ADVANCED

WELL TEST ANALYSIS

D. TIAB, Professor
University of Oklahoma
[email protected]
405 325 6777
 2

5.1
REAL-GAS PSEUDO-
PRESSURE
 3

The solution to the diffusivity equation, discussed in Chapter 1, is

applicable for slightly compressible fluids (i.e. liquids) with relatively
constant fluid properties such as viscosity.

However for flow of real compressible gases in the reservoir, the

solution to the diffusivity equation requires adjustment because the
physical properties, such as viscosity, isothermal compressibility and
the gas deviation factor, are strong functions of pressure.

5.1 – REAL-GAS PSEUDO-PRESSURE

The diffusivity equation for gas flow through porous media is:

1 ∂ P ∂P φ ∂ ( P / Z )
(r )= (5.1.1)
r ∂r µZ ∂r k ∂t
 4
This equation was derived by combining the conservation of mass
equation, Darcy's law and the following equation of state for gas
density:

PM
ρ= (5.1.2)
ZRT

Where:
R = Universal gas constant, 10.732 psia ft3/lb-mole/oR
M = Molecular weight of gas, lbm/lb-mole
Z = Gas deviation factor
P = pressure, psia
T = Reservoir temperature, °R

When the left-hand side of Eq. 5.1.1 is differentiated, the term

(∂P/∂r)2 is not negligible as in the case of slightly compressible fluids.
Therefore, Eq. 5.1.1 is a non-linear partial differential equation.
 5
Al Hussainy et al. introduced a transformation called the real gas
pseudo-pressure so that the pressure, the viscosity and gas deviation
factor could be combined in a way that allows the linearization of Eq.
5.1.1.

The real gas pseudo-pressure is defined as follows:

P
m( P ) = 2 ∫
P (5.1.3)
dP
Pn µ ( P ) Z ( P )

This transformation accounts for variations in gas properties with

pressure. P0 is an arbitrary reference pressure. From Eq. 5.1.3, we
have:

∂m( P) ∂m( P) ∂P 2 P ∂P
= =
∂t ∂P ∂t µ g Z ∂t (5.1.4)

∂m( P) 2 P ∂P
= (5.1.5)
∂r µ g Z ∂r
 6
The substitution of equations 5.1.4 and 5.1.5 into Eq. 5.1.1 gives the
following diffusivity equation:

∂ 2 m(P ) 1 ∂m(P ) φµ g ct ∂m(P ) (5.1.6)

+ =
∂r 2
r ∂r k ∂t
Eq. 5.1.6 has exactly the same form as the diffusivity equation for
slightly compressible fluids (Eq.1.1.2), with the pressure P replaced
with the pseudopressure m(P).

The close analogy between equations 1.1.2 and 5.1.6 suggests that
the solution for the flow of real gases should correlate as functions of
dimensionless time based on initial or average values of viscosity and
compressibility.

Define the dimensionless time tD and dimensionless pseudo-pressure

drop as:
tD =
0.0002637kt
m D (rD , t D ) =
kh
[m( pi ) − m( p)]
φµ g ct rW 2 1422.7 q sc T

(5.1.7) (5.1.8)
 7
Where:

qsc = gas flow rate, MSCF/D

T = reservoir temperature, oR
m(p) = pseudopressure,psia2/cp.
Psc = 14.7 psia, pressure at standard conditions.
Tsc, = 520 °R (60 °F), temperature at standard conditions.

The total system compressibility is:

ct = c g S g (5.1.9)

The gas compressibility, cg, can be expressed in terms of Z as,

1 1 dZ 1
cg = − ≅ (5.1.10)
P Z dP P
Combining equations 5.1.7, 5.1.8 and 5.1.6 yields:

∂ 2 mD (P ) 1 ∂mD (P ) ∂mD (P )
+ = (5.1.11)
∂rD2 rD ∂rD ∂t D
 8

Eq. 5.1.11 is identical in format to Eq. 1.1.11 for slightly compressible

fluids.

Wattenbarger and Ramey showed that the pressure transient

equations could be used, with very good approximation, in terms of
m(P).

Then the continuous line source solution given in section 1.1 is

applicable to gas well testing if the pressure is replaced by m(P), i.e.

1  rD 
2

m D ( P ) = PD = − Ei −  (5.1.12)
2  4t D 

Using the log-approximation to the Ei-function at the flowing well,
Eq. 5.1.12 becomes:

m D ( Pwf ) = PD =
1
[ln t D + 0.80907 + 2S '] (5.1.13)
2
 9
where S' is the total skin factor that includes the skin resulting from
true formation damage or stimulation, S, and D is turbulence or non-
Darcy coefficient:

S ' = S + Dqsc (5.1.14)

The effect of D is to create an increasing apparent skin for higher

production rates.

The value of D varies with pressure but for simplicity it can be

considered constant as an acceptable approximation.

In real units Eq. 5.1.13 becomes:

1422q scT   kt  
m( Pwf ) = m( Pi ) − log  − 3.23 + 0.869S '
  φµ g ct rw 
2 (5.1.15)
kh  

Thus a plot of m(Pwf) versus time will yield a straight line portion
corresponding to the infinite-acting radial flow regime of slope mP,
which can be used to calculate permeability:
 10
1422qscT
mP = − (5.1.16)
kh
The total skin factor S’ is then calculated from:
 m( P ) − m ( P )  k  
S ′ = 1.151 i 1hr
− log  + 3.23
(5.1.17)
 mP  φµ c r 2  
 g t w 

The reduced pseudo-pressure at time t = 1 hour, m(P1hr), should be

obtained from the semilog straight line portion (extrapolated if
necessary).
1.20E+09

1.10E+09

m(P), psi2/cp 1.00E+09

Figure 5.1.3 – Semilog 9.00E+08

plot of m(Pwf) vs. time 8.00E+08
m=-5.16E+07
7.00E+08 m(P)1hr=7.8E+08

6.00E+08
0.01 0.1 1 10 100 1000

time, hr
 11

5.1.1 – NON-DARCY FLOW

Darcy’s law applies to gases flowing at low rates (laminar

flow), which occur in the reservoir away from the wellbore.
However, in the vicinity of the wellbore, the flow rates of gas
can become extremely high because of the converging flow.

At these rates, inertial effects can become important and

Darcy’s law no longer applies.

The inertial effects take the form of distorted flow paths and
also turbulence in different locations in the pore structure.

Although the exact nature of this microscopic flow is not

known in the reservoir, the net effect is a higher-pressure
gradient when these inertial effects become considerable [6].
 12
For Laminar flow, Darcy's law can be rearranged in the following form:

∂P µ
=− u (5.1.18)
∂x k

where ∂ P/ ∂ x is the pressure gradient and u is the macroscopic

(Darcy) fluid velocity.

At higher rates, when the inertial effects become important, a

quadratic relation first introduced by Forchheimer [7] is used:

∂P µ
− = u + β ρ u2 (5.1.19)
∂x k
The right hand side of Eq. 5.1.19 (Forchheimer equation) contains a
term for viscous forces and a term for inertial forces, both of which
contribute to the pressure loss.

The parameter β (ft-1) is called either the Forchheimer turbulence

factor or inertial factor and can be determined experimentally or
estimated from one of the following correlations:
 13

4.851 × 104 1.88 × 1010 4.11 × 1010

β= β = 0.53 1.47 β=
φ 5.5 k φ k k4/3
(5.1.20) (5.1.21) (5.1.22)

Eq. 5.1.22 is applicable only to sandstones, dolomites and limestone.

The parameter D (D/Mscf) in Eq. 5.1.14 is called the non-Darcy flow

coefficient or turbulence coefficient.

It can be estimated from:

M = ∑ yi M i
γ kβ
−15 g γg =
M
D = 2.22 × 10 (5.1.23) 28.97 (5.1.24)
µ g rw h (5.1.25)

Where γg = the specific gravity of gas

M = apparent molecular weight of a gas mixture.

Mi is the molecular weight of each component and yi is the mole

fraction of a component with a molecular weight Mi.
 14
The non-Darcy coefficient D and mechanical skin S can be determined
more accurately if two consecutive flow tests are run at two different
rates, yielding:

S1' = S + Dqsc1
(5.1.26)
S 2' = S + Dqsc 2
Solving these equations simultaneously will give the values of S
and D.

As discussed above, accuracy is improved for semilog analysis tests

by replacing pressure with the real-gas pseudopressure function,
m(p), which is expressed in psi2/cp.

For type-curve analysis, particularly of wellbore-storage-distorted

data from both flow and buildup tests, accuracy also is improved by
replacing time with adjusted pseudotime, tap(p), which is expressed
in hr-psi/cp:
t
dt
t ap = ∫ (5.1.27)
0
µ g ( P )c t ( P )
 15
Pseudotime may influence the behavior of m(p) in closed reservoirs
of moderate to high permeability.

For convenience, although not by necessity, m(P) and tap can be

normalized to have units of psia and hours, respectively, like the
original variables, P and t.

Normalization also gives the pseudo-pressure and pseudo-time

variables magnitudes comparable with those of the untransformed
pressure and time; the unnormalized variables m(P) and tap typically
have values of 105 to 108.

Reference values of pressure used for normalization are arbitrary.

