et al. [12] has been incorporated in this model.
It also automatically identifies the optimum fracture dimensions,
for a given fracture volume, providing maximum mass flow rate.
The description of governing equations, the solution technique
and its verification can be found elsewhere [14].
In Fig. 9 the curves of FSS (xf) and effective wellbore radius have
been plotted using the production rate at optimum point
(15.97 m3/h), estimated from the curve of production rate versus
IX. However, as discussed in Appendix B, when the production rate
is unknown, its value is assumed and an iterative procedure should
be followed to find the optimum fracture design.
In Fig. 9, a very good agreement is noticed amongst the results
of the three approaches. In this illustration the predicted optimum
parameters are:
Optimum fracture penetration ratio: 0.59.
Optimum absolute fracture conductivity: 2.35.
Optimum effective fracture conductivity: 1.98.
33
25
F(x f)/10000 or Q/m hr or 1/2(r w-eff /m)
(rw-eff)/2
20
15
Volumetric Flow Rate
-1
3
10
5
FSS(xf)/10000
0
0.8 1 0 0.2 0.4 0.6 0.8 1
-5
IX
Fig. 9. Optimization curves of a HF for the case referred to as HFWS-O4 in Table 3.
There is two-phase flow within the whole drainage area. Production rate, calculated
by the in-house simulator, and effective wellbore radius become maximum where
FSS(xf) is zero. For Steady state (SS), optimum fracture penetration ratio is 0.59.
Optimum Reynolds number: 0.78.
3.5. Illustration 5: gas condensate flow, HF with two-phase region not
extending beyond the fracture
The parameters of this illustration (referred to as HFWS-O5 in
Table 3), with the exception of the exterior pressure, are the same
as the last illustration. Here the exterior pressure is 100 psi above
the dew point pressure whilst the bottom-hole pressure is still
55 psi below the dew point pressure. Thus, there are both single-
phase and two-phase flow regions within the system. Moreover,
as the drawdown is increased, the optimum production rate should
increase, resulting in higher inertial effect. Figs. 10 and 11 show the
optimization curves of this system and the gas fractional flow
(GTR) contour map around the wellbore of the optimum design,
respectively. Here again, a good agreement is seen between the re-
sults of different approaches for optimization and all three curves
show almost the same point as the optimum point.
In this illustration, optimum production rate is estimated as
80.9 m3/h, about 5 times more than that of the last illustration.
Thus, as expected the increase in inertia has decreased the opti-
mum fracture length from 146 m to 80 m.
The other optimum parameters are predicted to be:
Optimum fracture penetration ratio: 0.32.
Optimum absolute fracture conductivity: 7.71.
Optimum effective fracture conductivity: 2.46.
Optimum Reynolds number: 2.17.
It has to be emphasized that optimization of the fracture geom-
etry based on the methodology developed for single phase Darcy
flow systems could result in significant errors. For instance, for this
case, using the single phase Darcy flow concept will predict an
optimum fracture penetration ratio equal to 0.8. This results in
the production rate of around 65 m3/h, which is about 20% less
than the production rate gained at the correct optimum IX value
of 0.32.
3.6. Illustration 6: gas condensate flow, HF much longer than the
extent of two-phase region
In this example (referred to as HFWS-O6 in Table 3), the exte-
rior pressure is increased to 3741.5 psi, 200 psi above the dew
point pressure, whilst the bottom-hole pressure is still kept at
55 psi below the dew point pressure. Thus the drawdown is in-
creased to 255 psi (100 psi more than that of the last illustration).
34 H. Mahdiyar, M. Jamiolahmady / Fuel 119 (2014) 27–37

90
F(xf)/10000 or Q/m3hr-1 or rw-eff /m

Volumetric Flow Rate

70

50

30
rw,eff

10
F(xf)/10000

0 0.2 0.4 0.6 0.8 1

-10
IX

Fig. 10. Optimization curves of a HF for the case referred to as HFWS-O5 in Table 3.
The length of the two-phase region is almost equal to the fracture length (See
Fig. 13. GTR = qg/(qg + qc) contour map around the wellbore of the optimum design
Fig. 11). Production rate, calculated by the in-house simulator, and effective
for the case referred to as HFWS-O6 in Table 3. Optimum fracture half length is
wellbore radius become maximum where FSS(xf) is zero. For Steady state (SS),
about 67 m.
optimum fracture penetration ratio is 0.32.

