PRESSURE AND PRESSURE DERIVATIVE ANALYSIS FOR FRACTURED

HORIZONTAL WELLS IN UNCONVENTIONAL SHALE RESERVOIRS USING

DUAL-POROSITY MODELS IN THE STIMULED RESERVOIR VOLUME
Freddy Humberto Escobar1, Karla María Bernal1 and Guiber Olaya-Marin2
1
Universidad Surcolombiana/CENIGAA, Avenida Pastrana - Cra 1, Neiva, Huila, Colombia
2
Universidad Surcolombiana, Avenida Pastrana - Cra 1, Neiva, Huila, Colombia

E-Mail: fescobar@usco.edu.co

ABSTRACT
Nowadays, the oil industry efforts revolve around the exploitation of hydrocarbon-bearing shale formations. At
the current rate of exploitation several shale formations can produce for more than 200 hundred years. Therefore,
researches are conducted for a better characterization of these formations so all the local details can be captured by the
mathematical models. From the well test point of view, some mathematical models to observe transient pressure behavior
have been introduced. However, such models are very complex and may not be practical for a practicing engineering who
neither does have the time to code such models or their computer programming capabilities are not good enough. Then,
this work takes two mathematical models from the literature as the basis to develop a practical well test interpretation
methodology using characteristic points found of the pressure and pressure derivative curves. The resulting methodology
was successfully tested with synthetic examples.

Keywords: pseudosteady state, bilinear flow regime, transient-pressure analysis, shale

1. INTRODUCTION fractured horizontal wells in unconventional shale

Currently, pressure transient analysis is focused reservoirs presented by Brown et al. (2011) who
on the characterization of such unconventional systems as developed their solutions following the work of Cinco-Ley
shale formations which permeability is extremely low and Meng (1988) for the finite-conductivity fracture
going to the range of microdarcies to nanodarcies. solution in a dual-porosity reservoir. Brown et. al. (2011)
The mathematical model governing either rate- or derived their solution for the outer reservoir, inner
pressure-transient behavior are complex and none of the reservoir, and the hydraulic fracture, and then couple the
commercial well test interpretation softwares have solutions with the flux- and pressure- continuity
included them. Then, there is a need of providing practical conditions on the interfaces between the regions and then
interpretation techniques as those presented by Bernal, inverted numerically from the Laplace space. Their
Escobar, and Ghisays-Ruiz (2014) and Escobar, mathematical model is brought here and given below.
Montenegro, and Bernal (2014) who used the philosophy
of the TDS technique, Tiab (1993), for the development of Outer Reservoir Solution
easy-to-use interpretation methodology. The outer reservoir solution in the Laplace space
The models proposed by Brown et al. (2011) is given by:
allows to study the well-pressure behavior of shale
formations using the stimulated reservoir volume, SRV,
cosh  s OD ( xeD  xD )
concept around a horizontal well and divided into three  
POD  PID (1)
different zones which have individual properties. The first xD 1 xD 1,
cosh  s OD ( xeD 1)
one corresponds to the finite-conductivity hydraulic  
fracture. The second zone deals with the inner reservoir
region which can be either homogeneous or naturally The outer reservoir solution, POD, is given in
fractured, and the third one deals with the external portion. terms of the inner-reservoir pressure at the interface of the
Brown et al. (2011) consider that the naturally fracture inner and outer reservoir pressure at the interface of the
reservoir can be governed by a pseudosteady-state or PID x 1,
transient model; then, two models were developed by them inner and outer reservoirs, D

and use in this work to provide an interpretation

methodology using characteristic points found on the Inner-Reservoir Solution
pressure derivative plot. The developed equations were The solution for this case in the Laplace space is:
successfully tested with simulated cases.
 
2. MATHEMATICAL FORMULATION  cosh   o ( y eD  y D )  
PID  ( PFD )     (2)
2.1. Mathematical Model y D  wD / 2 
 wD  
This study is based on the solutions of the  cosh   o ( y eD  2 )  
trilinear-flow model for pressure-transient responses of    

1
 Where the outer-reservoir parameter used in the The dimensionless wellbore storage coefficient,
trilinear-flow model is: CD, is given:

o  s  O D tanh  s  O D ( x eD  1)  (3) CD 
5.615 C
(12)
2   cht  I x F2
And the outer-reservoir parameter, o, used in
trilinear- flow model is, Dual-Porosity Parameters
The naturally-fractured reservoir parameters
o introduced by Warren and Root (1963) were used by
o  u (4) Brown et al. (2011),
C RD y eD

( ct ) f
Hydraulic-Fracture Solution   (13)
The hydraulic fracture parameters used in the ( ct ) f  ( ct ) m
trilinear- flow model are:
k
 w     l 2 m (14)
 k
F   o tanh   o  y eD  D   (5) f
  2 
Where l is a reference length chosen as the half
2 F s length of the hydraulic fracture; that is, l = xf.
F   (6)
C FD  FD
Kazemi et al. (1976) proposed the following first
The dimensionless pressure solution for the approximation for the shape factor of rectangular matrix
hydraulic fracture is obtained as: blocks:

 cosh   F (1  x D )   1 1 1 
PFD  (7)   4  2  2  (15)
 L2 
C FD s  F sinh(  F )  x L y Lz 

Wellbore-Pressure Solution Serra et al. (1983) define the storativity and flow
The dimensionless wellbore pressure is obtained capacity ratios for the transient dual-porosity model,
at xD = 0 is given by, respectively, by:

 ( c t ) m
Pw D  PFD ( x D  0)  (8)   (16)
C FD s  F tanh(  F ) ( c t ) f

The skin factor caused by flow choking within  l 2   k m hm 

the fracture is provided by Mukherjee and Economides   12  2   (17)

 hm   k f h f 
(1991):
The dimensionless oil pressure and the pressure
k h   h   derivative, respectively, are:
sc  I I ln    (9)
k F wF   2rw  2 
k I h  Pi  Pwf 
The dimensionless wellbore pressure after the end PD  (18)
141.2 Bo q f o
of the radial flow in the hydraulic fracture:
k f h(t * P ')
t D * PD '  (19)
 s 141.2 q f Bo o
PD   c (10)
C FD s  F tanh(  F ) s
The dimensionless gas pseudopressure and the
The effect of wellbore storage was incorporated dimensionless pseudopressure derivative, respectively, for
into the solution by substituting PD from Equation 10 into the constant-production-rate solution are:
the following convolution expression in the Laplace
domain: k I h  m  Pi   m( Pwf ) 
Pw D PD  (20)
Pw D , storage  (11) 1424 T q f
1  C D s 2 Pw D

2
 k I h[t * m( P) '] The dimensionless fracture and reservoir
t D * m( PD ) '  (21) conductivities, respectively:
1424 q f T

The dimensionless time is defined by; k f wf

c fD  (31)
k xI f
I
tD  t (22) kI x f
x 2f cRD  (32)
k o ye
The inner-reservoir diffusivity,
For any dual-porosity model, CFD and CRD are
0.0002637k I defined on the basis of the bulk permeability of the inner
I  (23)
( ct ) I  reservoir kI . Brown et. al. (2011) also defined the
hydraulic fracture diffusivity and the outer-reservoir
The hydraulic-fracture diffusivity, diffusivity ratios, respectively, as:

0.0002637 k f f
f  (24)  fD  (33)
( ct ) f  I

The outer-reservoir diffusivity, O

OD  (34)
I
0.0002637ko
o  (25)
( ct )o  For the transient dual-porosity model:

The dimensionless distances in the x and y kI  k f (35)

direction are defined, respectively, by
hI  n f h f (36)
x
xD  (26)
xf And the intrinsic fracture storativity is;
y
yD  (27) ( ct ) I  ( ct ) f
xf (37)

For the pseudosteady dual-porosity model:

The dimensionless distances to the reservoir
boundaries are given by xeD and yeD.
k I  k f
(38)
According to Brown et. al. (2011), the pressure-
transient response of a horizontal well with nF identical
transverse hydraulic fractures can be modeled by
hI  h (39)
considering one of the fractures producing from a
rectangular reservoir section at a rate: And the intrinsic fracture storativity

qf 
q
(28) ( ct ) I  (ct ) f  (ct ) m (40)
nf

2.2. TDS Formulation for Pseudostady Dual-Porosity

The fracture is located centrally in the closed
Model
rectangular drainage area
The Equations will be developed for oil flow. For
gas flow refer to Appendix A and B. Even though, Brown
A, drainage f  2 x e  2 y e (29)
et. al. (2011) presented pressure solutions for each flow
regime, new simpler expressions based on the
The half-fracture length, xf, can be related to observations of the pressure derivative are derived in this
reservoir length, xe, by: work. However, a metholodology with the equations
proposed by Brown et. al. (2011) for each flow regime is
xe  2 x f (30) provided in Appendixes C and D.