Some engineers prefer properties evaluated at initial reservoir

pressure, others prefer at the average reservoir pressure.
 16
Lee and Wattenberger define normalized adjusted pressure, Pan, and
normalized adjusted time, tan, as:

 µg Z 
Pna =   m( P ) (5.1.28) t na = ( µ g ct )t ap (5.1.29)
 2P 
 

For a new well: P ≅ P* = Pi

In terms of adjusted variables, the unsteady state equation for gas

flow is:
162q sc B g µ g   kt  
(Pna )wf = (Pna )i − log  − 3.23 + 0.869S ' (5.1.30)
  φµ g ct rw 
2
kh  

Where the gas formation volume factor at standard conditions of

pressure (14.7 psia) and temperature (60 oF), Bg (bbl/SCF), is
calculated at the average or initial properties of Z, T and P from:

ZT
B g = 0.00504 (5.1.31)
P
 17
Note that the product qscBg is in bbl/D, similarly to the product qBo in
Eq. 1.1.24.

The similarity between equations 1.1.24 and 5.1.30 indicates that all
the methods discussed above for interpreting a pressure transient
test in an oil well are also applicable in a gas well.

For semilog analysis of buildup tests, adjusted pressure and adjusted

time should be used, but the adjusted producing time, tap used in the
Horner time ratio is evaluated at current average drainage area
pressure, and therefore tap = tp.

Adjusted shut-in time, ∆tna is evaluated from the integral, i.e.

∆t
d∆t
∆t na = ( µ g ct ) ∫ (5.1.32)
0
µ g ( P )c t ( P )

With µg and ct evaluated at shutin bottom hole pressures, Pws, at

values of shutin times ∆t.

For semilog analysis of a drawdown test, adjusted pressure and

actual flowing time should be used.
 18

5.1.3 – Pressure and Pressure-squared

Use of adjusted time and ad-justed pressure in formulating

equations for analysis of transient tests in gas wells is not always
necessary.

Pressure data may be graphed in terms of P2 rather than m(P) if

reservoir pressures are less than 2000 psi.

Also, pressure data may be plotted directly in terms of P if the

reservoir pressures are above 3500 psi.

If in doubt whether to plot P, P2 or the normalized adjusted m(P),

use m(P).

Wattenbarger and Ramey have proposed a simple method to

determine the best mode for plotting the data.

They suggest that the variation of the product µgZ with P examined
over the pressure range of interest.
 19
Figure 5.1.1 is a graph of the
variation in the product µZ with
pressure and is typical of many gases.

If the variation in the µgZ product is

small, then they suggest that a P2-
graph would suffice.

On the other hand, if the µgZ product

is a linear function of pressure, then a
P-graph would be adequate
regardless of the magnitude of the
pressure gradient.

If neither of these relations is

applicable, then the m(P) approach
should be used.

Thus, the procedure for graphing data

depends primarily on the variation of
the product µgZ.
 20
At high reservoir pressures, i.e. P>3500 psi, the plot of µgZ versus
pressure is almost a straight line, thus:

P Pi P
≅ ≅ (5.1.33)
µ g Z µ gi Z i µ g Z

Thus for P0 = 0 Eq. 5.1.3 can be expressed as:

2P
P
 2P 
m( P ) = ∫ dP =  P (5.1.34)
µg Z 0 µ Z 
 g 

Thus, the normalized adjusted real pseudo-pressure, Eq. 5.1.28,

becomes:

 µg Z  µ Z  2 P 
Pna =   m( P ) =  g  P = P (5.1.35)
 2P   2P  µ Z 
    g 
and the unsteady state equation 5.1.30 becomes:
162.6q sc B g µ g   kt  
Pwf = Pi − log  − 3.23 + 0.869S ' (5.1.36)
  φµ g ct rw 
2
kh  
 21
At low pressures, i.e. less than 2000 psi, the product µgZ is almost
constant, as can be observed in Figure 5.1.1. Thus:

µ g Z ≅ µ gi Z i ≅ µ g Z (5.1.37)

Thus for P0 = 0, Eq. 5.1.3 can be expressed as:

P
 1  2
m( P ) =
2
µg Z ∫P Pd P =  µ g Z P
 (5.1.38)
n  

The normalized adjusted pressure, Eq.5.1.28 becomes:

 µg Z  µ Z  P2  1  2
Pna =   m( P ) =  g  = P (5.1.39)
 2P   2P  µ Z  2P 
    g
And the unsteady state equation 5.1.30 can be written in terms of
pressure-squared:

1637q scTµ g Z   kt  
P = Pi −
2 2
log  − 3.23 + 0.869S ' (5.1.40)
  φµ g ct rw 
wf 2
kh  
 22
In summary, the three fundamental equations used to analyze the
pressure behavior of a gas well are:

 µg Z  q scT
Pwf = Pi − 1422
 2P

 kh D D
[P (1, t ) + S '] (5.1.41)
 

Pwf2 = Pi 2 − 1422(µ g Z ) [PD (1, t D ) + S ']

q sc T (5.1.42)
kh

m( Pwf ) = m( Pi ) − 1422
q sc T
[PD (1, t D ) + S '] (5.1.43)
kh
Where, at the well rD = 1 and Eq. 1.1.21 becomes:

    1
PD (1, t D ) = − ln −  + 0.5772 = [ln (t D ) + 0.80907 ]
1 1
(5.1.44)
2   4t D   2

Eq. 5.1.41 can be converted to Eq. 5.1.36 by observing that the gas
flow rate is expressed in MSCF/D and the constant 162.6 is actually
((1422/2/0.00504/2)*2.303)/1000.
 23
Carter has suggested that the average values of viscosity and gas
deviation factor should be obtained at the average reservoir pressure
over the duration of the test, i.e. for a drawdown test:

Pi + Pwf
P= (5.1.45)
2
Or, for the pressure-squared approach:

Pi 2 + Pwf2
P= (5.1.46)
2

Similarly, for a pressure buildup test:

Pws (∆t = 0) + Pws

P= (5.1.47)
2

Where Pwf is the flowing well pressure at the end of the test and Pws
is the shutin well pressure at the end of the buildup test.
 24
In order to test the accuracy of these three methods, Aziz et al.
calculated the sandface pressure for twenty five different gas well
tests conditions from reservoirs in Alberta, Canada.

They concluded that:

(1) the use of average gas properties in the definition of PD and tD
results in more accurate solutions than the use of initial conditions.
This is true regardless of the approach used;

(2) When gas properties must be assumed constant at the initial

values, the use of the m(P)-function yields the most reliable results;
and

(3) In very low permeability gas reservoirs, i.e. tight gas reservoirs,
the m(P)-function must be used, especially if the reservoir is
produced at very high rates.

Again, all of the rules developed in Chapter 1 to identify near wellbore

effects (skin flow efficiency and wellbore storage coefficient) and
the start of the semilog straight line for oil wells are also applicable
for gas wells, regardless of which function is used.
 25
5.1.4 – CONVERTING P TO m(P)

There are several methods for converting pressure data to real gas
pseudo-pressure data:

(a) the areal summation method, which involves plotting 2P/µZ

versus pressure, the varea under the curve is the value of m(P);

(b) the reduced properties correlation, which requires tables or

charts; and

(c) the numerical integration method, which the more common

method and will be illustrated by the following example. The step-
by-step procedure of the numerical integration method is:
1. Obtain viscosity and gas deviation factor. If PVT data are
not available, empirical correlations can be applied.

µ Z).
2. For each pressure value, calculate 2P/(µ

3. Estimate the mean value between the former and current

value of 2P/(µµ Z):
 26
 2P   2P 
  +  
 2P   µ Z i  µ Z  i −1
  =
 µ Z  mean 2

4. Estimate the pressure difference between the former and

current pressure value:

∆P = Pi −1 − Pi

5. Multiply results from steps 4 and 2. Thus,

*
 2P   2P 
  =   ∆P
 µ Z  i  µ Z  mean

6. Estimate pseudo-pressure as:

*
 2P 
m( P ) i =   + m( P ) i −1
 µ Z i
 27
EXAMPLE 5.1.1
P, psi µg, cp Z
Given the viscosity and gas 200 0.01117 0.9768
deviation factor data reported
400 0.01176 0.9545
in Table 5.1.1 of a gas which
has a specific gravity of 0.76, 600 0.01236 0.9332
estimate the pseudo-pressure 800 0.01297 0.9131
function, m(P). 1000 0.01360 0.8946
1200 0.01423 0.8778
1400 0.01488 0.8631
SOLUTION 1600 0.01555 0.8507
1800 0.01622 0.8409
The numerical integration
2000 0.01690 0.8338
procedure will be illustrated
only for P = 200 psia. 2200 0.01759 0.8295
2400 0.01828 0.8280
Table 5.1.2 shows the complete 2600 0.01898 0.8292
results. 2800 0.01969 0.8329
3000 0.02041 0.8389
 28
1. First calculate the group 2P/µgZ
2P 2 × 200
= = 36660.7 psi / cp
µ g Z 0.01117 × 0.9768

2. The pressure difference is constant and equal to 200 psi.

3. Calculate the mean value of 2P/µgZ between P = 0 and P = 200:

 2P  36660.7
  = = 18330.35 psi / cp
µ Z  2
 g  mean

4. Multiply the group 2P/µgZ with 200 psi:

*
 2P   2P 
  =   ∆P = 18330.35 × 200 = 3.666 × 10 6 psi 2 / cp
 µ Z  i  µ Z  mean