2.5 2
y = 14.592x - 29.166x + 15.573
2
R =1

1.5
fA

0.5
0.7 0.8 0.9 1
A
 
Fig. 11. GTR contour map around the wellbore of the optimum design of HFWS-O5 Fig. 14. Variation of fA versus A fA ¼ A  IX dIdAX .
defined in Table 3. Fracture half length is about 80 m.

1.2
150
1 y = -2.9905x2 + 5.3765x - 1.3789
F(xf)/10000 or Q/m3hr-1 or rw-eff /m

130 Volumetric Flow Rate

R2 = 0.9999
0.8
110
0.6
90
0.4
fB

70
0.2
50
0
30
1 1.1 1.2 1.3 1.4 1.5 1.6
rw,eff
10 -0.2
F(xf)/10000

-10 0 -0.4
0.2 0.4 0.6 0.8 1
IX B
 
Fig. 15. Variation of fB versus B fB ¼ B  IX dIdBX .
Fig. 12. Optimization curves of a HF for the case referred to as HFWS-O6 in Table 3.
The length of two-phase flow region is smaller than the fracture length (See Fig. 13).
Production rate, calculated by the in-house simulator, and effective wellbore radius
become maximum where FSS(xf) is zero. For Steady state (SS), optimum fracture of the two-phase region is almost 36 meters, about half of the frac-
penetration ratio is 0.27. ture half length.
The optimum parameters of this illustration are estimated as
Fig. 12 shows the optimization curves of this system. According follows:
to all three curves in this figure, the optimum fracture penetration
ratio is 0.27; i.e. optimum fracture half length is 67 m, which is Optimum fracture penetration ratio: 0.27.
13 m shorter than that of HFWS-O5, as a result of the increase in Optimum absolute fracture conductivity: 11.00.
the inertial effect. The GTR contour map around the optimum de- Optimum effective fracture conductivity: 3.00.
sign of this HF system is shown in Fig. 13. Here, the half length Optimum Reynolds number: 3.00.
 H. Mahdiyar, M. Jamiolahmady / Fuel 119 (2014) 27–37 35

For this case, using the single phase Darcy flow optimization corporation, Petrobras and Total Exploration UK plc. Their sponsor-
concept will predict an optimum fracture penetration ratio equal ship is gratefully acknowledged.
to 0.8. This results in the production rate of around 100 m3/h,
which is about 30% less than the production rate gained at the cor- Appendix A. Solution of Eq. (3)
rect optimum IX value of 0.27.
The illustrations presented in this section confirm the applica- Fracture conductivity can be written as a function of fracture
bility of the semi-analytical method developed here, which can half length as follows:
be applied to gas or gas condensate flow systems with square
kf wf kf V f C1
drainage areas operating under either SS or PSS conditions, albeit C fD ¼ ¼ 2
¼ 2; ðA:1Þ
kxf 2kxf xf
without the presence of mechanical damage affecting the optimum
fracture geometry.
kf V f
C1 ¼ ; ðA:2Þ
2k
4. Summary and conclusions
Besides, Re, which depends on the flow velocity (v), is also an
A new methodology for optimum fracture design (OFD) of a implicit function of the fracture half length as expressed by,
HFW in gas and gas-condensate reservoirs was proposed. In this q w v w kf bf ðkrgbf þ krcbf Þw
approach the correlation developed recently by Mahdiyar et al. Rew ¼ )
l w
[15] for estimation of the effective wellbore radius of an equivalent
open-hole (EOH) system replicating flow around the actual q w qtw
hydraulic fractured well model was used. Two sets of formulae cor- Rew ¼ ðkrgbf þ krcbf Þw kb ¼ C 2 qtw xf ; ðA:3Þ
l w f f 2hwf
responding to Steady state (SS) and Pseudo-steady state PSS condi-
tions, were developed for calculation of optimum values of the where
fracture conductivity, effective fracture conductivity and half q w kf bf
length of the fracture with given volume that maximize the pro- C 2 ¼ ðkrgbf þ krcbf Þw : ðA:4Þ
l w hV f
ductivity of a HFW, i.e., maximum (r0w ) of the EOH. Several illustra-
tions confirmed the applicability of these newly developed Here, b, h and qtw refer respectively to the single-phase inertial
formulae for single phase and gas condensate systems. factor, formation (or fracture) height and total (gas plus conden-
Some of the important observations are: sate) volumetric flow rate.
Substituting from Eqs. (4) and (9) into Eq. (3) and ignoring the
1. Optimum effective fracture conductivity is independent of the variation of fluid properties with variation of xf gives the following
mobility ratio but depends on the Reynolds number and frac- equation:
ture penetration ratio. !
2. When optimum fracture penetration ratio(Ixopt) is less than @ xf
¼ 0: ðA:5Þ
0.2, the optimum fracture design is the same for both SS and @xf 2  d þ MRpw C1 ð1 þ 0:62Rew Þ
fD V f ;P w
PSS and, depending on the Reynolds number, optimum effective
fracture conductivity (CfDeff,opt) can be between p/2, for low Therefore:
velocity systems (Re  0), and p for very high velocity systems   
(Re  1). In general, covering all range of Ix variation, (CfDeff,opt)
2d þ MRpw 1
C fD
dd
ð1 þ 0:62Rew Þ  xf 2 dx þ MRpw @x@ 1þ0:62Re
C
w
f f fD
 2
varies between 0.785 and p for SS systems and p/2 and 22.6 for
2d þ MRpw C1 ð1 þ 0:62Rew Þ
PSS systems. fD