2.2.1. Initial Flow Regime

3
 The flow at early times displays a pressure
derivative which slope goes from 0.25 to 0.5 depending Once the dimensionless quantities given by
upon the value of the storativity ratio, ω, as defined by Equations 19, 22 and 23 are replaced into Equation 41, it
Equation 13. This flow regime shows up at diferent times yields;
depending on the ω values, as shown in Figure-1. The
early time portion of the log-log plot of the dimensionless k I h t * P '  0.0002637 k I t 
0.25
0.4159
pressure derivative versus dimensionless time curve under    (42)
constant flow capacity ratio, λ, and variable storativity 141.2 q f Bo o (1   )  ( ct ) I  x f 
2

ratio, ω, is used for the determination of the initial flow

regime governing equations. Equation 42 allows solving for the permeability
For this model the developed equations depend of the inner reservoir;
on the storativity ratio value, ω. The estimated
permeability value correponds to the one in the inner zone 1

of the stimulated reservoir volume, SRV. To obtain an  24.9496 q B   t 

0.25
 0.75
kI   f o o
 (43)
 
appropriate flow overning equation is necessary to adjust  (1   )h(t * P ')   ct  I  x 2f  
the application range according to the values the  
storativity ratio, ω.
Notice that the hydraulic half- fracture length, xf,
1.E+04 can be solved from Equation 43,
P
s
e
u
d
o
s
t
e
a
d
y
M
o
d
e
l

P
s
e
u
d
o
s
t
e
a
d
y
s
t
a
t
e
C
o
n
s
t
a
n
t
f
l
o
w
c
a
p
a
c
i
t
y
r
a
t
i
o

1.E+03
 1x10-3 2
 24.9496 q B   t BL  
0.25

xf   f o o
 (44)
 
 (1   )k I0.75 h(t * P ') BL   ct  I   
1.E+02

1x10
9x10 -1
0
 
t D * PD'

1.E+01
5x10 -1
4x10 -1
3x10 -1
2x10 -1 When 0.01 ≤ ω < 0.1, the early flow is defined by
F
i
r
s
t
-
P
s
e
u
d
o
s
t
e
a
d
y
s
t
a
t
e

1.E+00
1x10 -1
:
:
-4
a slope of 0.33 in a log plot of the pressure derivate. It
gives a relationship of three log cycles in the time axis
1.E-01 1x10
1x10 -5
I
n
i
t
i
a
l
F
l
o
w

1.E-02
against one log cycle in the pressure derivative axis. As
1.E-04 1.E-03 1.E-02 1.E-01 1.E+00

tD
1.E+01 1.E+02 1.E+03 1.E+04
shown in Figure-3, the flow behavior does not present a
Figure-1. Effect of storativity ratio, ω, on the initial flow (early uniformity related to the variation of ω, then, it was
time) for the dual-porosity pseudosteady model under constant necessary to determine the most representative
flow capacity ratio, λ, and yD=xD=1 mathematical representation of the initial flow behavior
which was performed by using a probabilistic average,
The determination of the governing equation for
for ω ≥ 0.1 at early time requires a log-log plot of tD*PD’ t D * PD '  0.87847 t D0.33 (45)
or tD*m(P)D’ versus tD/(1+ω)4 using a constant flow
capacity ratio, λ, as shown in Figure-2. The initial flow is 1.E+07

characterized by a typical 0.25-slope line (bilinear flow

P
s
e
u
d
o
s
t
e
a
d
y
M
o
d
e
l

1.E+06
regime) on the pressure derivative curve.
C
o
n
s
t
a
n
t
f
l
o
w
c
a
p
a
c
i
t
y
r
a
t
i
o

1.E+05
 1x10-3

1.E+04 1.E+04
P
s
e
u
d
o
s
t
e
a
d
y
M
o
d
e
l
C
o
n
s
t
a
n
t
f
l
o
w
c
a
p
a
c
i
t
y
r
a
t
i
o

1.E+03
t D * PD'

1.E+03
  1x10-3
1.E+02

1.E+02 8x10 -2
Average
1.E+01 6x10 -2
value
4x10 -2
t D * PD'

1.E+01  1.E+00 1x10 -2

1x10 0
9x10 -1
5x10 -1
Initial Flow
1.E+00 1.E-01
4x10 -1 Slope of 0.33
3x10 -1
2x10 -1
1.E-02
I
n
i
t
i
a
l
F
l
o
w

1x10 -1
1.E-01 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07
S
l
o
p
e
o
f
0
.
2
5

tD
1.E-02
1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04
Figure-3. Initial flow behavior for the case 0.01 ≤ ω < 0.1 on the
tD
(1+ 4
pressure derivative behavior governed by the dual-porosity
Figure-2. Effect storativity ratio, ω, on the initial flow for the pseudosteady model with constant flow capacity ratio, λ, and
case ω ≥ 0.1 on the pressure derivative behavior using the the yD=xD=1
dual-porosity pseudosteady model with constant flow capacity
ratio, λ, and yD=xD=1
0.25 Once the dimensionless quantities given by
 t  Equations 19, 22 and 23 are replaced into Equation 45, it
t D * PD '  0.4159  D
 (41)
 1   4  results;
 

4
  0.0002637k I t 
0.33 Notice that the half-fracture length, xf, can also be
k I h(t * P ')
 0.87847  (46) solved from Equation 51,
141.2 q f Bo o  ( ct ) I  x 2f 
 
 38.5384q B   tL 
0.5

xf   f o o
 (52)
Equation 46 allows solving for the inner-reservoir  
permeability;  h(t * P ') L   ct  I  k I  
 
1
 25.685 q B   For the three early-time cases the reservoir lenght
 
0.33 0.67
tTL (47) can be solved from Equation 30 which depends only on
kI   f o o
  
 h(t * P ')TL   ct  I  x 2f   the half- fracture length, xf.
 
2.2.2. Second flow regime on first pseudosteady state
Notice that the half-fracture length, xf, can be
Once the initial flow regime vanishes, a second
obtained from Equation 47,
unit-slope flow regime is observed on the prssure
1 derivative log-log plot. It has been arbitriarily refered here
 25.685 q B   t  
0.33 0.66
as “first pseudosteady-state regime” which is affected by
x f   0.67 (48)
 
f o o
 TL
both the flow capacity, λ, and the storativity ratios, ω. The
 k I h(t * P ')TL   ct  I    former affects the duration time when the regime is
 
presented as shown in Figure-5; thereby, the starting time
Finally, the last case of ω variation at early time of the first pseudosteady state regime converges at the
takes places for 0 < ω < 0.01 which is characterized by a same point for different  values. With this parameter the
half-slope (linear flow regime) in the pressure derivative flow regime varies in length but it is the same in terms of
log-log plot. A typical representation of the homogeneus location (no parallel displacement along the time axis). In
reservoir occurs when ω=0 as depicted in Figure-4. contrast, the storativity ratio, ω, causes a parallel
displacement along the time axis as shown in Figure-6.
1.E+04
1.E +06
P
s
e
u
d
o
s
t
e
a
d
y
M
o
d
e
l

1.E+03
C
o
n
s
t
a
n
t
f
l
o
w
c
a
p
a
c
i
t
y
r
a
t
i
o

1.E +05

 1x10-3 

Ps
st
ea
ue
d
o
s
t
e
a
d
y
-4
1x10
1.E+02

t
1.E +04 -3
9x10
-2
5x10
-1
4x10
t D * PD'

1.E+01 1.E +03

t D * PD'

Linear 
Flow 0 1.E +02
1.E+00 1x10 -5
1x10 -4
1x10 -3 1.E +01

P
s
e
u
d
o
s
t
e
a
d
y
M
o
d
e
l
1.E-01 Initial Flow

C
o
n
s
t
a
n
t
s
t
o
r
a
t
i
v
i
t
y
r
a
t
i
o
Slope of 0.33
1.E +00
  1x10 -2
1.E-02
1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E-01
1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E +05
tD tD
Figure-4. Initial flow pressure derivative behavior for ω < 0.01
Figure-5. First pseudosteady state behavior observed in the dual-
with dual-porosity pseudosteady model under constant flow
porosity pseudosteady model under constant storativity ratio, ω,
capacity ratio, λ, and yD=xD=1
and yD=xD=1

t D *( PD ) '  5.35 tD (49) The general equation for the unit-slope

pseudosteady-state period is:
After plugging the dimensionless quantities into
Equation 49, we obtain; t D *( PD ) '  b t D (53)

k I h t *(P) '
0.5
 0.0002637 k I t  As the displacement along the time axis is
 5.35  (50)
141.2 q f Bo o  ( ct ) I  x 2f  defined by the value of storativity ratio, ω, it is possible to
  develop an expression to obtain b as a function of ω.
These observations lead to obtain the following
Equation 50 allows solving for the permeability expression:
of the inner reservoir;
2
 38.5384 q B   t  
0.5

kI   (51)
 
f o o
 L

 h(t * P ') L x f   ct  I   
 

5
 1.E+04 Finally, drainage area and half-reservoir length,

P
s
e
u
d
o
s
t
e
a
d
y
s
t
a
t
e
xe, can be computed using equations 29 and 30,
P
s
e
u
d
o
s
t
e
a
d
y
M
o
d
e
l
C
o
n
s
t
a
n
t
F
l
o
w
c
a
p
a
c
i
t
y
r
a
t
i
o
1.E+03 respectively.
 1x10-3