5. Calculate pseudo-pressure from:

*
 2P 
m( P ) i =   + m( P ) i −1 = 3.666 × 10 6 + 0 = 3.666 × 10 6 psi 2 / cp
 µ Z i
 29
Table 5.1.2 – Pseudo-pressure data for Example 5.1.1

P, psi µg, cp Z 2P/µµZ, psi/cp ∆P, psi m(P), psi2/cp

200 0.01117 0.9768 36660.7 200 3.67E+06
400 0.01176 0.9545 71270.0 200 1.45E+07
600 0.01236 0.9332 104037.1 200 3.20E+07
800 0.01297 0.9131 135102.0 200 5.59E+07
1000 0.01360 0.8946 164385.0 200 8.59E+07
1200 0.01423 0.8778 192136.9 200 1.22E+08
1400 0.01488 0.8631 218018.8 200 1.63E+08
1600 0.01555 0.8507 241904.1 200 2.09E+08
1800 0.01622 0.8409 263941.3 200 2.59E+08
2000 0.01690 0.8338 283864.7 200 3.14E+08
2200 0.01759 0.8295 301557.7 200 3.72E+08
2400 0.01828 0.8280 317128.1 200 4.34E+08
2600 0.01898 0.8292 330405.9 200 4.99E+08
2800 0.01969 0.8329 341467.6 200 5.66E+08
3000 0.02041 0.8389 350427.4 200 6.35E+08
 30
EXAMPLE 5.1.2 t, hr Pwf,, psia

Given the viscosity and Z-factor data in 0 5000

0.021 4407
Table 5.1.1, the pressure drawdown data 0.042 4121.6
presented in Table 5.1.3 and the 0.084 3829.2
information given below, determine: 0.146 3639.6
0.209 3556.4
0.418 3464.6
1. Reservoir permeability 0.626 3432.1
2. Total skin factor 0.835 3410.4
3. Forscheimer turbulence factor and 1.044 3394
1.46 3372.1
4. Total skin (due to turbulence and 2.088 3347.4
mechanical skin). 4.176 3306
6.246 3278.4

γg = 0.76 Pi = 5000 psia 8.352

10.44
3261.7
3247.8
T = 255 °F h = 12.4 ft 12.528 3236.6
14.616 3225.5
q = 5000 Mscf/D rw = 1/3 ft 16.704 3217.2
18.792 3211.6
Ø = 10% -5 -1 µgi = 0.0253 cp 20.88 3203.2
cti = 13.1x10 psi Swi = 65.5% 41.76 3161
62.64 3185.7
100.44 3104.5
 31
SOLUTION

Since the initial reservoir pressure is greater than 3500 psia, the
pressure test can be analyzed using the P-function.

However during the flow test the pressure dropped below 3500 psia,
so it maybe advisable to use the m(P)-function.

(A)Using the m(P) function

m(P) - vs - P
7.00E+08
Step 1 – Convert the y = -0.0081x3 + 95.488x2 - 1614.7x + 306828
flowing pressure Pwf data 6.00E+08

to m(Pwf). Figure 5.1.2 is a 5.00E+08

Cartesian plot of P versus 4.00E+08

m(P)
m(p) data in Table 5.1.2. 3.00E+08

2.00E+08

The curve fit function 1.00E+08

shown in this figure is used 0.00E+00

to calculate the m(Pwf) 0 500 1000 1500 2000 2500 3000 3500

shown in Table 5.1.4. P

 32
Step 2 – Figure 5.1.3 is a semilog 1.20E+09
plot of m(Pwf) versus time. 1.10E+09

1.00E+09

m(P), psi2/cp
This plot shows a late-time
straight line corresponding to the 9.00E+08

infinite-acting radial flow line. 8.00E+08

m=-5.16E+07
The slope of this line is mp = - 7.00E+08 m(P)1hr=7.8E+08
5.16E+07 psi2/cp. 6.00E+08
0.01 0.1 1 10 100 1000
Step 3 – Equation 5.1.16 is used time, hr
to calculate permeability:
1422 qscT 1422 × 5000 × (460 + 255)
k =− =− = 8 md
mph − 5.16 × 10 7 × 12.4
Step 4 – Obtain the skin factor from Eq. 5.1.17, where m(P)1hr =
7.8x108 psi2/cp and m(Pi) = 1.37x109 psi2/cp:

 m( P ) − m ( P )  k  
S ′ = 1.151 i 1hr
− log  + 3.23
 mP  φµ c r 2  
 g t w 
13.7 − 7.8  8  
= 1.151 − log −5 2 
+ 3.23 = 6
 0.516  0.10 × 0.0253 × 13.1 × 10 × 0.333  
 33
Step 5 – Forscheimer turbulence factor and non-Darcy coefficient:

4.851× 10 4 4.851× 10 4
β= = = 5.44 × 10 9
ft −1

φ 5.5 k 0.15.5 8

−15 γg kβ −15 0.76 × 8 × 5.44 × 10

9
D = 2.22 × 10 = 2.22 × 10 = 6.9 × 10− 4 D / Mscf
µ g rwh 0.0253 × 0.333 × 12.4

Step 6 – The additional “skin” due to turbulence is:

Dq sc = 6.9 × 10 −4 × 5000 = 3.5

Step 7 – The mechanical skin factor is:

S = S '− Dqsc = 6 − 3.5 = 2.5

 34
4600

B) Using the pressure function 4400

4200

Figure 5.1.4 is a semilog plot of 4000

Pwf, psia
Pwf versus test time. 3800

3600
m=150
3400
The slope of the straight line is 3200
P1hr=3400

m = 150 psi/logcycle, and the 3000

permeability is: 0.01 0.1 1 10 100 1000

time, hr

162.6qsc µ g Bg 162.6 × 5000 × 0.0253 × 0.000728

k=− =− = 8.0 md
mh − 150 × 12.4

Where Bgi is calculated as follows:

1. Knowing the specific gas gravity (0.76), use the following

correlations (for dry gas) to calculate Ppc [psia] and Tpc [oR]:
p = 756.8 − 131.0γ g − 3.6γ g2
pc
= 756.8 − 131.0 × 0.76 − 3.6 × 0.762 = 655.16
 35
T = 169.2 + 349.5γ g − 74.0γ g2
pc
= 169.2 + 349.5 × 0.76 − 74.0 × 0.762 = 392

2. Calculate the pseudoreduced

pressure (Ppr) and temperature (Tpr)

P 5000
p pr = = = 7.63
Ppc 655.16
T 460 + 255
Tpr = = = 1.82
Tpc 392

3. The factor Z is then determined from

the Standing-Katz chart (Fig. 5.1.5): Zi
= 1.01. The gas almost ideal, therefore
the P-function is applicable.

4. The gas formation volume factor at initial pressure is:

Z iT 1.01 × (460 + 255)

Bg = 0.00504 = 0.00504 = 0.000728 bbl / scf
Pi 5000
 36

5.2 – TYPE-CURVE
MATCHING
 37
5.2. - TYPE-CURVE MATCHING TECHNIQUE

The step-by-step presented in Chapter 2 for oil wells is also applicable

to gas wells.

The following example illustrates the procedure.

In tight formations it may take too long to reach the infinite acting
line.

Consequently very often the semilog straight line portion of the

pressure versus test time plot, which is crucial for calculating the
permeability and skin factor, is not observed.

In this case the type-curve matching method based on Figure 5.2.1 is

particularly applicable, because of the presence of a unique hump for
different values of CDe2s during the early time portion of the pressure
derivative curve.
 38
1.E+02
Type Curve for a Gas Well with Wellbore Storage and Skin
100
Maximun-points line

CD e ^2S
Unit-slope line

Start of
infinite-acting line
m(p) D & (t D/C D)m(p)'D

10
1.E+01

1
1.E+00 Intersection
point

Infinite-acting line

0.1
1.E-01
0.1
1.E-01 1.E+00
1 1.E+01
10 100
1.E+02 1000
1.E+03 1.E+04
10000
tD/CD

Figure 5.2.1 – Type Curves for Gas Reservoirs

 39
EXAMPLE
A 120-hour pressure buildup test
was run in gas well before
hydraulic fracturing.

The well is located in a very tight

formation. Fig. 5.2.2 is a
diagnostic log-log plot of the real-
gas pseudo-pressure data ∆m(p)
against shut-in time ∆t.

The well and reservoir

characteristics are:
h = 77 ft µ = 0.028
Ø=5% ct = 0.00014 psi-1
rw = 0.25 ft q = 306 MSCF/D
T = 386 oF = 846 oR

Calculate permeability, skin and

wellbore storage coefficient.
Fig. 5.2.2
 40
SOLUTION
The equations for oil reservoirs can also be used for gas reservoirs
(especially if P>3000 psia and the formation is not tight), by
converting gas flow rates from SCF/D to STB/D and calculating the
gas formation volume factor Bg from:

ZT
B g (bbl / SCF ) = 0.00504 (5.2.1)
P
Where T = formation temperature, oR
Z = Gas deviation factor at the average reservoir pressure

For gas reservoirs (especially in gas reservoirs where 2000<P<3000

psi and in tight gas formations), the permeability is preferably and
more accurately obtained from:
1424qT PDM
k= (5.2.2)
h ∆m( p) M

Where: m(p) = Real-gas pseudo-pressure, psi2/cp, q = MSCF/D, T =

oR and h = ft.
 41
Fig. 5.2.2 shows a good match of both
real-gas pseudopressure and
pressure derivative curves: ∆m(p)
and ∆t*∆m(p)’

The coordinates of the match point are:

PD = 7.2 × 10 −1 t D / C D = 1.4 C D e 2 s = 10 8
∆m( p ) = 10 8 psi 2 / cp ∆t = 1 hr

The permeability is obtained from Eq.

