SPE 104056

Fracture Porosity of Naturally Fractured Reservoirs

D. Tiab, D.P. Restrepo, and A. Igbokoyi, SPE, U. of Oklahoma
Copyright 2006, Society of Petroleum Engineers
This paper was prepared for presentation at the First International Oil Conference and Exhibition
in Mexico held in Cancun, Mexico, 31 August2 September 2006.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to an abstract of not more than
300 words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.

Abstract
The storage capacity ratio, , measures the flow capacitance
of the secondary porosity and the interporosity flow
parameter, , is related to the heterogeneity scale of the
system. Currently, both parameters and are obtained from
well test data by using the conventional semilog analysis,
type-curve matching or the TDS Technique. Warren and Root
showed how the parameter can be obtained from semilog
plots. However, no accurate equation is proposed in the
literature for calculating fracture porosity.
This paper presents an equation for the estimation of the
parameter using semilog plots. A new equation for calculating
the storage capacity ratio and fracture porosity from the
pressure derivative is presented. The equations are applicable
to both pressure buildup and pressure drawdown tests. The
interpretation of these pressure tests follows closely the
classification of naturally fractured reservoirs into four types,
as suggested by Nelson1.
The paper also discusses new procedures for interpreting
pressure transient tests for three common cases: (a) the
pressure test is too short to observe the early-time radial flow
straight line and only the first straight line is observed, (b) the
pressure test is long enough to observe the late-time radial
flow straight line, but the first straight line is not observed due
to inner boundary effects, such as wellbore storage and
formation damage, and (c) Neither straight line is observed for
the same reasons, but the trough on the pressure derivative is
well defined. Analytical equations are derived in all three
cases for calculating permeability, skin, storage capacity ratio
and interporosity flow coefficient, without using type curve
matching.
In naturally fractured reservoirs, the matrix pore volume,
therefore the matrix porosity is reduced as a result of large
reservoir pressure drop due to oil production. This large
pressure drop causes the fracture pore volume, therefore
fracture porosity, to increase. This behavior is observed
particularly in reservoir where matrix porosity is much greater

than fracture porosity. Fractures in reservoirs are more

vertically than horizontally oriented, and the stress axis on the
formation is also essentially vertical. Under these conditions,
when the reservoir pressure drops, the fractures do not suffer
from the stress caused by the drop. Using these principles, a
new method is introduced for calculating fracture porosity
from the storage capacity ratio, without assuming the total
matrix compressibility is equal to the total fracture
compressibility.
Several numerical examples are presented for illustration
purposes.
Introduction
Nelson1 identifies four types of naturally fractured reservoirs;
based on the extent the fractures have altered the reservoir
matrix porosity and permeability: In Type 1 reservoirs,
fractures provide the essential reservoir storage capacity and
permeability. Typical Type-1 naturally fractured reservoirs are
the Amal field in Libya, Edison field California, and preCambrian basement reservoirs in Eastern China. All these
fields contain high fracture density.
In Type 2 naturally fractured reservoirs, fractures provide the
essential permeability, and the matrix provides the essential
porosity, such as in the Monterey fields of California, the
Spraberry reservoirs of West Texas, and Agha Jari and Haft
Kel oil fields of Iran.
In Type 3 naturally fractured reservoirs, the matrix has an
already good primary permeability. The fractures add to the
reservoir permeability and can result in considerable high flow
rates, such as in Kirkuk field of Iraq, Gachsaran field of Iran,
and Dukhan field of Qatar. Nelson includes Hassi Messaoud
(HMD) in this list. While indeed there are several lowpermeability zones in HMD that are fissured; in most zones
however the evidence of fissures is not clear or unproven.
In Type 4 naturally fractured reservoirs, the fractures are filled
with minerals and provide no additional porosity or
permeability. These types of fractures create significant
reservoir anisotropy, and tend to form barriers to fluid flow
and partition formations into relatively small blocks. Nelson
discusses three main factors that can create reservoir
anisotropy with respect to fluid flow: fractures, crossbedding
and stylolite. The anisotropy in Hassi Messaoud field, for
instance, appears to be the result of a non-uniform
combination of all three factors with varying magnitude from
zone to zone. Stylolites, just like fractures, are a secondary
feature. They are defined as irregular planes of discontinuity
between two rock units. Stylolites, which often have fractures
associated with them, occur most frequently in limestone,