1.E+02
2.2.3. Late-time Pseudosteady-State Regime
Pseudosteady-state is the last flow regime that
t D * PD'

1.E+01

 can be seen in a test run long enough in the systems we are

1x10 0
1.E+00
5x10 -1 dealing with. This is recognized by a unit-slope straight
F
i
r
s
t
p
s
e
u
d
o
s
t
e
a
d
y
s
t
a
t
e
1x10 -1

1.E-01
6x10 -2
1x10 -2
line. It is important to remark that in transient-pressure
1x10 -3
0
analysis for the dual-porosity pseudosteady model, this
1.E-02 behavior is independent of the flow capacity ratio, λ, and
1.E-04 1.E-03 1.E-02 1.E-01 1.E+00
tD
1.E+01 1.E+02 1.E+03 1.E+04
the storativity ratio, ω; see Figures-5 and 6; however, the
Figure-6. First pseudosteady state behavior observed in the behavior of the late pseudosteady-state regime is particular
dual-porosity pseudosteady model under constant flow capacity for different dimensionless length values (yD= 0.35, 0.5,
ratio, λ, and yD=xD=1 075 and 1), as shown in Figure-6.
For this reason the determination of the
governing equation for the late pseudosteady period
1) Equation 54 which constants are given in Table-1 was requires a log-log plot of tD*PD’ or tD*m(P)D’ versus tDA
developed to estimate b using the storativity ratio, ω. Its using a constant parameter of flow-capacity ratio and
range of application is for values of 0 ≤ ω < 1. storativity ratio (λ and ω = constant) and different
dimensionless length values (yD= 0.35, 0.5, 075 and 1), as
1
b   A  B   C ( 0.5 )  (54) shown in Figure-7. In each case, a uniform behavior was
found by dividing the dimensionless time by the
dimensionless length of the stimulated reservoir volume
Table-1. Constants for Equation 54 for each case respectively.
Constant Value
A 0.00361707 tD
t DA 
B 0.646008952 yD2 (56)
C -0.00058982
1.E+08
P
s
e
u
d
o
s
t
e
a
d
y
M
o
d
e
l

P
s
e
u
d
o
s
t
e
a
d
y
s
t
a
t
e
1.E+07
C
o
n
s
t
a
n
t
F
l
o
w
c
a
p
a
c
i
t
y
r
a
t
i
o
a
n
d
c
o
n
s
t
a
n
t
The b value obtained from Equation 54 is
s
t
o
r
a
t
i
v
i
t
y
r
a
t
i
o

1.E+06
replaced into equation 53. It is important to point out that  1x10-3  1x10-2
1.E+05
when ω=1, the first pseudosteady-state regime does not
1.E+04
exist; then, the late-time pseudosteady-state period is only
t D * PD'

1.E+03
seen inidicating a homogeneous behavior. The storativity yD
1.E+02
ratio, ω, only depends on fracture storativity. This 0.25
1.E+01
represents the other extreme case where there is no matrix 0.50
0.75
1.E+00
contribution. 1
1.E-01
1.E-02
Once the dimensionless quantities given by 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07
Equations 19, 22 and 23 are replaced into Equation 53, it tD
yields;
Figure-7. Effect of the variation of the dimensionless length on
the pseudosteady-state period for the late pseudosteady-state
k I h t *(P) '  0.0002637 k I t  period in the dual-porosity pseudosteady model with constant
 b (54)
141.2 q f Bo o  ( ct ) I  x 2f  flow capacity ratio, λ, and constant storativity ratio, ω
 
Then, it is possible to write a general
The first pseudosteady-state regime can be used dimensionless pressure or pseudopressure derivative
to estimate either the hydraulic half-fracture length, xf, or expression for the pseudosteady-state period, as follows,
the well drainage area whether or not the permeability
value is known.
tD * PD 'PSS t 
 0.4973  D2  (57)
1 yD y 
 b 0.03723444 q B   2  D  PSS
tqPSS
xf    
f o


    P 'qPSS 
 
h ( c ) t * Once the dimensionless terms given by Equations
 t I

(55) 19, 22 and 23 are replaced in Equation 57, an expression

for the determination of the half- fracture length, xf, is
obtained,

6
 1.E+08 1
 

P
s
e
u
d
o
s
t
e
a
d
y
M
o
d
e
l
C
   A  B tD * PD 'max 
1.E+07
C
o
n
s
t
a
n
t
F
l
o
w
c
a
p
a
c
i
t
y
r
a
t
i
o
a
n
d
c
o
n
s
t
a
n
t
 (60)

s
t
o
r
a
t
i
v
i
t
y
r
a
t
i
o
tD * PD 'max 

P
s
e
u
d
o
s
t
e
a
d
y
s
t
a
t
e
1.E+06
1.E+05
 1x10-3  1x10-2

1.E+04
t D*P D' /yD

yD
1.E+03
0.25
The range of application of Equation 60 is for
1.E+02 0.50
0.75 1x10-5 ≤ λ < 1. That applies to extreme cases. Constants A
1.E+01 1
through C are given in Table-2.
1.E+00
1.E-01
1.E-02
Table-2. Constants for Equation 60
1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 1.E+08

t DA
Constant Value
Figure-8. Effect of the variation of the dimensionless length on A -0.22892364
the late pseudosteady-state period for the dual-porosity B 1.693500511
pseudosteady model with constant flow capacity ratio, λ and C 0.004208008
constant storativity ratio, ω
The minimum point (trough) of the pressure
0.058172 Bo q f t PSS derivative, (tD *PD’)min, is used to evaluate the storativity
xf  (58)
hI  ct  I y  ( P)'PSS
t *  ratio, ω. This point, however, is affected by λ, then, a
correct equation for ω computation should include, so that:
Notice that xf can be solved from Equations 55
and 58 whether or not the permeability value is known.   A  B  (t D * PD ') min *    C  (tD * PD ') min *  
3
(61)
For this reason is possible calculate the distance y-
coordinate.
Constants A through C are given in Table-3.
0.058172 Bo q f t PSS Table-3. Constants for Equation 61
y (59)
hI  ct  I x f t * ( P) 'PSS Constant Value
A -0.00674835
2.2.4. Dimensionless storativity, ω, ratio and flow B 0.090506793
capacity ratio, λ
C 0.205670929
The calculation of the flow capacity ratio, λ, and
storativity ratio, ω, use the transition period observed in a 1.E+04
P
s
e
u
d
o
s
t
e
a
d
y
M
o
d
e
l

log-log plot of the dimensionless pressure derivative

C
o
n
s
t
a
n
t
f
l
o
w
c
a
p
a
c
i
t
y
r
a
t
i
o

1.E+03

versus dimensionless time which is characterized by a  1x10-3

trough and permits defining correlations to estimate the 1.E+02

above-mentioned parameters.
t D * PD'

1.E+01 
The maximum peak found on the pressure 1x10 0
9x10 -1
derivative was correlated to calculate, λ, This 1.E+00 5x10 -1
4x10 -1
characteristic points were chosen since it is independent of 1.E-01
3x10 -1
2x10 -1
ω as pointed out in Figure-9.
1.E+04 1.E-02
P
s
e
u
d
o
s
t
e
a
d
y
s
t
a
t
e

1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04

P
s
e
u
d
o
s
t
e
a
d
y
M
o
d
e
l

tD
C
o
n
s
t
a
n
t
F
l
o
w
c
a
p
a
c
i
t
y
r
a
t
i
o

tD * PD 'max
1.E+03
 1x10-3
Figura-10. Absence of the trough when ω > 0.1 in the dual-
porosity pseudosteady model, λ constant and yD=xD
1.E+02
t D * PD'