5.2.2:

1424 × 306 × 846  0.72 

k=   = 0.034 mD
77  108 

Using Equations 2.1.12, 2.1.13 and 2.1.16, the wellbore storage

coefficient and apparent skin factor are:
 42

0.0002637k  t 
CD =  
φµct rw2  t D / CD  M
3.6 ×10−6 × 0.034  1.0 
=   = 525
0.05 × 0.028 × 0.00014 × 0.252  1.4 

 φ c t hr w2 
C =   C D
 0 . 89359 
 0 . 05 × 0 . 00014 × 77 × 0 . 25 2  bbl
=   525 = 0 . 002
 0 . 89359  psi

S ' = 0.5 ln

(
 CD e2 S ' ) M
  108 
 = 0.5 ln  = 6

 CD   525 

From the results of this calculation, i.e. k = 0.034 md and s = 6,

this gas well is a good candidate for hydraulic fracturing.
 43

5.3 – TDS
TECHNIQUE
 44
5.3 – TDS TECHNIQUE

The classical assumptions normally used in conjunction with a single

gas well producing at a constant rate from a homogeneous, isotropic
and uniform porous media are applicable in this section.

The dimensionless wellbore pressures for a well with storage and

skin, m(PD), and its derivative, dm(PD)/dtD, are obtained from:

∞  −u 2t 
4 1− e D 
m D ( rD , t D ) = ∫ 
π 0  u 3U J
2 du

(5.3.1)
 

 −u 2tD
∞ 
d (m D ) 4 e  (5.3.2)
dt D
=
2 ∫ 
π 0  uU J
du

 
where

[
U J = uCD J 0 (u) − (1 − C D su 2 ) J 1 (u) ] [
2
( ) ]
+ uCDY0 (u ) − 1 − C D su 2 Y1 (u )
2
(5.3.3)
 45
The dimensionless pseudo-pressure, mD, dimensionless time, tD and
dimensionless wellbore storage coefficient are expressed as follows:

m D (rD , t D ) =
kh
[m( pi ) − m( p)] (5.3.4)
1422.7 q sc T

hφ ( µct ) i rw
2

m( p )' D = 2.32 ∆m( p )' (5.3.4a)

qT

 0.0002637k   0.8935 
t D =  t CD =  C
 φc hrw 
 φµi ct rw 
2 2
 t 

(5.3.5) (5.3.6)

Where:
qsc (Mscf/D) = Flow rate at standard conditions of pressure (14.7
psia) and temperature 520 0R (60 0F).
T (0R) = Reservoir temperature, and
µ (cp) = gas viscosity, calculated at initial reservoir pressure
(preferably), or at the average reservoir pressure
 46
Using the same approach as in Chapter 2,
several equations describing the behavior 10000
∆ PR
of the loglog plot of of the pseudo-

t*∆m(P)'
pressure and its derivative with respect to 1000
time.

∆m(P)
Since using either real time or equivalent 100 (t*∆P')R
time in both semilog analysis and type- tR
curve matching techniques gives almost 10
the same results, the following derivations 0.01 0.1 1 10 100
will be done with respect to the time
dimensionless test time, tD.

The log-log plot of the pseudo-pressure versus time has three

major features:
(a) Both the pressure and derivative show an early-time unit-slope
straight line corresponding to the wellbore storage effect,
(b) The pressure derivative has a hump corresponding to the
combined effects of skin and wellbore storage, and
(c) The pressure derivative shows a late-time horizontal straight
line portion,which corresponds to the infinite acting radial flow
regime (Fig. 5.3.1).
 47
1) - Wellbore Storage Coefficient

The early-time pseudo-pressure curve has a unit slope line. This line
corresponds to pure wellbore storage effect. The equation of the
straight line is:

tD
mD ( P ) = (5.3.7)
CD

Combining Equations 5.3.5 and 5.3.6 yields;

tD  kh  t
=  2.95 × 10 −4  (5.3.8)
CD  µi  C

Substituting Equations 5.3.4 and 5.3.8 into Equation 5.3.7 and solve
for C, leads to:

 0.42qscT  t 
C =    (5.3.9)
 µi  ∆m( P) 
 48
Similarly, the unit slope line of the derivative curve:

 t D   tD
 m' ( p ) D  = (5.3.8)
 CD   CD

Taking the derivative of Equation 5.3.4 with respect to the natural log
of tD/CD (or simply the derivative with respect to tD/CD), and applying
the chain rule method yields:

tD '  kh  ∂∆m( P) ∂t  kh 
mD ( P) = 
  =
 (t × ∆m' ( P))
 1422.52qscT  ∂t ∂(ln(t D CD ))  1422.52qscT 
CD 

(5.3.10)

Combining Equations 5.3.10 and 5.3.8, and solving for C yields:

 0.42q scT  t 
C =    (5.3.11)
 µi  ∆m( P)' 
 49
2) - Permeability

The infinite-acting radial flow portion of the pressure derivative is a

horizontal straight line.

In terms of the pseudo-pressure function the Equation of this line is

similar to Equation 2.2.15.

The pseudopressure derivative curve has also an early-time straight

line of unit slope.

The equation of this line is obtained by taking derivative of Eq. 3.24

with respect to natural log of tD/CD. Thus,

1   tD 
m D ( P ) r = ln 
2   C D

(
 + 0.80907 + ln C D e 2 S ') )
 (5.3.12)
r 

Where S’ is the total skin and D (D/Mscf) is

the non-Darcy flow coefficient: S ' = S + Dqsc (5.3.13)
 50
The derivative of Equation 5.3.12 with respect to the natural log of
tD/CD (or simply the derivative with respect to tD/CD) is:

 tD 
 m ' ( P ) D = 0.5 (5.3.13)
CD R

Combining Equation 5.3.10 and 5.3.13 and solving for permeability

gives:

711.26qscT
k= (5.3.14)
h(t × ∆m' ( P)) R

3) - Skin Factor

The skin factor is determined from a relationship between the

pseudo-pressure and its derivative during the infinite acting radial
flow.

This relation is derived by dividing Equation 5.3.12 with Equation

5.3.13:
 51

= ln(tDR ) + 0.80907+ 2S'

(mD (P))R
(5.3.15)
((tD / CD )m'D (P)) R
Combine Equations 5.3.4, 5.3.5, and 5.3.13 with Equation 5.3.15 and
solve for the skin factor yields:

 ( ∆m( P)) R  kt R  
S ' = S + Dqsc = 0.5 − ln  + 7.43
 (t × ∆m' ( P ) )R
2 
 φ ( µct ) i rw  
(5.3.16)
4) - Drainage Area

Just like the pressure derivative function (Chapter 2), the

pseudopressure derivative function also yields a late-time
straight line of unit slope in a closed system for long producing
times (Fig. 5.3.2).

This line, which corresponds to the pseudosteady state flow

regime, starts at a tDA value of approximately 0.2.
 52
The equation of this straight line is

t DA * m' ( p )wD = 2πt DA

1000

(5.3.17)

t*∆m(p)'
An expression to calculate the
drainage area can be derived by 100

∆m(p)
substituting the dimensionless
terms in Eq. 5.3.17: t*dm(p)R

PSS line
 2.355qT 
t * m' ( p )w
10

=  t
 φ (µct )i Ah 
0.001 0.01 0.1 1 10

 (5.3.18) time, hr

Solving Eq. 5.3.18 for A (ft2) gives:

  t pss
 2.355 qT 
A =   (t ×∆m ( P )' ) pss
 φ ( µct )i h 
(5.3.19)

Where tPSS and (tx∆m(P)’)PSS are the coordinates of any point taken
on the pseudo-steady state line.
 53
5) - Maximum Point of the “hump”
The coordinates of the maximum point of the wellbore storage hump
(tx and (t*∆m(p)’)x ) can be used to estimate the wellbore storage
coefficient if the infinite acting line is well defined:
 kh  tX
C =  
 4708µ   (t * ∆m' ( p ) ) X  (5.3.20)
 + 1
 (t * ∆m ' ( p ) ) R 
The coordinates can also be used to estimate permeability if the
infinite acting line has not been observed, e.g. short test:
 qT 
 722.26 
k=  h 
(5.3.21)
 qT  t
 0.151  X − (t * ∆m' ( p) ) X
 µ C

Equations 5.3.20 and 5.3.21 are extremely sensitive to the

coordinates of the maximum point of the hump.
They should only be used to obtain an estimate of C and k, in the
absence of the early-time unit slope line or the absence of the infinite
acting line.
 54
6) - Verification

The slope mP of the straight line (infinite acting line), obtained from
the semilog plot of m(p) versus time, and the value of (t*∆m(p)’)R are
related by the flowing equation:

m p = 2.303(t × ∆m( p )')R (5.3.22)

8.01E+08
1.00E+09
7.01E+08
mp= 5.15x107
6.01E+08

t*∆m(p)'
t*∆m(p)'

5.01E+08

4.01E+08
1.00E+08

∆m(p)
3.01E+08
∆m(p)

2.01E+08
t*∆m(p)R' = 2.24x107
1.01E+08
t*∆m(p)R' = 2.24x107
1.00E+07 1.00E+06

0.01 0.1 1 10 100 0.001 0.01 0.1 1 10 100

time, hr time, hr
 55
EXAMPLE 5.3.1 1.00E+09

Using data in Example 5.1.2 and

t*∆m(p)'
the TDS technique, recalculate 1.00E+08

permeability (k) and total skin (S’).