3. For HFW systems with constant exterior pressure (at SS condi- ¼ 0; ðA:6Þ
tions), optimum values of fracture conductivity and effective
fracture conductivity decrease with increasing fracture penetra- while:
   
tion ratio. @ 1 þ 0:62Rew @ 1 1 @
4. For HFW systems with a closed exterior boundary (at PSS con- ¼ ð1 þ 0:62Rew Þ þ ð1
@xf C fD @xf C fD C fD @xf
ditions), optimum values of fracture conductivity and effective
þ 0:62Rew Þ: ðA:7Þ
fracture conductivity absolute/effective fracture conductivity
increase with the increase in fracture penetration ratio. According to Eqs. (A.2) and (A.3), it can be said that:
5. It was demonstrated that in HFWs with a large propped number  
@ 1 2xf
and a penetration ratio greater than 0.2, whereby the well per- ¼ ðA:8Þ
formance and hence, optimum fracture geometry depend on @xf C fD C1
the exterior boundary condition, the applicability of the UFD and
method, which was originally developed for single-phase PSS  
Darcy flow regime, to single-phase SS Darcy flow regime is @Rew @ðC 2 qtw xf Þ @q
¼ ¼ C 2 qtw þ ðxf tw Þ : ðA:9Þ
erroneous. @xf @xf @xf
6. It was demonstrated that the application of the methods such At the optimum point for a fixed drawdown, the flow rate is the
as UFD, which are based on single phase Darcy flow systems, maximum, @q tw
¼ 0. Therefore, Eq. (A.9) can be simplified as
@xf
for gas condensate two-phase flow systems could result in sig- follows:
nificant errors.
@Rew
¼ C 2 qtw : ðA:10Þ
@xf
Acknowledgments
Combining Eq. (A.6), (A.7), (A.8), and (A.10) results in:
The above study was conducted as a part of the Gas-condensate  
Recovery Project at Heriot-Watt University. This research project is
dd p
2 d  xf  ð1 þ 2  0:62Rew Þ ¼ 0: ðA:11Þ
sponsored by: The Ecopetrol/Equion, Eni Petroleum Co, Inpex dxf MRw C fD;opt
36 H. Mahdiyar, M. Jamiolahmady / Fuel 119 (2014) 27–37