1.E+01 The range of application of Equation 61 is 1x10-3

 ≤ ω ≤ 0.1. This range applies when the transition period is
1x10 0
1.E+00
evident in the pressure derivative vs. time log-log because
F
i
r
s
t
-
P
s
e
u
d
o
s
t
e
a
d
y
s
t
a
t
e

5x10 -1
1x10 -1

1.E-01
6x10 -2
1x10 -2
for ω values greater than 0.1, see Figure-10, the minimum
1x10 -3
0
point (trough) is not seen which prevents finding a
1.E-02 characteristic point to develop an expression in these
1.E-04 1.E-03 1.E-02 1.E-01 1.E+00
tD
1.E+01 1.E+02 1.E+03 1.E+04
cases. It is important to clarify that the above developed
Figure-9. Effect of storativity ratio, ω, on the maximum point
equations were obtained considering yD=xD=1.
(peak) pressure derivative during the start of the transition period
for the dual-porosity pseudosteady model 2.3. TDS Formulation for Transient Dual-Porosity
Model
Using the dimensionless pressure derivative value 2.3.1 Initial Flow Regime
at the peak during the start of transition period (tD*PD’)max This is presented for all values of the flow
allows obtaining: capacity ratio, λ, and the storativity ratio, ω defined by
Equations 16 and 17 respectively, during the bilinear flow

7
regime on the pressure derivative curve as shown in tD * PD '  0.2052  tD (61)
Figures-10 and 11. This means that the initial flow is
independent of the λ and ω values.
1.E+07 Once the dimensionless quantities given by
Equations 19, 22 and 23 are replaced into Equation 61, it
T
r
a
n
s
i
e
n
t
D
u
a
l
-
P
o
r
o
s
i
t
y
M
o
d
e
l
1.E+06
Constant storativity ratio   = 0.1 )
yields;

P
s
e
u
d
o
s
t
e
a
d
y
r
e
g
i
m
e
1.E+05

1.E+04

k I h t *(P ) '
0.25
  0.0002637 k I t 
 0.2052   (62)
1.E+03

 ( ct ) I  x 2f 
0.1
t D * PD'

141.2 q f Bo o
0.01
1.E+02 0.001
0.0001  
B
i
l
i
n
e
a
r
F
l
o
w

1.E+01
s
l
o
p
e
0
.
2
5

1.E+00 Equation 62 allows solving for the inner-reservoir

1.E-01 permeability as follows;
1.E-02
1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 1
tD 11.60 q B   t BL 
0.25
 0.75
kI   f o o
 (63)
Figure-11. Effect of the flow capacity ratio, λ, on the bilinear  
 hI (t * P ') BL   ct  I  x 2f  
flow regime for the transient dual-porosity model, with ω = 0.1  

1.E+05
Notice that the hydraulic half-fracture length, xf,
can also be solved from Equation 63,
T
r
a
n
s
i
e
n
t
D
u
a
l
-
P
o
r
o
s
i
t
y
M
o
d
e
l

1.E+04
Constant flow capacity ratio  = 0.5 )

1.E+03
2
 11.60 q B   t  
0.25

x f   0.75 (64)
 
f o o
1.E+02 
 BL

 k I h (t * P ') BL   ct  I   
t D * PD'

 
0.5
1.E+01 0.1
0.01
B
i
l
i
n
e
a
r
F
l
o
w

0.001
s
l
o
p
:
0
.
2
5

0.0001
1.E+00
2.3.2. Pseudosteady-State Period
0.00001
0

1.E-01
This is recognized by a unit-slope straight line on
1.E-02 the pressure derivative curve. It is important to note that,
1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05
contrary to the dual-porosity pseudosteady-state model,
tD the transient model has only one pseudosteady-state period
Figure-12. Effect of the storativity ratio, ω, on the bilinear flow since this models assumes that kI = kf, therefore, the first
regiem for transient dual-porosity model, with λ = 0.5 pseudosteady-sate period is absent. Additionally, for this
model, the pressure transient behavior is not fully
The bilinear flow regime also takes place at about independent of both the flow capacity ratio, λ, and
the same period of time for the different yD values as storativity ratio, ω, since there exists a restricction in the
shown in Figure-13. Therefore, its behavior does not uniformity ranges of the model showing a variation of the
depend upon neither the variation of the dimensionless intersection point, b, as a function of ω for λ values
reservoir length, ye, nor the λ and ω parameters. greater than 0.5 or for the contrary case where the
1.E+08 intersection point, b, is a function of λ for ω values greater
T
r
a
n
s
i
e
n
t
D
u
a
l
-
P
o
r
o
s
i
t
y
M
o
d
e
l

1.E+07
than 0.5 as observed in Figure-14. In other words, the
Constant flow capacity ratio  = 1x10-4 )
1.E+06
developed equation for pseudosteady state in restricted to
Constant storativity ratio  = 0.1 )

1.E+05
ω and λ > 0.5. However, the uniformity is kept for the
1.E+04
remaining values of flow-capacity-, λ, and storativity-, ω,
values. This becomes more evident at the beginning of the
t D * PD'

1.E+03
yD
pseudosteady-state period as shown in Figures-15 and 16.
1.E+02
The behavior of the late pseudosteady-state
B
i
l
i
n
e
a
r
F
l
o
w

0.25
1.E+01
period is particular for different dimensionless reservoir
s
l
o
p
e
0
.
2
5

0.50
0.75
1.E+00 1
length values (yD= 0.35, 0.5, 075 and 1), as shown in
1.E-01
Figures-17.
1.E-02
1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07
As for the pseudosteady-state model, the
tD determination of the governing equation for the late
Figure-13. Effect of the dimensionless reservoir length, yD, pseudosteady period requires a log-log plot of tD*PD’ or
variation on the bilinear flow regime for the transient dual- tD*(mP)D’ versus tDA using a constant parameter of flow
porosity model, with constant values of λ and ω capacity ratio and storativity ratio (λ and ω constant) and
different dimensionless reservoir length values (yD= 0.35,
Then, at early time the flow is governed by the 0.5, 075 and 1), as shown in Figure-18.
following equation:

8
 1.E+05
1.E+08

P
s
e
u
d
o
s
t
e
a
d
y
r
e
g
i
m
e

T
r
a
n
s
i
e
n
t
D
u
a
l
-
P
o
r
o
s
i
t
y
M
o
d
e
l
T
r
a
n
s
i
e
n
t
D
u
a
l
-
P
o
r
o
s
i
t
y
M
o
d
e
l

P
s
e
u
d
o
s
t
e
a
d
y
S
t
a
t
e
-
R
e
g
i
m
e
u
n
i
t
s
l
o
p
e
1.E+04 1.E+07
Constant flow capacity ratio  = 1x10-4 )

u
n
i
t
s
l
o
p
e
Constant flow capacity ratio  = 0.5 )
1.E+06 Constant storativity ratio  = 0.1 )
1.E+03
1.E+05

1.E+02  1.E+04

t D * PD'

t D * PD'
1
0.5 1.E+03
1.E+01 0.1
0.01 yD
0.001
1.E+02
1.E+00 0.0001 0.25
0.00001 1.E+01 0.50
0 0.75
1.E-01 1.E+00 1

1.E-01
1.E-02
1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E-02
tD 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07
tD
Figure-14. Effect of the storativity ratio, ω, on the
Figure-17. Effect of the dimensionless reservoir length variation,
pseudosteady-state regime for the transient dual-porosity model,
yD, on the bilinear flow regime for the transient dual-porosity
with constant flow capacity ratio, λ = 0.5
model, with constant λ and ω
1.E+08
1.E+07

T
r
a
n
s
i
e
n
t
D
u
a
l
-
P
o
r
o
s
i
t
y
M
o
d
e
l

P
s
e
u
d
o
s
t
e
a
d
y
r
e
g
i
m
e
T
r
a
n
s
i
e
n
t
D
u
a
l
-
P
o
r
o
s
i
t
y
M
o
d
e
l

1.E+07

u
n
i
t
s
l
o
p
e
1.E+06 Constant flow capacity ratio  = 1x10 -4 )
Constant flow capacity ratio  = 1x10 )
-4 1.E+06 Constant storativity ratio  = 0.1 )
1.E+05 
1.E+05
1
0.9
1.E+04 1.E+04
0.5
t D * PD'

0.4 yD
1.E+03
1.E+03 0.3
P
s
e
u
d
o
s
t
e
a
d
y
r
e
g
i
m
e
t D * PD'

0.2 0.25
s
l
o
p
e
0
.
2
5

0.1 1.E+02 0.50

1.E+02 0.08 0.75
0.06 1.E+01 1
1.E+01 0.04
0.02 1.E+00
0.01
1.E+00 0.001 1.E-01
0.0001
1.E-01 0.00001 1.E-02
0 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 1.E+08