∆m(p)
t*∆m(p)R' = 2.24x107

SOLUTION 1.00E+07
0.01 0.1 1 10 100

time, hr

Table 5.3.1 shows the pseudo-

pressure and the derivative Figure 5.3.1
pseudo-pressure data, which is
calculated from: (5.3.23)

t × ∆m( P )' =   =  +
(
 ∂ (∆m( P ) )   ln (t i / ti −1 )∆m( P ) i +1 ln t i +1t i −1 / t i2 ∆m( P) i
−
)
ln (t i +1 / t i )∆m( P ) i −1 

 ∂ ln t ln (t
 i  i +1 i/ t ) ln (t i +1 / t i −1 ) ln (t i +1 / t i ) ln (t i / t i −1 ) ln (t i / t i −1 ) ln (t i +1 / t )
i −1 

Figure 5.3.1 is the loglog plot of ∆m(p) and (t*∆m(p)’) vs. test time.

The horizontal line, which corresponds to the infinite acting radial

flow regime, yields:
(t*∆m(p)’)R = 2.24x107 psia2/cp
 56
Table 5.3.1
t, hr m(p) ∆m(p) t*∆m(p)'
0 1.37E+09
0.021 1.15E+09 2.16E+08
0.042 1.05E+09 3.21E+08 1.55E+08
0.084 9.39E+08 4.31E+08 1.41E+08
0.146 8.69E+08 5.01E+08 1.02E+08
0.209 8.38E+08 5.32E+08 7.34E+07
0.418 8.04E+08 5.66E+08 3.67E+07
0.626 7.92E+08 5.78E+08 2.85E+07
0.835 7.84E+08 5.86E+08 2.73E+07
1.044 7.78E+08 5.92E+08 2.57E+07
1.46 7.70E+08 6.00E+08 2.46E+07
2.088 7.61E+08 6.09E+08 2.41E+07
4.176 7.46E+08 6.24E+08 2.38E+07
6.246 7.36E+08 6.34E+08 2.26E+07
8.352 7.30E+08 6.40E+08 2.18E+07
10.44 7.25E+08 6.45E+08 2.24E+07
12.528 7.21E+08 6.49E+08 2.43E+07
14.616 7.17E+08 6.53E+08 2.41E+07
16.704 7.14E+08 6.56E+08 1.97E+07
18.792 7.12E+08 6.58E+08 2.33E+07
20.88 7.09E+08 6.61E+08 2.79E+07
41.76 6.93E+08 6.77E+08 2.22E+07
62.64 6.84E+08 6.86E+08 2.30E+07
100.44 6.73E+08 6.97E+08
 57
Equation 5.3.14 gives

711.26qscT 711.26 × 5000 × (460 + 255)

k= = = 9.2 md
h(t × ∆m' ( P )) R 12.4 × 2.24 × 10 7

Selecting the time tR = 41.76 hr on the infinite-acting line gives:

∆m(p) R= 6.77E+08 psi2/cp.

The skin factor is then obtained from Eq. 5.3.16:

 (∆m( P )) R  kt R  
S ' = 0.5 − ln  + 7 .43 
 (t × ∆m ' ( P ) )R  φ ( µc ) r 2 
t i w  
 6.77 E + 08  9.2 × 41.76  
= 0 .5  − ln −5  + 7 .43 = 7.8
 2 . 24 E + 07  0. 10 × 0 .0253 × 13 . 1 × 10 × 0. 3332
 

Substituting the value of (t*∆m(p)’)R into Eq. 5.3.22 gives:

m p = 2.303(t × ∆m( p )')R = 2.303 × 2.24 E + 07 = 5.15 E + 07

 58
This value is practically equal to the 1.20E+09

slope obtained in Fig. 5.1.3. 1.10E+09

1.00E+09

m(P), psi2/cp
Figure 5.3.2 shows the semilog plot of 9.00E+08

∆m(p) and (t*∆m(p)’) versus test 8.00E+08

m=-5.16E+07
time. 7.00E+08 m(P)1hr=7.8E+08

6.00E+08

The slope of the straight line portion 0.01 0.1 1 10 100 1000

time, hr
of the ∆m(p) curve is equal to mP
obtained from Eq. 5.3.20.
Fig. 5.1.3.
8.01E+08

7.01E+08
mp= 5.15x107
6.01E+08
t*∆m(p)'

5.01E+08

4.01E+08
∆m(p)

3.01E+08

2.01E+08

1.01E+08
t*∆m(p)R' = 2.24x107
1.00E+06
0.001 0.01 0.1 1 10 100

time, hr

Figure 5.3.2
 59

5.4 -
HYDRAULICALLY
FRACTURED GAS
WELLS
 60
The system considered in this study consists of a gas well centered in
a closed system, intercepted by a hydraulic fracture, Fig. 5.4.1.

The hydraulic fracture penetrates the entire thickness, h, of the pay

zone, and has a half-length of xf , width of wf and permeability of kf..

Permeability, k, and the thickness, h, of the reservoir are assumed to

be constant and uniform.

w
Well

Fracture

Thickness

Impermeable
boundaries

xf

Fig. 5.4.1A Fig. 5.4.1B

 61
100
100

The constant rate production case p s e u d o s te a d y st a t e

is assumed and for simplicity 10 10

m(p)wD and tDA * m(p)'wD

m (p ) wD

wellbore storage effects are

x e /x f

1 1
8 4 1

neglected, these effects may mask r a d ia l f lo w

the early time flow regimes in a 0 .1 0

fractured system. 0 .0 1 0
l in e a r f lo w
u n if o r m f l u x

0 0 0 0 1 10
0 .0 0 0 1 0 .0 0 1 0 .0 1 0 .1 1 10
tD A

1 0 0100

Using the same approach as in

p se u d o stea d y sta te

10 10

m(p)wD and tDA * m(p)'wD

Chapter 3, figures 5.4.3 – 5.4.5 x e /x f
m (p )wD

were generated for gas flow.

16 4 2
1
1

r a d ia l fl o w

b i- r a d ia l flo w

0 .1 0

It can be noted from these figures,

lin e a r flo w

in fi n i t e c o n d u c t iv it y
0 .0 1 0
00 0 . 00 0 1 0 . 00 1 0 0. 1 11
.0 0 0 1 1 01 0

that the curves, just like for oil tD A

reservoirs, have several unique 1 0 0 10 0

characteristics.
p se u d o ste a d y sta te

10 10

tDA * m(p)'wD
These characteristics can be used to
4:1R
xe/xf squa re
1 1

16
r a d ia l fl o w

interpret pressure transient tests in 0 .1 0

4

fractured gas wells without using

l i n e a r f lo w

u n i f o r m fl u x

type-curve matching procedure. 0 .0 1 0

0
0 .0 0 0 1
0
0 .0 0 1 0 .0 1
0

tDA
0 .1
0 1
1 10
10
 62
5.4.1 - UNIFORM FLUX & INFINITE CONDUCTIVITY FRACTURES

Figure 5.4.2 shows the distinct

flow regimes theoretically
expected in the fracture and
formation around a hydraulically
fractured well.

Initially fracture-linear flow (Fig.

5.4.2a) occurs into the fracture,
which is characterized by a slope
of 0.5 in a log-log plot.

Figure 5.4.2

During this fluid period, most of the fluid entering the wellbore comes
from fluid expansion in the fracture, the occurrence of this flow
period is too short and normally is never seen.
 63
Bilinear flow14 (Fig. 5.4.2b), which
is characterized by a slope of 0.25 on
the log-log plot, takes place in finite-
conductivity fractures as fluid in the
surrounding formation flows linearly
into the fracture, most of the fluid
entering the wellbore during this flow
period comes from the formation.

Fractures are considered to have finite-conductivity when

dimensionless fracture conductivity30, CDf<100.

Fracture conductivity has been described as having values

varying typically from 1 to 500.

A low value of dimensionless conductivity indicates low fracture

permeability or long fracture lengths, or possibly both.

On the other hand, a high value of conductivity implies high

fracture permeability, small fracture length, or both.
 64
Formation-Linear flow (Fig.
5.4.2c) occurs only in high
conductivity fractures (CDf>100),
this flow is also identified by a slope
of 0.5, once the linear flow in the
formation -or the bilinear flow-
vanishes, the pseudo-radial flow
takes place (Fig. 5.4.2d).