Solving the above equation for optimum fracture conductivity Optimum fracture half-length for SS is gained when FSS(xf) is
(CfD,opt) gives: equal to zero. When none of the roots of Eq. (A.19) is between zero
p and reservoir half length, the plot of FSS(xf) versus IX does not inter-
C fD;opt ¼ dd
ð1 þ 2  0:62Rew Þ: ðA:12Þ sect the x-axis, and optimum practical fracture design is achieved
2ðd  xf dxf
ÞMRw
at IX = 1.
Since d has different expressions for SS and PSS, further steps of
this part of the derivation will come in the next two separate sec- A.2. Pseudo-steady state systems
tions for SS and PSS systems.
For Pseudo-steady state (PSS) systems, d is equal to B (Eq. (6))
A.1. Steady state systems and Eq. (A.12) is rewritten as follows:
p 
For Steady state (SS) systems, d is equal to A (Eq. (5)) and Eq.
C fD;opt ¼  ð1 þ 2  0:62Rew Þ: ðA:24Þ
dB
2 B  xf dxf
MRw
(A.12) is rewritten as:
p  Similarly to what was explained in the previous section:
C fD;opt ¼  ð1 þ 2  0:62Rew Þ: ðA:13Þ
dA
2 A  xf dxf
MRw dB xf dB dB
xf ¼ ¼ IX : ðA:25Þ
dxf xe dIX dIX
On the other hand:
Thus Eq. (A.24) converts to:
dA xf dA dA
xf ¼ ¼ IX : ðA:14Þ p
dxf xe dIX dIX C fD;opt ¼ ð1 þ 2  0:62Rew Þ; ðA:26Þ
2f B MRw
Thus Eq. (A.13) converts to:
where
p
C fD;opt ¼ ð1 þ 2  0:62Rew Þ; ðA:15Þ dB
2f A MRw fB ¼ B  IX : ðA:27Þ
dIX
where
Fig. 15 shows fB as a function of B for the complete range of IX
dA (0 6 IX 6 1:0). Here again a second order polynomial function can
fA ¼ A  IX : ðA:16Þ
dIX represent fB for the whole domain of B [1.0, 1.57], which is:
In Eq. (A.16), A is a function of fracture penetration ratio (Ix) (see
fB ¼ 2:9905B2 þ 5:3765B  1:3789: ðA:28Þ
Eq. (5)). Fig. 14 shows fA as a function of A for the complete range of
IX (0 6 IX 6 1:0). According to this figure, the following second or- For small fracture penetration ratios, where B should be one, Eq.
der polynomial function can represent fA for the whole domain of A (A.28) estimates fB as 1.0071, which shows the good accuracy of
[0.723, 1.00]; i.e.: this equation. However, it is seen from Fig. 15 that fB becomes zero
at B = 1.488 (IX = 0.972), which is due to the lack of accuracy of the
fA ¼ 14:592A2  29:166A þ 15:573: ðA:17Þ slope of B versus IX. Similar behavior is also noticed from data pro-
For small fracture penetration ratios, where A is about one, fA cessing of the outcomes of Eq (D.3) of the Meyer et al. [16] paper
should be 1.0 while it is estimated as 0.999 by Eq. (A.17), verifying studying single-phase Darcy PSS flow around a HFW. That is, both
the high precision of this equation. equations slightly overestimate the slope of B versus IX when IX ap-
Combining Eq. (A.15) with Eq. (9), CfDeff,opt can be calculated as proaches 1.0.
follows: As fB becomes zero and then negative for IX P 0:97;there is a
need to restrict the applicability of the results of this part. That
MRw p ð1 þ 2  0:62Rew Þ is, IX should be less than or equal to 0.95. It should be noted, how-
C fDeff ;opt ¼ C fD;opt ¼
1 þ 0:62Rew 2f A 1 þ 0:62Rew ever, that this restriction is not very critical. In the illustration sec-
 
p 0:62Rew tion it is shown that around the optimum point, the variation of
) C fDeff ;opt ¼ 1þ : ðA:18Þ
2f A 1 þ 0:62Rew effective wellbore radius with IX is insignificant and that when
the optimum IX is greater than 0.95, its performance is almost
Substituting CfDopt and the Reynolds number in terms of xfopt
the same as that at IX = 1.
in Eq. (A.15) results in the following equation, which is a useful tool
Combining Eq. (A.26) with Eq. (9), the optimum effective frac-
for finding optimum fracture half length.
ture conductivity can be calculated as follows:
aSS  x3f opt þ bSS  x2f opt þ cSS ¼ 0; ðA:19Þ MRw p ð1 þ 2  0:62Rew Þ
C fDeff ;opt ¼ C fD;opt ¼ )
where 1 þ 0:62Rew 2f B 1 þ 0:62Rew