1.E-02 tD
1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07
Figure-18. Effect of the dimensionless reservoir length variation
tD
on the pseudosteady-state regime for the pseudosteady-state
Figure-15. Effect of the storativity ratio, ω, on the period for the dual-porosity transient model, with constant flow
pseudosteady-state period for the transient dual-porosity model, capacity ratio, λ, and constant storativity ratio, ω
with constant flow capacity ratio (λ = 1x10-4).
tD
1.E+07
t DA 
yD2 (65)
T
r
a
n
s
i
e
n
t
D
u
a
l
-
P
o
r
o
s
i
t
y
M
o
d
e
l

1.E+06
Constant storativity ratio   = 0.1 )
1.E+05
P
s
e
u
d
o
s
t
e
a
d
y
r
e
g
i
m
e

1.E+04
Then, it is possible to write a general

dimensionless pressure or pseudopressure derivative
1.E+03 0.1
expression for the pseudosteady-state period, as follows,
t D * PD'

0.01
1.E+02 0.001
0.0001

1.E+01
tD * PD 'PSS t 
 0.4933  D2  (66)
1.E+00
yD y 
1.E-01
 D  PSS
1.E-02
1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 After the dimensionless terms given by Equations
tD
19, 22, 23 and 65 are replaced in Equation 66, an
Figure-16. Effect of the flow capacity ratio, λ, on the bilinear
expression for the determination of the half- fracture
flow regime for the transient dual-porosity model, with constant
length, xf, is obtained,
storativity ratio (ω = 0.1)

In each case, a uniform behavior was found by 0.05770 Bo q f t PSS

xf  (67)
dividing the dimensionless time by the dimensionless hI  ct  I y t * ( P)'PSS
length of the stimulated reservoir volume for each
particular case, respectively. Notice that the hydraulic-fracture half-length, xf,
can be solved from Equation 59, since Equation 67 makes
possible the calculation of the distance y-coordinate.

9
 0.05770  q f t PSS 1.E+11
y (68)
hI  ct  I x f t * ( P)'PSS 1.E+10
t QPSS = 8.49 X10+5 hr
(t *P') MAX= 2.9465 X10+7psi

1.E+09 (t *P') QPSS = 2.838 X10 +7 psi (t *P') PSS = 6.527 X10+7 psi
4. EXAMPLES

 P, t *P ', psi

Two synthetic examples are worked out for the 1.E+08 t PSS = 1.751 X10 +9 hr

applicability of the above developed equations, one for 1.E+07

each model. Table-4 provides relevant information of the
1.E+06
reservoir, well and fluid properties used for the dual- (t *P') MIN = 6.4425 X10 +6 psi

porosity pseudosteady-state model. 1.E+05

t BL = 2636.43 hr
1.E+04
Table-4. Relevant information for example of the dual-porosity (t *P') BL= 7.531 X10+4 psi

pseudosteady model 1.E+03

1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 1.E+08 1.E+09 1.E+10 1.E+11 1.E+12 1.E+13 1.E+14
Parameter Value Parameter Value t, hr
h (ft) 200 kf, md 1x104 Figure-19. Log-log plot of the pressure and pressure derivative
rw (ft) 0.3 kI, md 1x10-2 vs. time for the dual-porosity pseudosteady model synthetic
example
μO (cp) 2.2 ko, md 2.0x10-3
βO (bbl/STB) 1.08 qf (bbl/day) 2500
It is also required to read the parameters during
ω 0.0052 λ 0.0083 the transition period which are useful for estimating  and
x (ft) 1325 y (ft) 1325 .
ϕf 0.45 ctf (1/psi) 3x10-1
ϕI 0.12 ctI (1/psi) 1x10-4 [t*P’]max = 2.9465x107 psi
ϕO 0.1 ctO (1/psi) 3x10-4 [t*P’]min = 6.4425x106 psi
xf (ft) 1325 wf (ft) 0.01
It is recommended to use the unit-slope
ηfD 88.9 ηoD 8 x10-2 pseudosteady state periods to estimate xf without needing
the inner reservoir permeability value which is carried out
For the case under consideration the pressure by using Equation 58 to provide xf = 1324.97 ft. An xe
drop and its derivative against time are reported in value of 2649.94 ft is found with Equation 30. Now, by
Figures-19 and 20 with the purpose of determining the taking a glance to Equation 55, which applies on the late
inner reservoir permeability or the half-fracture length time pseudosteady-state period, requires the value of b,
depending upon the model type, half-fracture length and which resulted to be 144.22 by using Equation 54.
reservoir length, which will be obtained by using the Applying this value into Equation 55 gives an xf value of
equations developed in this work. The other permeability 1344.76 ft and xe 2689.54 ft. The last value is the one
values could be estimated using the expressions provided recommended for the reservoir area/length.
by Brown et al. (2011) according to the diffusivity ratio. The estimation of kI uses the initial time period
The computations will be outlined as follows. which is dominated by linear flow regime (half slope).
Then, the use of Equation 47 provides a value of 0.013 md
4.1. Example 1. Pseudosteady Model and Equation 48 gives an xf value of 1332.18 ft.
Considerations given by Equations 35 through 37 Finally, application of Equations 60 and 61
should be taken into account in this model. The allows for the estimation of the λ and ω naturally-fractured
information provided in Table-4 and Figure-19 is useful parameters. For such purpose the peak and trough pressure
for the estimation of the inner reservoir permeability, kI, derivative values have to be converted to dimensionless
the reservoir length and half-fracture length. form. This is performed by using Equation 19 which
provides 70.260(max) and 15.363(trough), respectively.
Solution. As expected for this example, linear flow is Then, Equations 60 and 61 allow obtaining values of
observed at early times followed by the first pseudosteady- 0.0084 and 0.0054 for λ and ω, respectively.
state period taking place before the transition period
(trough). Finally, the late pseudosteady-state period is 4.2. Example 2. Transient Model
observed. The following information was read from It is required to find permeability, reservoir
Figure-19. length and half-fracture length from the data reported in
Figure-20 and the information given in Table-5.
tL = 2636.4334 hr
[t*P’]L = 7.5305x104 psi Solution. As expected for this example, only two flow
tqPSS = 8.4981x105 hr regimes are observed. First, linear flow regime which is
[t*P’]qPSS = 2.8389x106 psi followed by late pseudosteady-state period. The naturally-
tPSS = 1.7508x109 hr fractured typical transition period is absent. The below
[t*P’]PSS = 6.5267x107 psi parameters were read from Figure-20;

10
 Table-5. Relevant information for the Transient model of dual- the input data. This study did not consider the variations of
porosity the diffusivity parameters ηfD and ηoD.
Parameter Value Parameter Value 2. Contrary to the transient model, the pseudosteady state
h (ft) 200 kf, md 1x104 dual-porosity model displays the transition period (trough)
rw (ft) 0.3 kI, md 1x10-2 characteristic of naturally-fractured formations. Besides,
μO (cp) 2.2 ko, md 2.0x10-3 this model displays a unit-slope pseudosteady-state period
BO (bbl/STB) 1.08 qf (bbl/day) 2500
called here as “First pseudosteady state”. The transient
model has no transition period and the influence of  and
ω 0.0052 λ 0.0083
 is evident when this parameters are greater than 0.5.
x (ft) 725 y (ft) 350 Then, it is recommended to analyze the behavior of the
ϕf 0.45 ctf (1/psi) 3x10-1 ranges of  and  to find expressions to allow their
ϕI 0.12 ctI (1/psi) 1x10-4 estimations.
ϕO 0.1 ctO (1/psi) 3x10-4 3. The models were used assuming square reservoir
725 geometry. However, the expressions for the late
xf (ft) wf (ft) 0.01
pseudosteady-state period were corrected for rectangular
ηfD 88.9 ηoD 8 x10-2 geometries.
1.E+13

1.E+12 Nomenclature
A Drainage area, ft2
1.E+11
Bo Volumetric factor, bbl/STB
1.E+10 ct System total compressibility, 1/psi
 P, t *P ', psi