100 100

For the uniform-flux fracture pseudosteady state

case (Figure 5.4.3) it was 10 10

m(p)wD and tDA * m(p)'wD

shown that in gas systems as xe/xf
m(p) wD

well as for oil systems

1 1
1

(Tiab23), pressure derivative

8 4

radial flow

plots for various xe/ xf ratios

may reveal three dominant 0.1 0

flow regimes: linear flow

uniform flux

0.01 0
0 0 0 0 1 10
0.0001 0.001 0.01 0.1 1 10
t DA
 65
100 100

pseudosteady state

(a) The first flow regime is called

“linear” and is identified by a straight
10
10

m(p)wD and tDA * m(p)'wD

m(p) wD
xe/xf

line of slope 0.5; the half-fracture

length is calculated from this line.
1 1
8 4 1

radial flow

This flow regime is very short and

0.1 0

may be masked by wellbore storage linear flow

uniform flux

effects.
0.01 0
0 0 0 0 1 10
0.0001 0.001 0.01 0.1 1 10
tD A

During this flow regime, most of the fluid entering the wellbore
comes from fluid expansion in the fracture, and the flow pattern is
essentially linear.

(b) The infinite-acting radial-flow regime, which can be identified by

a horizontal straight line, is also dominant for xe/ xf > 8 in gas
systems. This flow regime is used to calculate permeability and skin.

(c) The third straight line, which corresponds to the pseudosteady

flow regime, has a unit-slope line. This line is used to calculate the
drainage area and shape factor.
 66
For the infinite-conductivity fracture case, pseudopressure derivative
plots reveal a fourth dominant flow regime, referred as bi-radial (or
elliptical) flow.

This flow regime, which can be identified by a straight line of slope

0.36, also can be used to calculate the half- fracture length and
permeability.
100
100

pseudosteady state

10 10
m(p)wD and tDA * m(p)'wD

m (p) wD
xe/xf

16 4 2
1
1

radial flow

bi-radial flow
0.1 0

linear flow

infinite conductivity
0.01 0
0 0 0 0 1
0.0001 0.001 0.01 0.1 1 1010
tDA
 67
1. LINEAR FLOW REGIME

If the unit-slope line corresponding to wellbore storage is observed,

then the equations developed in the previous section can be applied
to calculate the wellbore storage coefficient.

For short producing times, fluid flow into fracture is linear. The
duration of this flow regime is a function of the penetration ratio
xe/xf.

The equation corresponding to this early-time straight line is:

 xe 
t DA * m' ( p )wD = 1.772  t DA (5.4.1)
x 
 f 
Taking the logarithm of both sides of the equation yields:

 xe π 
log(t DA * m' ( p )wD ) = 0.5 log t DA + log  (5.4.2)
 x 
 f 
A straight line of slope 0.5 identifies the linear flow regime.
 68
Substituting for the dimensionless terms in Eq. 5.4.1 a solving for the
derivative of the pseudo-pressure we have:

t * ∆m' ( p) w = 0.5mL t (5.4.3)

1.00E+10

where 1.00E+09

t*∆m(p)'
t*∆m(p)R' = 2.54x108
qT 1
mL = 40.915
h φk (µCt )i x f 2 (5.4.4) 1.00E+08
t*∆m(p)'L1hr= 4.6x107

∆m(p)
slope = 1, PSS flow

1.00E+07

Taking the logarithm of both

sides of Eq. 5.4.3 gives slope = 0.5, linear flow

log(t * ∆m' ( p )w ) = 0.5 log t + log(0.5mL )

1.00E+06
1.00E-03 1.00E-02 1.00E-01 1.00E+00 1.00E+01 1.00E+02 1.00E+03 1.00E+04

time, hr

(5.4.5)

It can be noticed from this expression that a plot of measured

t*m(P)’ versus time on a log-log graph will yield a straight line of
slope 0.5 when the linear flow regime is dominant.
 69
Combining Eqs. 5.4.3 and 5.4.4 and
solving for ½-length xf :
1.00E+10

1.00E+09

20.457 qT 1
xf =

t*∆m(p)'
h(t * ∆m' ( p )w )L1 φk (µct )i
t*∆m(p)R' = 2.54x108

1.00E+08
t*∆m(p)'L1hr= 4.6x107

∆m(p)
slope = 1, PSS flow

Where: t*m(P)’ )L1 = value of 1.00E+07

t*m(P)’ at time t=1 hr on the slope = 0.5, linear flow

linear-flow straight line

1.00E+06
1.00E-03 1.00E-02 1.00E-01 1.00E+00 1.00E+01 1.00E+02 1.00E+03 1.00E+04

time, hr
(extrapolated, if necessary);

2. INFINITE-ACTING RADIAL FLOW REGIME

The infinite-acting radial-flow line 100 100

(horizontal) occurs after the linear pseudosteady state

flow, Fig. 5.4.1. 10 10

m(p)wD and tDA * m(p)'wD

m(p) wD
xe/xf

Notice that, in a closed reservoir, this 1 1

8 4 1

radial flow

flow regime is dominant only if the

penetration ratio xe/xf >8. 0.1 0

linear flow
uniform flux

The equation corresponding to this 0.01 0

0
0.0001 0
0.001
0
0.01
tD A
0
0.1
1
1 10
10

second-straight line is described by:

 70
t DA * m' ( p )wD = 0.5 (5.4.7)
10000

t*∆m(p)'
Substituting for the dimensionless 1000

terms and solving for k yields:

infinite-acting

∆m(p)
slope=1, PSS flow
711.26qT 100

k=
slope=0.25, bi-linear

h(t * ∆m' ( p )w )R
(5.4.8)
10
0.1 1 10 100 1000

where the subscript R stands for time, hr

radial flow line.

The total apparent skin factor (S’) can be calculated from:

 ∆m( p )  kt  
S ' = 0.5 R
− ln R  + 7.43 (5.4.9)
 t * ∆m' ( p )R
 φ (µc ) r 
2
 t i w  

where tR is any convenient time during the infinite acting radial flow
line and ∆m(P)R is the value of ∆m(P)R corresponding to tR.
 71
3. PSEUDO-STEADY FLOW REGIME 100 100

pseudosteady state

The pseudopressure derivative

function yields a straight line of unit
10
10

m(p)wD and tDA * m(p)'wD

m(p) wD
xe/xf

slope in closed system for long

producing times.
1 1
8 4 1

radial flow

This line, which corresponds to the 0.1 0

PSS flow regime, starts at a tDA value linear flow

uniform flux

of approximately 0.2. 0.01 0

0 0 0 0 1 10
0.0001 0.001 0.01 0.1 1 10
tDA

The equation of this straight line is: t DA * m' ( p )wD = 2πt DA (5.4.10)

An expression to calculate A can be 10000

derived by substituting the

dimensionless terms in Eq. 5.4.10.

t*∆m(p)'
1000

 2.355qT  t PSS
A =   (5.4.11)
 φ (µCt )i h  (t * m' ( p )w )PSS
infinite-acting

∆m(p)
100 slope=1, PSS flow
slope=0.25, bi-linear

Where and tPSS and (t*∆m(P)’)PSS 10

are the coordinates of any point on 0.1 1 10

time, hr
100 1000

the pseudosteady state line.

 72
SHAPE FACTOR
The equation corresponding to the pseudosteady-state flow regime is:

m( p )wD = 2πt DA + ln (x / x ) + ln
2.2458 10000

e f (5.4.14)
CA

t*∆m(p)'
1000

Taking the ratio of Eqs. 5.4.14 and 5.4.10:

infinite-acting

∆m(p)
m( p )wD
slope=1, PSS flow

 x  2.2458 
100
slope=0.25, bi-linear
1
= 1+ ln  e  
t DA × m' ( p )wD 2πt DA   x f 
 C A  10
0.1 1 10 100 1000

time, hr
(5.4.15)
Substituting for the dimensionless terms and solving for CA:

 xe  
 exp1 −
(∆m( p )w ) pss  0.000527kt pss 

C A = 2.2458 
x
 f



 (t * ∆m ' ( p ) ) 
w pss  φ ( µc )
t i A 
(5.4.16)

Where (∆m(P))PSS and (t*∆m(P)’)PSS are obtained at any time line,tPSS

, during the pseudosteady state flow.
 73
4. BI-RADIAL (OR ELLIPTICAL) FLOW REGIME

A log-log plot of dimensionless 100100

pseudopressure and pseudosteady state

pseudopressure derivative versus 10 10

tDA for an infinite-conductivity

m(p)wD and tDA * m(p)'wD

m(p)wD
xe/xf

vertical fracture inside a square

system is shown in Fig. 5.4.4. 1 1
16 4 2

radial flow

In this case the bi-radial (or 0.1 0

bi-radial flow

elliptical) flow regime requires linear flow

special attention, it is why only the infinite conductivity

0.01

characteristics of the bi-radial flow

0
0 0 0 0 11
0.0001 0.001 0.01 0.1 1010
t DA

regime will be discussed.

5.4.4 - Infinite conductivity fracture
inside a square gas system.

The characteristics and interpretation of the other three flow regimes

(linear, radial and pseudo steady state) are exactly the same as
discussed for the uniform-flux fracture.
 74
In Fig. 5.4.4 the existence of four 100100

straight lines can be noted: pseudosteady state

10 10

m(p)wD and tDA * m(p)'wD

(a) the linear-flow line of slope 0.5, xe/xf
m(p)wD

(b) bi-radial flow line of slope 0.36, 16 4 2

1

(c) infinite-acting radial flow line

1

radial flow

(horizontal line), and

bi-radial flow

(d) pseudosteady state flow line of 0.1 0

slope unity. linear flow

infinite conductivity
0.01 0

For xe/ xf > 8, the linear flow regime

0 0 0 0 11
0.0001 0.001 0.01 0.1 1010
tDA

is almost non-existent, and the bi-

radial flow line is observed first.
For xe/ xf < 8, it is the radial flow line which disappears.