0:62pC 2 qtw  
aSS ¼ ; ðA:20Þ p 0:62Rew
fA  MRw C fDeff ;opt ¼ 1þ ðA:29Þ
2f B 1 þ 0:62Rew
p Substituting CfDopt and Reynolds number in terms of xfopt in
bSS ¼ ; ðA:21Þ
2  fA  MRw Eq. (A.26) results in the following equation, which is a useful tool
for finding optimum fracture length in PSS systems:
kf V f
cSS ¼ C 1 ¼  : ðA:22Þ aPSS  x3f opt þ bPSS  x2f opt þ cPSS ¼ 0; ðA:30Þ
2k
This equation can be solved simply by drawing FSS(xf), defined where
below, versus xf or IX. 0:62pC 2 qtw
aPSS ¼ ; ðA:31Þ
fB  MRw
F SS ðxf Þ ¼ aSS  x3f þ bSS  x2f þ cSS : ðA:23Þ
 H. Mahdiyar, M. Jamiolahmady / Fuel 119 (2014) 27–37
p References
bPSS ¼ ; ðA:32Þ
2  fB  MRw
kf V f
cPSS ¼ C 1 ¼  : ðA:33Þ
2k
Here, again drawing FPSS(xf) (defined below) versus IX provides
the optimum fracture penetration ratio for PSS systems:
F PSS ðxf Þ ¼ aPSS  x3f þ bPSS  x2f þ cPSS : ðA:34Þ
Appendix B. Optimum fracture design problems
In this section, the corresponding procedures for solving three
possible types of optimum fracture design problems that could oc-
cur in practical field applications are described.
Type I – Production rate and drawdown known, minimum frac-
ture volume and its optimized dimensions unknown.
1- Assume fracture volume.
2- Calculate optimum fracture length and thickness.
3- Calculate optimum effective wellbore radius.
4- Simulate the drainage area and calculate the production
rate.
5- Check if the difference between the calculated production
rate and the known value is less than the acceptable value.
YES: Go to step 6.
NO: Read the calculated value for fracture volume and go
back to step 2.
 6. Report the designed fracture geometry.
Type II – Production rate and fracture volume known, optimized
fracture geometry giving the lowest drawdown unknown.
1- Assume the bottom-hole pressure (Pw).
2- Estimate fluid properties and then mobilities at Pw.
3- Calculate optimum fracture length and thickness.
4- Calculate optimum effective wellbore radius.
5- Simulate the drainage area and calculate the bottom-hole
pressure.
6- Check if the difference between the calculated bottom-hole
pressure and the assumed value is less than the acceptable
value.
YES: Go to step 7.
NO: Read the calculated value for Pw and go back to step 2.
7- Report the designed fracture geometry and bottom-hole
pressure.
Type III- Bottom-hole pressure and fracture volume known, the
optimized fracture geometry giving the highest production rate
unknown.
1- Assume the production rate (qw).
2- Calculate optimum fracture half length and thickness.
3- Calculate optimum effective wellbore radius.
4- Simulate the drainage area and calculate the production
rate.
5- Check if the difference between the calculated production
rate and the assumed value is less than the acceptable value.
YES: Go to step 6.
NO: Read the calculated value for q and go back to step 2.
6- Report the designed fracture geometry and production rate.
37
[1] Economides M, Oligney R, Valko P. Unified fracture design. Alvin, Texas,
USA: Orsa Press; 2002.
[2] Cinco-Ley H, Samaniego VF. Effect of wellbore storage and damage on the
transient pressure behaviour of vertically fractured wells. SPE 6752. In:
Presented at 52nd annual fall technical conference, Denver, USA, 1977.
[3] Giddley JL. A method for correcting dimensionless fracture conductivity for
non-Darcy flow effects. SPE Prod Eng J 1991;6(4):391–4 [SPE-20710-PA].
[4] Guppy KH, Cinco-Ley H, Ramey Jr HJ, Samaneigo VF. Non-Darcy flow in wells
with finite-conductivity vertical fractures. SPE J 1982;22(5):681–98 [SPE 8281].
[5] Huang H, Ayoub JA. Modeling non-Darcy flow and perforation convergence for
vertically fractured wells. SPE 107853. In: Presented at European formation
damage conference, Scheviningen, Netherland, 2007.
[6] Henderson GD, Danesh A, Tehrani DH, Al-Shaldi S, Peden JM. Measurement and
correlation of gas condensate relative permeability by the steady-state method.
SPE J 1996;1(2):191–201. http://dx.doi.org/10.2118/31065-PA [SPE-31065-PA].
[7] Henderson GD, Danesh A, Tehrani DH. Effect of positive rate sensitivity and