1.E+09
ctO Total compressibility of the inner reservoir,
(t *P')PSS = 6.503 X10+4 psi 1/psi
1.E+08
ctI Total compressibility of the outer reservoir,
1.E+07
(t *P') BL = 2.451 X10+8 psi t PSS = 9.656 hr 1/psi
1.E+06 ctf Total hydraulic compressibility, 1/psi
c Wellbore-storage coeffi cient, bbl/psi
1.E+05 t BL = 19101.89 hr
CD Dimensionless wellbore-storage coefficient
1.E+04 h Reservoir thickness, ft
1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 1.E+08 1.E+09 1.E+10 1.E+11 1.E+12 1.E+13 1.E+14
t, hr
kI Permeability of the inner reservoir, md
Figure-20. Log-log plot of the pressure and pressure derivative ko Permeability of the outer reservoir, md
vs. time for transient dual-porosity model synthetic example kf Hydraulic-fracture permeability, md
nf Number of main fracture planes
tL = 2636.4334 hr Pi Initial reservoir Pressure, psi
[t* (P)’]BL = 7.5305x104 psi P Laplace-space pressure
tPSS = 1.7508x109 hr Pwf Bottomhole flowing pressure, psi
q Laplace-space flow rate
[t*(P)’]PSS = 6.5267x107 psi
q production rate, STB/D, Mscf/D
It is recommended as a methodic procedure to rw wellbore radius, ft
use Equation 63 which applies on the late pseudosteady- t Time, hr
tD Dimensionless time
state period to obtain an xf value de 730.6 ft. A value of xe
t*m(P)’ Pseudopressure derivative, psi2/cp
equal to 1461.2 ft is found with Equation 30. Also, the
pseudosteady-state period is used to estimate the reservoir t*P’ Pressure derivative, psi
T Absolute Temperature, °R
distance in the y direction, Equation 68, which resulted to
wf Hydraulic-fracture width, ft
be 352.7 ft.
u Laplace space variable
Now, Equation 63 is used to find a kI value of
xe effective reservoir width, ft
0.0103 md and an xf of 754.98 ft is estimated with
xf Hydraulic half-fracture length, ft
Equation 64.
xeD Dimensionless reservoir size in x-direction.
5. CONCLUSIONS yeD Dimensionless reservoir size in y-direction.
1. New expressions for the interpretation of pressure- αf Hydraulic-fracture parameter used in trilinear-
transient tests in shale formations using both transient and flow model
pseudosteady-state models are obtained by using αO Outer-reservoir parameter used in trilinear-flow
characteristic points found on the pressure and pressure model
derivative versus time log-log plot. The equations were βF hydraulic-fracture parameter used in trilinear-
successfully tested with simulated examples so half- flow model
fracture length, inner-reservoir permeability and reservoir βO Outer-reservoir parameter used in trilinear-flow
dimensions were found with a very good agreement with model
ηf Hydraulic-fracture diffusivity, ft2/hr

11
ηI Inner-reservoir diffusivity, ft2/hr
ηo Outer-reservoir diffusivity, ft2/hr
ηfD Hydraulic-fracture-diffusivity ratio
ηoD Outer-reservoir diffusivity ratio
λ Flow-capacity ratio
ω Storativity ratio
Greeks
 Porosity, fraction
 Viscosity, cp
Suffices
o Crudo
i Initial
D Dimensionless
DA Dimensionless based on drainage area
BL Bilinear
L Linear
PSS Pseudosteady state
sc Standard conditions
TL Trilinear
qPSS First pseudosteady state regime
ACKNOWLEDGEMENTS
The authors gratefully thank the Most Holy
Trinity and the Virgin Mary mother of God for all the
blessing received during their lives.
REFERENCES
Bernal, K.M., Escobar, F.H., and Ghisays-Ruiz, A.
Pressure and Pressure Derivative Analysis for
Hydraulically-Fractured Shale Formations Using the
Concept of Induced Permeability Field. Journal of
Engineering and Applied Sciences. ISSN 1819-6608. Vol.
9. Nro. 9. Accepted for publication September 2014.
Brown, M., Ozkan, E. Raghavan, R. and Kazemi, H. 2011.
“Practical Solutions for Pressure-Transient Responses of
Fractured Horizontal Wells in Unconventional Shale
Reservoirs”. SPE Reservoir Evaluation & Engineering. p.
663-676. December.
Cinco-Ley, H., and Meng, H.Z. 1988. Pressure Transient
Analysis of Wells with Finite Conductivity Vertical
Fractures in Double Porosity Reservoirs. Society of
Petroleum Engineers. doi:10.2118/18172-MS. , January 1.
Escobar, F.H., Montenegro, L.M. and Bernal, K.M.
Transient-Rate Analysis For Hydraulically-Fractured Gas
Shale Wells Using The Concept Of Induced Permeability
Field. Journal of Engineering and Applied Sciences. ISSN
1819-6608. Vol. 9. Nro. 8. P. 1244-1254. August 2014.
Chen, C.-C. and Raghavan, R. 1997. A Multiply-Fractured
Horizontal Well in a Rectangular Drainage Region. SPE J.
2 (4): 455–465. SPE-37072-PA.
El-Banbi, A., H. and Wattenbarger, R.A. 1998. Analysis
of Linear Flow in Gas Well Production. Paper SPE 39972
presented at the SPE Gas Technology Symposium,
Calgary, Alberta, Canada.
Ge, J. and Ghassemi, A. 2011. Permeability Enhancement
in Shale Gas Reservoirs after Stimulation by Hydraulic
Fracturing. Paper AR11-514 presented at the Rock
Mechanics / Geomechanics Symposium, San Francisco,
CA.
Kazemi, H., Merrill, L. S., Porterfield, K. L., and Zeman,
P. R. 1976. Numerical Simulation of Water-Oil Flow in
Naturally Fractured Reservoirs. Society of Petroleum
Engineers. doi:10.2118/5719-PA. December 1.
Mukherjee, H., and Economides, M. J. 1991. A Parametric
Comparison of Horizontal and Vertical Well Performance.
Society of Petroleum Engineers. doi:10.2118/18303-PA. ,
June 1.
Serra, K., Reynolds, A. C., and Raghavan, R. 1983. New
Pressure Transient Analysis Methods for Naturally
Fractured Reservoirs(includes associated papers 12940
and 13014 ). Society of Petroleum Engineers.
doi:10.2118/10780-PA. December 1.
Tiab, D. 1993. Analysis of Pressure and Pressure
Derivative without Type-Curve Matching: 1- Skin and
Wellbore Storage. Journal of Petroleum Science and
Engineering. 12: 171-181.
APPENDIX A. GOVERNING EQUATIONS FOR
GAS FLOW DUAL-POROSITY PSEUDOSTEADY
MODEL
A.1. Initial Flow
 ω ≥ 0.1
The initial flow is characterized by a typical 0.25-
slope line on the pressure derivative curve, the behavior is
given by:
0.25
 t 
t D *( mPD ) '  0.4159  D
 (A.1)
 1   4 
 
Once the dimensionless terms given by Equations
19, 22 and 23 are plugged into Equation A.1 and solving
for the inner reservoir permeability, it yields,
12
 1
 237.097 q f T  t 
0.25
 0.75
kI    (A.2) A.2. Second Flow Regime or First Pseudosteady state
 
 (1   )h t * (mP) '   ct  I  x 2f   The general equation for the pseudosteady state
  with unit-slope line is:
From the above equation it is possible to find the t D *( mPD ) '  b t D (A.10)
value of the hydraulic half- fracture length, xf:
2 Since the displacement along the time axis is a
 237.097 q f T  t  
0.25
function of the storativity ratio, ω, is possible to develop
xf     
 (1   )k I0.75 h t * (mP) '   ct  I    an expression to obtain b as a function of it.
 
(A.3) 1

 0.01 ≤ ω < 0.1 b   A  B   C ( 0.5 )  (A.11)

The initial flow characterized by a typical 0.33- Constants A through C are given in Table-6.
slope line on the pressure derivative curve, the behavior is
Table-6. Constants for Equation (A.11)
given by:
Constant Value
t D *(mPD ) '  0.87847 tD0.33 (A.4) A 0.00361707
B 0.646008952
Once the dimensionless quantities expressed by C -0.00058982
Equations 19, 22 and 23 are replaced into Equation A.4,
the permeability of the inner reservoir; can be solved for: The range of application of Equation A.11 goes
for 0 ≤ ω < 1. The first pseudosteady-state regime can be
1
 259.0318 q T  0.33
 0.67 used for the computaion of , xf,
t  (A.5)
kI   f
  
 h t * (mP) '   ct  I  x 2f   1
   b 0.3755 q f T  t  2
xf      (A.12)
From the above equation it is also possible to  h ( ct  ) I  t * (P) '  
solve for the value of the half- fracture length, xf:
1 A.3. Pseudosteady-State Regime
 259.0318 q T  t  
0.33 0.66
Once the dimensionless terms given by Equations
x f   0.67 (A.6)
 
f

 k I h t * (mP) '   ct  I    19, 22, 23 and 65 are replaced in Equation 57, an
  expression for the determination of of the hydraulic half-
 0 < ω < 0.01 fracture length, xf , is obtained,