1 – Half-Fracture Length from Bi-Radial

0.72
x 
t DA * m' ( p )wD = 0.769 e  0.36
t DA (5.4.17)
x 
 f 
 75
Taking the logarithm of both sides of the equation yields:
  xe 
0.72


log(t DA * m' ( p )wD ) = 0.36 log t DA + log 0.769  
 x   (5.4.18)
  f  
Thus the slope = 0.36 is unique to the bi-radial flow regime.

In real parameters, Eq. 5.4.17 becomes:

0.72
x 
t * ∆m' ( p )w = 0.769GBR  e  t 0.36 (5.4.19)
x 
 f 
0.36
73.192qT  k  (5.4.20)
GBR =  
 φ (µct )i A 
kh  

Taking the logarithm of Eq. 5.4.19 gives:

  xe 
0.72


log(t * ∆m' ( p )w ) = 0.36 log t + log 0.769GBR   
 x   (5.4.21)
  f  
 76
A straight line of slope 0.36 identifies the bi-radial flow regime.

Notice that the bi-radial flow regime cannot be identified from the
∆m(P) curve.

If the linear flow regime is not observed, the half-fracture length xf,
can be calculated from the bi-radial straight line:
1.388
 GBR 
x f = 0.694 xe   (5.4.22)
 (t × ∆m ' ( p )w )BR1 

Where
GBR is calculated from Eq. 5.4.20, and k from the infinite-acting
radial flow line (Eq. 5.4.8).

(t*∆m(P)’)BR1 is the value of t*∆m(P)’ at time t = 1 hour on the

straight line of slope = 36 (extrapolated if necessary)
 77
2 – Permeability from Bi-Radial

The coordinates of the intersection point of the bi-radial flow and the
linear flow lines can be obtained by combining Eqs. 5.4.1 and 5.4.22:
2
 xf 
t DALBRi = 0.00257 
(5.4.23)
 xe 
Substituting for dimensionless time and solving for k yields:

39φ (µct )i x f
2

k= (5.4.24)
t LBRi

Another expression to calculate k can be derived by combining Eqs.

5.4.6 and 5.4.24:
 127.75qT  1
k =   (5.4.25)
 h (t × ∆ m ' ( p ) )  (t )0.5
w L1  LBRi

Thus, when the infinite acting radial flow regime is too short-lived or
essentially non-existent (xe/ xf < 8), we can still determine k, as long
as the linear and bi-radial flow regimes are well defined.
 78
5. BOUNDED GAS SYSTEMS
100 100

For a rectangular system in gas pseudosteady state

wells, the transition between the 10 10

infinite-acting radial flow and the

tDA * m(p)'wD
pseudosteady-state flow regimes, xe/xf
4:1R
square

for both types of fracture, is much

1
1

16
radial flow

longer than it is for the square.

4
0
0.1
linear flow

For a 4:1 rectangle, for instance,

this transition period yields a 0.01 0
0 0 0 0
uniform flux

1 10
0.0001 10

straight line of slope 0.5, as shown

0.001 0.01 0.1 1

t DA

in Fig. 5.4.5.

This straight line corresponds to the effect of the two closest parallel
boundaries

t DA * m' ( p )wD = 3.545 t DA (5.4.26)

Substituting for the dimensionless terms we obtain:

 79

t * ∆m' ( p ) = mCBG t (5.4.27)

Taking the logarithm of both sides of Eq. 5.4.27 yields an equation of

a straight line of slope 0.5 (reflecting the effect of the two closest
parallel boundaries or channel flow):

Log (t × ∆m' ( p )) = 0.5 log(t ) + log(mCB ) (5.4.27a)

where the subscript CGB stands for closest parallel boundaries in gas
systems, and mCBG is given by:
0.5
qT  1 
mCBG = 81.85   (5.4.28)
h  kφ (µct )A 
At time t = 1 hour on the straight line (extrapolated if necessary) of
slope 0.5 (Channel flow), Eq. 4.4.27a yields:
0.5
 
(t × ∆m' ( p ))CB1 = 81.85 qT  1
 (5.4.28a)
h  kφ (µct )A 
 80
Eq. 5.4.28a can be used to estimate A if the PSS steady flow regime is
not observed, or k if the infinite-acting radial flow line is not
observes, such as when Xe/Xf<8 (very long fracture).

Solving Eq. 5.4.28a for k gives:

2
 1  qT 

k = 6700   (5.4.29)
 φ (µct )i A  h(t * ∆m' ( p ))CB1 


Solving Eq. 5.4.28a for A (ft2) gives:

2
 1  qT 

A = 6700  
 φ (µct )i k  h(t * ∆m' ( p ))CB1 
 (5.4.29a)
 81
5.4.2 – FINITE CONDUCTIVITY FRACTURE

The equations presented above for calculating C, k, S’ and A for the

uniform flux and infinity conductivity fracture cases, are also
applicable for the finite conductivity fracture case.

The only thing unique about the finite conductivity fracture is the
presence of another flow regime, called Bi-linear.

Fig. 5.4.6 shows a typical curve of curves for a finite-conductivity

vertically fractured gas well.

10000
t*∆m(p)'

1000

infinite-acting
∆m(p)

100 slope=1, PSS flow

slope=0.25, bi-linear

10
0.1 1 10 100 1000

time, hr
 82
During the bilinear flow regime, the dimensionless well pseudo-
pressure derivative behavior is given by:

 0.6127 
t Dxf * m' ( p ) D =  t 0.25
 C fD  Dxf (5.4.30)
 

where the dimensionless fracture conductivity, CfD, is defined by:

wf k f
C fD = (5.4.31)
xf k

The dimensionless time, tDxf, and dimensionless well pseudo-

pressure derivative, ∆m(p)D’, are given by:

0.0002637kt hφ ( µct )i rw2

tD = m' ( p) D = 2.667 ∆m( p)
φµct x 2f qscT
(5.4.32) (5.4.33)
 83
Combining Eqs. 5.4.30, 5.4.31, 5.4.32 and 5.4.33 and solving for the
well pseudo-pressure derivative, ∆m(p)’, yields:

t * ∆m' ( p ) = 0.25mBLt 0.25 (5.4.34)

where the subscript BL stands for bilinear and mBL is defined by:

444.3  qscT 
mBL = (5.4.35)
(φµct k )0.25  h k f w f 

Taking logarithm of both sides of Eq. 5.4.34 yields,

log (t * ∆m( p)' ) = 0.25 log t + log(0.25mBL ) (5.4.36)

This expression indicates that a plot of versus time on a log-log

graph will yield a straight-line portion of slope 0.25, if the fracture
has a finite conductivity.
 84
at time t=1 hr on the bilinear flow straight line, extrapolated if
necessary, then Eq. 5.4.36 becomes:

(t * ∆m' ( p)) BL1 = 0.25mBL (5.4.37) 10000

t*∆m(p)'
The conductivity of the fracture is 1000

obtained from the pseudo-pressure

derivative line by combining Eq. infinite-acting

∆m(p)
100 slope=1, PSS flow
5.4.35 and 5.4.37: slope=0.25, bi-linear

2
12272.76  qscT 
k f wf =  
10

φ ( µct )i k h (t * ∆m ' ( p )) (5.4.38) 0.1 1 10 100 1000

 BL1  time, hr

If the linear flow and radial flow lines are present, but the
bilinear flow line of slope 0.25 is not well defined or not observed
due to wellbore phenomena, then the fracture conductivity, wfkf
can be estimated from23:
3.31739k
k f wf =
e S ' 1.92173
− (5.4.39)
rw xf
 85
For a low-conductivity fracture, CfD≤1, the straight line corresponding
to the linear flow regime will probably not be observed.

After the bilinear flow line, the pressure derivative curve generally
enters a transition flow.

If the test is run long enough, a horizontal line on the pressure

derivative curve is observed corresponding to the infinite-acting
radial flow regime.

In this case the half-fracture length can be estimated from:

1.92173
xf = S' (5.4.40)
e 3.31739k
−
rw wk f
 86

5.5 – NUMERICAL
EXAMPLES
 87
EXAMPLE 5.4.1

A constant-rate drawdown test was run in a gas well following a

fracture treatment. The data in Table 5.4.1 shows the flowing well
pressure, pseudopressure and time data; other known data are given
below:

q sc = 3000 Mscf / D Bg = 0.7085 bbl / Mscf µ g = 0.01961 cp

h = 60 ft rw = 0.25 ft φ = 0.10
ct = 0.0002084 psi −1 T = 250 o F Pi = 5000 psia Zi ≈ 1

Determine as many parameters as possible from this pressure test.