inertia on gas condensate relative permeability at high velocity. Pet Geosci
2001;7(1):45–50.
[8] Indriati Sh, Wang X, Economides MJ. Adjustment of hydraulic fracture design
in gas condensate wells. In: Paper SPE 73751 presented at the international
symposium and exhibition on formation damage control, Louisania, USA,
February 20–21, 2002. <http://dx.doi.org/10.2118/73751-MS>.
[9] Jamiolahmady M, Danesh A, Tehrani DH, Duncan DB. A mechanistic model of
gas-condensate flow in pores. Transp Porous Media 2000;41(1):17–46. http://
dx.doi.org/10.1023/A:1006645515791.
[10] Jamiolahmady M, Danesh A, Tehrani DH, Sohrabi M. Variations of gas-
condensate relative permeability with production rate at near wellbore
conditions: a general correlation. SPE Reservoir Engineering and Evaluation
Journal 2006;9(6):688–97 [SPE-83960-PA, Also presented in the Offshore
Europe Conference and Exhibition, September 2–5 2003, Aberdeen, United
Kingdom. http://dx.doi.org/10.2118/83960-MS.].
[11] Jamiolahmady M, Sohrabi M, Ireland S. Gas condensate relative permeabilities
in propped fracture porous media: coupling versus inertia, SPE 115726. SPE J
2008;15(1):208–22 [Also presented at the 2008 SPE Annual Technical
Conference and Exhibition, Denver, Colorado, September 21–24, 2008.
http://dx.doi.org/10.2118/115726-MS.].
[12] Jamiolahmady M, Sohrabi M, Ireland S, Ghahri P. A generalized correlation for
predicting gas-condensate relative permeability at near wellbore conditions. J
Pet Sci Eng 2009;66:98–110. http://dx.doi.org/10.1016/j.petrol.2009.02.001.
[13] Lopez-Hernandez HD, Valko PP, Pham TT. Optimum fracture design minimizes
the impact of non-Darcy flow effects. In: Paper SPE 90195 presented at the SPE
annual technical conference and exhibition, Houston, USA, September 26–29,
2004. http://dx.doi.org/10.2118/90195-MS.
[14] Mahdiyar H, Jamiolahmady M, Sohrabi M. Gas condensate flow around
hydraulically fractured wells. Transp Porous Media 2011. http://dx.doi.org/
10.1007/s11242-011-9783-0.
[15] Mahdiyar H, Jamiolahmady M, Sohrabi M. Improved Darcy and non-Darcy flow
formulations around hydraulically fractured wells. J Pet Sci Eng
2011;78:149–59. http://dx.doi.org/10.1016/j.petrol.2011.05.013.
[16] Meyer BR, Jakot RH. Pseudo-steady state analysis of finite conductivity vertical
fractures. In: Paper SPE 95941 presented at the SPE annual technical
conference, Texas, USA, October 9–12, 2005. http://dx.doi.org/10.2118/
95941-MS.
[17] Mohan J, Pope GA, Sharma MM. Effect of non-Darcy flow on well productivity
of a hydraulically fractured gas/condensate well. In: Paper SPE 103025
presented at the SPE gas technology symposium, Texas, USA, September 24–
27 2006. http://dx.doi.org/10.2118/103025-MS.
[18] Prats M. Effect of vertical fractures on reservoir behaviour-incompressible
fluid case. SPE J 1961;1(2):105–18 [SPE-1575-G].
[19] Settari A, Bale A, Batchman RC, Floisand V. General correlation for the effect of
non-Darcy flow on productivity of fractured wells. In: Paper SPE 75715
presented at gas technology symposium, Calgary, Canada, 2002.
[20] Valko PP, Economides MJ. Heavy crude production from shallow formations:
long horizontal wells versus horizontal fractures. In: Paper SPE 50421
presented at the SPE international conference on horizontal well technology,
Calgary, Canada, November 1–4 1998. http://dx.doi.org/10.2118/50421-MS.
[21] Wang X, Indriati S, Valko PP, Economides MJ. Production impairment and
purpose built design of hydraulic fracture in gas condensate reservoirs. In:
Paper SPE 64749 presented at the SPE international oil and gas conference and
exhibition, Beijing, China, November 7–10, 2000. http://dx.doi.org/10.2118/
64749-MS.
[22] Bardon C, Longeron DG. Influence of very low interfacial tension on relative
permeability. SPE J 1980:391–401.
[23] Cinco-Ley H, Samaniego-V F, Dominguez AN. Transient pressure behaviour for
a well with a finite conductivity vertical fracture. SPE J 1978;6:253–64 [SPE
6014].
[24] McGuire W, Sikora V. The effect of vertical fractures on well productivity. J Pet
Technol 1960;12(10):72–4 [paper SPE 1618-G].