The initial flow is characterized by a typical 0.5- 0.5867 T q f t PSS

slope line on the pressure derivative curve which behavior xf  (A.13)
hI  ct   I y t *  (mP )'PSS
is governed by:
A.4. Dimensionless Storativity (ω) ratio and flow-
t D *(mPD ) '  5.35 tD (A.7)
capacity ratio (λ)

Once the dimensionless terms given by Equations The calculation of the parameters flow-capacity ratio (λ)
34, 35 and 36 are plugged into Equation A.7 and solving and storativity ratio (ω), is the same for goberning
for the inner reservoir permeability; expressions for oil model.
2
 388.69 q T   
0.5 APPENDIX B. GOVERNING EQUATIONS FOR
t (A.8)
kI   f
   GAS FLOW TRANSIENT DUAL-POROSITY
 h t * (mP) ' x f   ct  I    MODEL


Also, from the above equation the half-fracture B.1. Initial Flow
length, xf, can be solved for; The initial flow characterized by a 0.25-slope line
on the pressure derivative curve is governed by;
 388.66 q T  t 
0.5

xf  
f
   (A.9)
tD * PD '  0.2052  tD (B.1)
 h t * (mP) '   ct  I  k I  
 

13
 After replacing the dimensionless terms provided
by Equations 19, 22 and 23 in Equation B.1 and solving  tD * PD '  0 (C.7)
for the of the inner reservoir permeability, we obtain:
1 tD * m( PD ) '  0 (C.8)
 116.98 q T  t 
0.25
 0.75
kI   f
 (B.2)
 
 hI t * (mP ) '   ct  I  x 2f   For pressure,
 
 1     yeD 
From the above equation it is possible to find the PD     sc (C.9)
value of the hydraulic half- fracture length, xf: 2 yeD  6 3CFD

 116.98 q T
2
If sc=0
 
0.25
 t (B.3)
x f   0.75 f
    
 k I h t * (mP) '   ct  I    141.2 q f  o    1      ye / y F   
  kI      (C.10)
hI P  2  ye / y F   6 k F wF 
 3 
B.2. Pseudosteady-State Regime  k I xF 
After plugging the dimensionless terms given by
Equations 19, 22, 23 and 65 in Equation 57, an expression
 
for the determination of xf is obtained, 1424 q f T   1    
  ye / y F   (C.11)
kI     
hI  mP   2  ye / yF   6 k F wF 
0.5819 T q f t PSS  3 
xf  (B.4)  k I xF 
hI  ct   I y t * (mP)'PSS

The slope of straight line on log PD vs. log tD and

APPENDIX C. GOVERNING EQUATIONS FOR
log [tD*PD’] vs. log tD is 0 and -∞ respectibility.
ASYMPTOTIC SOLUTIONS FOR PRESSURE AND
DERIVATIVE: PSEUDOSTEADY DUAL-
POROSITY MODEL  tD * PD '  0 (C.12)

C.1. Early Time tD * m( PD ) '  0 (C.13)

The slope of straight line on log PD vs. log tD and For pressure,
log [tD*PD’] vs. log tD is 1.
1
tD   1   4

 tD * PwD '  (C.1) PD     sc (C.14)

CD 2CFD   

0.04167 q f Bo ht t (C.2) If sc=0

C
hI (t * P ')  
 1/ 4 
141.2 q f  o    1   
0.4202q fgTt t kI     (C.15)
C (C.3) hI P k F wF    
hI t * (mP ')  2



 k I xF 
For pressure;  
 1/ 4 
1424 q f T   1  
tD kI    (C.16)
PD  (C.4) hI m( P)  k F wF    
CD  2 
 k I xF 
 
0.04167 q f Bo ht t
C (C.5)
hI P The slope of straight line on log PD vs. log tD and
log [tD*PD’] vs. log tD is 0 and -∞ respectibility.
0.4202q f Tt t
C (C.6)  tD * PD '  0 (C.17)
hI m( P)

C.2 Intermediate Time tD * m( PD ) '  0 (C.18)

The slope of straight line on log PD vs. log tD and
log [tD*PD’] vs. log tD is 0 and -∞ respectibility.

14
 For pressure, 
1/ 0.75

 1/ 4 
 570.0821 q f T  t  
 1   kI      (C.27)
PwD    sc
 2k f w f  5 / 4 h m( P )   c TI  x f
(C.19) 2

2  3CFD 
 k x   I 
 I f 
If sc=0
  The slope of straight line on log PD vs. log tD and
221.796 q f Bo o  1     log [tD*PD’] vs. log tD is is (1/4).
kI     (C.20)
hI P   k
3 F F
w 
   t1/D 4
 k I xF   tD * PwD '  (C.28)
4 2CFD   5 / 4  1/ 4

 
2236.814 q f T  1    1/ 0.75
  
kI     (C.21)  1/ 4 
hI m( P )   k F wF   14.132 q f Bo o  t  
3 kI      (C.29)
 k I xF   2k f w f  5 / 4 h (t * P ')   TI
  
2
 c x  
 
f
 k x I 
 I f 
The slope of straight line on log PD vs. log tD and
log [tD*PD’] vs. log tD is 0.25. 1/ 0.75
 
 1/ 4 
 142.521 q f T   
 t1/D 4 t (C.30)
 tD * PD '  (C.22) kI   
    
 
 5 / 4
2
 2 k w  c x 
4 2CFD   
 k x   5 / 4  hI t *  ( mP ') 
f f TI f

 I f 
1/ 0.75
 
 1/ 4  For
 14.132 q f Bo o  t   pressure,
kI      (C.23)
 2k f w f  5 / 4 h (t * P ')   TI  x f
 2
c  
 k x   I   t1/D 4
 I f  PwD   sc (C.31)
2CFD   5 / 4  1/ 4
1/ 0.75
 
 1/ 4  If sc=0
 142.521 q f T  t  
kI     (C.24)

 2 k w 
   c   x 2f    
1/ 0.75

 k x
f f
  5 / 4  h  t *  ( mP ')  TI
  1/ 4 
I
 I f   
 56.5278 q f Bo o t 
kI      (C.32)
  TI
  
2
For pressure,  2k f w f c x  
 k x   5 / 4  hI P
f

 I f 
 t1/D 4
PwD   sc (C.25)
2CFD  5 / 4 1/ 0.75
 
 1/ 4 
 570.0821 q f T  t  
kI      (C.33)
If sc=0  2 k w 
   c   x 2
 
 
 k x   5 / 4  hI m( P )
f f TI f

1/ 0.75
 I f 
 
 1/ 4 
The slope of straight line on log PD vs. log tD and
 56.5278 q f  o   t  
kI      (C.26) log [tD*PD’] vs. log tD is 0.5.
 2k f w f  5 / 4 h P   c TI  x f
2
 
 k x   I 
  1  tD
I f
 tD * PD '  (C.34)
2 

2
 2.03205 q B  t 
kI   
f o o
(C.35)
 hI (t * P ')

 c T I
xf  
2


15
 2
 tD * PD '  2 tDA (C.46)
 20.4932 q T t 
kI   
f
(C.36) 0.23395 q f Bo  
 hI t * m( P) '  c TI  x f  
2
A
t
   (C.47)
 T I 
 c h
I
(t *  P ') 
For pressure,
2.3954 q f T  t 
 tD  A   (C.48)
PD  
 3CFD
 sc (C.37)  c TI  hI  t * m( P) ' 

If sc=0 and CFD → ∞, For pressure,

2  yeD 
 4.0641 q B  t  PD  2 t DA    sc (C.49)
kI   
f o o
(C.38) 6 3CFD


hI P  c T I
 x 2f  

 yeD 1 s
2 When t DA    c
 40.9864 q T t  12 6CFD 2
kI   
f
(C.39)
 hI  mP   c T x   2
 I  f
0.23395 q f Bo  t 
A
 c TI hI  P 
(C.50)
The slope of straight line on log PD vs. log tD and
log [tD*PD’] vs. log tD is 0.5.
2.3954 q f T  t 
1 A
 c T  hI  m( P) 
(C.51)
 tD * PD '   tD (C.40)
2 I

2
 2.03205 q B   APPENDIX D. GOVERNING EQUATIONS FOR
t
kI    ASYMPTOTIC SOLUTIONS FOR PRESSURE AND
f o o
(C.41)
 hI (t * P ')

 c T I
 x 2f 
 DERIVATIVE: TRANSIENT DUAL-POROSITY
MODEL
2
 20.4932 q T t  D.1 Early Time
kI   
f
(C.42) The slope of straight line on log PD vs. log tD and
 hI t * (mP ')  c T xf 
2
 I  log [tD*PD’] vs. log tD is is 1.