 88
Table 5.4.1 – Pressure Drawdown and
Pseudo-pressure data for Example 5.4.1

t, hr Pwf, psia m(p) t, hr Pwf, psia m(p)

0 5000 1.9029E+09 35.473 4142.8 1.4758E+09
0.00141 4993.1 1.8994E+09 56.275 3977.9 1.3933E+09
0.002813 4990.2 1.8979E+09 89.236 3792.9 1.3007E+09
0.005613 4986.2 1.8960E+09 141.91 3588.9 1.1987E+09
0.011199 4980.5 1.8931E+09 225.78 3369.2 1.0893E+09
0.028131 4969.1 1.8875E+09 360.02 3135.3 9.7391E+08
0.070662 4951 1.8786E+09 455.26 3016.8 9.1601E+08
0.1775 4922.4 1.8644E+09 576.26 2892.6 8.5592E+08
0.44586 4877 1.8419E+09 730.43 2766.7 7.9573E+08
1.12 4806.2 1.8069E+09 927.6 2637.6 7.3487E+08
1.7751 4758.3 1.7831E+09 1180.7 2502.5 6.7233E+08
3.5421 4668.5 1.7385E+09 1507.4 2355.7 6.0597E+08
5.6144 4594.6 1.7017E+09 1932 2188.4 5.3256E+08
8.8995 4507.3 1.6581E+09 2488.9 1984.4 4.4693E+08
11.205 4458.4 1.6337E+09 3227.3 1720.3 3.4299E+08
17.764 4347.7 1.5784E+09 4221.8 1347.9 2.1698E+08
22.368 4284.9 1.5470E+09 5588.5 683.8 5.7312E+07
 89

SOLUTION

Figure 5.4.1 is a log-log plot of 1.00E+10

∆m(p) and (t*∆m(p)’ versus time.

The pseudopressure derivative 1.00E+09

curve reveals the existence of

t*∆m(p)'
three straight lines. t*∆m(p)R' = 2.54x108

The first straight line of slope 0.5 1.00E+08

t*∆m(p)'L1hr= 4.6x107
corresponds to the linear flow

∆m(p)
slope = 1, PSS flow

regime
1.00E+07

The second straight line is

horizontal and corresponds to the slope = 0.5, linear flow

infinite acting radial flow regime. 1.00E+06

1.00E-03 1.00E-02 1.00E-01 1.00E+00 1.00E+01 1.00E+02 1.00E+03 1.00E+04

The presence of the third straight time, hr

line, with a unit slope, corresponds
to the pseudo-steady state line, Figure 5.4.1
indicating a closed system.
 90
1 – Permeability

The infinite acting line yields: (t*∆m(p)’)R=2.54x108 and therefore

the permeability is (Eq. 5.4.8):

711.26qscT 711.26 × 5000 × ( 460 + 250)

k= = = 0.097 md
h(t × ∆m' ( P )) R 60 × 2.54 × 10 8

2 – Skin Factor

The skin factor is calculated from Eq. 5.4.9, where at tR = 730.43 hr

and ∆m(p)R = 7.9572E+08:

 ( ∆m( P )) R  kt R  
S ' = 0.5 − ln  + 7 . 43
 (t × ∆m ' ( P ) )R  φ ( µc ) r 2 
t i w  
 7.9572 E + 08  0.097 × 730.43  
= 0. 5  − ln −4 2 
+ 7 . 43 = −5.0
 2.54 E + 08  0.10 × 0.01961 × 2.304 × 10 × 0.25  
 91
3 – Drainage Area

Selecting any point on the late-time pseudosteady state flow line

of slope, tPSS = 4221.8 hr and (t*∆m(p)’)PSS = 5.18x108, the
drainage area is (Eq. 5.4.11):

  t pss
 2.355 qT 
A =   (t ×∆m ( P )' ) pss
 φ ( µct )i h 
 2.355×3000×( 460 + 250 )  4221.8 2
=  
 = 1 . 67 E + 06 ft
 0.10×0.01961×2.304×10 − 4 ×60  5.18×10
8

1.67 E + 06
A = = 38.3 acres
43560

4 – Half-Fracture Length

The half fracture length is calculated from Eq. 5.4.6, where at time
t=1 hr on the linear-flow straight line
(t * ∆m' ( p )w )L1 = 4.6 × 107
 92
0.5
20.457 qT  1 
xf =  
h(t * ∆m' ( p )w )L1  φk (µct )i 


20.457 × 3000 × (460 + 250) 1

= −4
= 74 ft
60 × 4.6 × 10 7
0.10 × 0.097 × 0.01961× 2.304 × 10

5 – Fracture Conductivity

Since the bi-linear flow regime is not observed, the fracture

conductivity is estimated from Eq.5.4.39:

3.31739k 3.31739 × 0.097

wf k f = s
= −4.98
= 131.6 md − ft
e 1.92173 e 1.92173
− −
rw xf 0.25 74

The dimensionless conductivity fracture is:

wf k f 131.6
C fD = = = 18.4
xf k 74 × 0.097
 93
According to Cinco-Ley this
fracture has a low conductivity 1.80E+09

since CfD<100. 1.60E+09

1.40E+09

Also, in such a low permeability

t*∆m(p)'
1.20E+09

formation (k = 0.097 md), the 1.00E+09

mp= 5.85x108

target of hydraulic fracturing is to 8.01E+08

achieve at least a ratio of wfkf

∆m(p)
6.01E+08
/kh>1000.
4.01E+08

t*∆m(p)R' = 2.54x108
In this case, the ratio is only:
2.01E+08

1.00E+06
1.00E-03 1.00E-02 1.00E-01 1.00E+00 1.00E+01 1.00E+02 1.00E+03 1.00E+04

wf k f 131.6 time, hr
= = 22.6
kh 0.097 × 60
Fig. 5.4.2
6 – Verification
Substituting the value of (t*∆m(p)’)R into Eq. 5.3.22 gives:

m p = 2.303(t × ∆m( p )')R = 2.303 × 2.54 E + 08 = 5.85 E + 08

Fig. 5.4.2 is a semilog plot of ∆m(p) and (t*∆m(p)’) versus test time.
The slope of the straight line portion of the ∆m(p) curve is equal to
mP obtained from Eq. 5.3.22.
 94
t Pws m(Pws)
EXAMPLE 5.4.2 0 4664 1248126794
0.048 4673.8 1.2517E+09
A pressure buildup test was run 0.144 4677.1 1.2528E+09
following a fracture treatment in a 0.336 4680.7 1.2541E+09
gas well. Table 5.4.2 shows the 0.72 4685.1 1.2557E+09
pressure, pseudo-pressure and time 1.488 4691.1 1.2579E+09
data. Other data are given below: 3.024 4698.9 1.2607E+09
6.096 4709.8 1.2646E+09
12.24 4725.1 1.2701E+09
A pre-frac buildup test yielded 23.129 4742.6 1.2763E+09
k = 0.10 md. 39.257 4761.6 1.2831E+09
62.537 4781 1.2900E+09
Determine fracture length and 86.537 4796 1.2954E+09
fracture conductivity. 110.54 4808.7 1.2999E+09
134.54 4818.6 1.3034E+09
158.54 4827.9 1.3067E+09
168 4831.1 1.3078E+09

q sc = 1000 Mscf / D Bg = 0.6992 bbl / Mscf µ g = 0.02442 cp

h = 100 ft rw = 0.25 ft φ = 0.20
ct = 0.0001057 psi −1 T = 250 o F Pi = 5000 psia Zi ≈ 1
 95

SOLUTION

Figure 5.4.3 is a log-log plot of 1.00E+08

∆m(p) and (t*∆m(p)’ versus time.

slope = 0.25
The pseudopressure derivative slope = 0.5

t*∆m(p)'
curve reveals the existence of two
straight lines. t*∆m(p)BL1' = 5.2x106
1.00E+07

The first straight line of slope 0.5

∆m(p)
corresponds to the linear flow tLBLi = 11
regime.
t*∆m(p)L1' = 2.9x106
The second straight line has a slope 1.00E+06

of 0.25 and corresponds to the 0.1 1 10 100 1000

test time, hr
bilinerar flow regime.

The test is short since the infinite Figure 5.4.3

acting radial flow line is missing.
Since the infinite-acting radial flow is not available the skin factor can only be estimated from Eq. 3.6.51:
96
1 – Half-Fracture Length
The half fracture length is calculated from Eq. 5.4.6, where at time
t=1 hr on the linear-flow straight line: 7
(t * ∆m' ( p )w )L1 = 5.2 × 10
0. 5
20.457 qT  1 
xf =  
h(t * ∆m' ( p )w )L1  φk (µct )i 


20.457 × 1000 × (460 + 250) 1

= −4
= 220 ft
100 × 2.9 × 10 7
0.20 × 0.1 × 0.02442 × 1.057 × 10

2 – Fracture Conductivity
The fracture conductivity is obtained from Eq. 5.4.38, where at time
t=1 hr on the bilinear flow straight line: 7
(t * ∆m' ( p )w )L1 = 2.9 × 10
2
12272.76  qscT 
k f wf =  
φ ( µct ) i k  h(t * ∆m' ( p )) BL1 
12272.76 1000 × 710
= = 101 md − ft
0.20 × 0.10 × 0.02442 × 1.057 × 10 −4 100 × 2.9 × 10 7
Since the infinite-acting radial flow is not available the skin factor can only be estimated from Eq. 3.6.51:
97
3 – Skin Factor

The skin factor can only be estimated since the infinite acting line is
not available. Eq. 3.6.51 gives:

  1.92173 3.31739k 
s = ln rw  − 
 w f k f 
  x f
  1.92173 3.31739 × 0.10 
= ln 0.25 ×  −  = −6.6
  220 101 
End of PART II