For pressure, tD
PD  (D.1)
CD

PD   t D   sc (C.43)
3CFD 0.04167 q f Bo ht t
C (D.2)
hI P
If sc=0 and CFD → ∞
0.04167q f Bo ht t
2 C (D.3)
 4.0641 q B  t  hI (t * P ')
kI   
f o o
(C.44)


hI P  c TI  x f 
2
D.2 Intermediate Time
The slope of straight line on log PD vs. log tD and
2 log [tD*PD’] vs. log tD is 1/8.
 40.9864 q T t 
kI   
f
(C.45)
 hI  mP   c T xf 
2 1
 t1D8
 tD * PD '  
3 8
 I  9 9 
(D.4)
    8 2C fD (9 8)
C.3 Late Time
The slope of straight line on log PD vs. log tD and
log [tD*PD’] vs. log tD is 1.

16
 8
 7 1
    0.75
 1   
 19.7940 q fo Bo o   8

3t 1 
kI       142.5205 q f g T   
 9 9  4
 ( ct ) o x f    
  
2
3t
kI   
h k w
2  f f  (9 8) (t * P ')     
t Ig  g x f
2
    ( c ) 
 I     h 2 k f w f  (5 4) t *  (mP ') 
k x
 I f 
   I    
(D.5) k x
 I f 
 
(D.12)
8
  7
For pressure,
 
 1 
 199.622 q f g T  3t  8
 t1D 4
kI      PwD   Sc (D.13)
 9 9 
 ( ct ) Ig  g x f    
2
   2C fD (5 4)
k w
h 2  f f  (9 8) t * m( P ') 
 I   
k x
 I f  If sc=0
 
(D.6)
1

For pressure,   0.75

 
 1 
1  56.5278 q fo Bo o  t  4
 3  8
 t1D8 kI      (D.14)
PD   9 9   Sc (D.7)  2 
( ct ) o x f  
   2C fD (9 8)    
k w
h 2  f f  (5 4) (P ) 
 I   
If sc=0   kI x f  

1
8
 7   0.75
   
  1 
1 
 158.3519 q fo Bo o   8  570.0820 q f g T  t  4
3t kI     
kI   
   (D.8)
9 9   ( c )  x 2 
   ( c )  x 2
       t Ig g f  
 t o f  k w
h k w
2  f f  (9 8) (P)  h 2  f f  (5 4) (mP) 
 I     I   
k x k x
 I f 
  I f    
(D.15)
8
  7 The slope of straight line on log PD vs. log tD and
  log [tD*PD’] vs. log tD is 0.25.
 1 
 1596.98 q f g T  3t  8

kI       t1D 4
 9 9 
 ( ct ) Ig  g x f     t 
2
   * PwD'  (D.16)
k w D
(5 4) (1   )1 4
h 2  f f  (9 8) m( P)  4 2C fD
 I   
k x
 I f 
 
1
(D.9)   0.75
 
The slope of straight line on log PD vs. log tD and log  1 
 14.1319 q fo Bo o  t  4
[tD*PD’] vs. log tD is 0.25. kI   
    ( c )  x 2
(1   )  
k w  t o f 
 t1D 4
h 2  f f  (5 4) (t * P ') 
 tD * PD '  (D.10)  I   
(5 4) k
 I f x
4 2C fD  
(D.17)
1
  0.75
 
 1 
 14.1319 q fo  t  
4

kI      (D.11)
 2 
 
k w
  ( ct ) o x f  
h 2  f f  (5 4) (t * P ') 
 I   
k x
 I f 
 

17
 1 
  0.75 If sc=0 and CfD ∞ 0
  3 C fD
 1  1
 142.5205 q f g T   4
 28.2639 q B  
t  4 
1 0.75
kI     
3t
 kI   (D.26)
fo o o
    ( ct ) Ig  g x f (1   )  
2
 
 hI (5 4) (P )  ( ct ) o x f ( )  
2
k w
 h 2  f f  (5 4) t *  (mP ')   
 I    
k x
 I f 
 
1
(D.18)  285.041 q T   4 
1 0.75
 fg 3t (D.27)
kI   
For pressure,  hI (5 4) (mP ) ( ct ) Ig  g x f ( ) 
2

   

 t1D 4 (D.19) The slope of straight line on log PD vs. log tD and
PD   Sc
2C fD (5 4) (1   )1 4 log [tD*PD’] vs. log tD is 0.5, and if sc 0 and CfD ∞

If sc=0, 1
 tD * PD '   tD (D.28)
2
1
  0.75
   2.032 q B   1

2

 1  t  2

kI    (D.29)
fo o o
 56.5278 q fo Bo o   4  
t  hI (t * P ')  ( ct ) o x 2f 
kI   
    ( c )  x 2
(1   )      
k w  t o f 
h 2  f f  (5 4) (P ) 
 I  k I x f  
 20.4932 q T  1

2
    t  2

kI    (D.30)
fg
(D.20)  
 hI t * m( P )'  ( ct ) Ig  g x 2f 
   
1
  0.75 For pressure,
 
 1 
 570.0820 q f g T  t  4

kI    PD   t D   Sc (D.31)
    ( c )  x 2
(1   )   3 C fD
k w  t Ig g f 
h 2  f f  (5 4) m( P ) 
 I   

k x
 I f   
If sc=0 and CfD ∞ 0
(D.21) 3 C fD
2
 4.0641 q B   
1

The slope of straight line on log PD vs. log tD and log t
2

kI    (D.32)
fo o o
 
[tD*PD’] vs. log tD is 0.25, and if sc 0 and CfD ∞  hI (P) ( c )  2
o xf

  t  
1
  3t D  4
 tD * PD '  (D.22) 2
   40.9864 q T  
1
2 
8(5 4)    t
kI    (D.33)
fg
 
 hI m( P)  ( ct ) Ig  g x 2f 
   
1
 7.066 q B    4 
1 0.75
3t
kI   (D.23) The slope of straight line on log PD vs. log tD and log
fo o o
 
 hI (5 4) (t * P ')  ( ct ) o x f (  )  
2

  [tD*PD’] vs. log tD is 0.5,

1 1 3 t D
 71.2603 q f g T   4 
1 0.75  tD * PD '  (D.34)
 3t (D.24) 2 yeD 
kI   
 hI (5 4) t * m( P ')  ( ct ) Ig  g x 2f (  )  
    2
 3.5196 q B    2 
1
t
kI   (D.35)
fo o o
For pressure,  
 hI ye (t * P ')  ( ct ) o ( )  
 
1/ 4
  3tD   (D.25)
PD      Sc  35.4953 q T  1

2
2 (5 4)  (  )  3 C fD t  2

kI    (D.36)
fg
 
 hI ye t * (mP ')  ( ct ) Ig  g (  ) 
   

18
 2 t DA
For pressure,  tD * PD '  (D.46)
1 

1 3 t D  yeD  0.23395 q f Bo 
PD     Sc (D.37) t 
 A
 c T hI (1   )  (t * P ') 

yeD 6 3 C fD (D.47)
I

If sc=0
  2.3594 q fg T  t 
  A   (D.48)
141.2 q f    x k I t  ye    c T  g hI (1   )  t * m( P ) ' 
P  0.08635 f    I

k I hI  ye   6 xf 2k w 
 3 f f 
 k I x f  For pressure,

(D.38)
2 t AD  yeD 
  PD     sc (D.49)
  1  6 3CFD
1424 q fg T  x k I t  ye  
mP  0.08635 f   
k I hI  ye   6 xf 2k f w f   yeD 1 s
 3   When t DA    c

 k I x f  12 6CFD 2
(D.39)
0.23395 q f Bo  t 
A
 c TI hI (1   )  P 
The slope of straight line on log PD vs. log tD and (D.50)
log [tD*PD’] vs. log tD is 0.5, and if sc 0 and CfD ∞

1  tD 2.3594 q f T  t 
 tD * PD '  (D.40) A
 c T  hI (1   )  m( P) 
2 (1   )
(D.51)
I

2
 2.032 q B   t  2 
1

kI   (D.41)
fo o o
 
 hI (t * P ')  ( ct ) o (1   )  
 

2
 20.4932 q T  
1
2 
t
kI    (D.42)
fg
 
 hI t * (mP ')  ( ct ) Ig  g (1   )  
 

For pressure,

 tD  (D.43)
PD    Sc
(1   ) 3 C fD


If sc=0 and CfD ∞ 0
3 C fD
2
 4.0641q B    2 
1
t
kI   (D.44)
fo o o
 
 hI P  ( ct ) o (1   )  


2
 40.9864 q T  
1
2 
t
kI    (D.45)
fg
 
 hI (mP ) ( ct Ig g (1   )
)  
   

D.3. Late Time

The slope of straight line on log PD vs. log tD and
log [tD*PD’] vs. log tD is 1.

19