SPE 104056

dolomite, and sandstone formations. Mineral-filled fractures

and stylolites can create strong permeability anisotropy within
a reservoir. The magnitude of such permeability is extremely
dependent on the measurement direction, thereby requiring
multiple-well testing. Interference testing is ideal for
quantifying reservoir anisotropy and heterogeneity, because
they are more sensitive to directional variations of reservoir
properties, such as permeability, which is the case of type 4
naturally fractured reservoirs.
It is important to take this classification into consideration
when interpreting a pressure transient analysis for the purpose
of identifying the type of fractured reservoir and its
characteristics. Each type of naturally fractured reservoir may
require a different development strategy. Ershaghi2 reports
that: (a) Type 1 fractured reservoirs, for instance, may exhibit
sharp production decline and can develop early water and gas
coning; (b) Recognizing that the reservoir is a type 2 will
impact any infill drilling or the selection of improved recovery
process; (c) In Type 3 reservoirs, unusual behavior during
pressure maintenance by water or gas injection can be
observed because of unique permeability trends.
PROPERTIES
OF
MATRIX
BLOCKS
AND
FRACTURES
A naturally fractured reservoir is composed of a
heterogeneous system of vugs, fractures, and matrix which are
randomly distributed. Such type of system is modeled by
assuming that the reservoir is formed by discrete matrix block
elements separated by an orthogonal system of continuous and
uniform fractures which are oriented parallel to the principal
axes of permeability. Two key parameters, and , were
introduced by Warren and Root3 to characterize naturally
fractured reservoirs. These dimensionless parameters and
are mathematically expressed as3:

( ct ) f
( ct ) t

k r2
= m w2
k f xm

( ct ) f
( ct ) f + ( ct ) m

.. (1)

(2)

The geometry parameter, , is defined as:

= 4n(n + 2) .. (3)
where n is 1, 2 or 3 for the slab, matchstick and cube
models, respectively.
Assuming:(a) the flow between the matrix and the
fractures is governed by the pseudo-steady state condition, but
only the fractures feed the well at a constant rate, and (b) the
fluid is single phase and slightly compressible, the wellbore
pressure solution and the pressure derivative in an infiniteacting reservoir are given by4,5:
t D
t
(4)
1
Ei D + s
PD = ln t D + 0.80908 + Ei

(
1

)
(
1

)
2

t
t D .. (5)
1

t D PD ' = 1- exp D +exp

(
1
)

(1- )

The second pressure derivative of the dimensionless

pressure equation is:
t D ... (6)
t D
1
(t D PD ' )' =

2(1 )

exp
exp
(1- )
(1- )

(A) Semilog Analysis

A plot of the well pressure or pressure change (P) versus
test time on a semilog graph should yield two parallel straight
line portions as shown in Figure 1. The pressure change P
during a drawdown test is (Pi - Pwf). During a buildup test P
= (Pws Pwf(t=0)).
1. Fracture Permeability
Figure 1 shows two well defined parallel straight lines of
slope m. The slope m of the straight lines may be used to
calculate the average permeability of the fractured system or
the kfh product:
162.6qBo
kh =
.(7)
m
Assuming the sugar cube model is valid and Types 1
naturally fractured reservoirs, the product kh is essentially
equal to (kh)f, so the slope of either straight line can be used to
determine kh.
In Type 2 naturally fractured reservoirs the first straight
line is mostly related to fracture flow, and therefore the kh
product in Eq 7 is essentially (kh)f. The second straight line is
however related to both fracture flow and matrix flow, thus the
kh product in Eq 7 reflects both (kh)m and (kh)f. In this case it
is unlikely that the two straight lines will be perfectly parallel.
If however (kh)m << (kh)f then kh can be approximated by
(kh)f.
In Type 3 reservoirs, both straight lines are related to
fracture flow and matrix flow, the product kh in Eq 7 is
therefore equivalent to (kh)t.

2. Skin Factor
The skin factor is obtained using conventional technique,
i.e.:
( P )

k
+ 3.23 .. (8)
1hr
log
s = 1.1513
(c )
m

r 2
t f +m w

(P)1hr is taken from the second straight line.

3. Fracture Storage Capacity Ratio
The vertical distance between the two semilog straight
lines, P, may be used to estimate3 the storage capacity ratio,
:
P

= exp 2.303
... (9)
m

or
= 10 P / m (10)
In Type 4 naturally fractured reservoirs the value of is
close to unity. The sugar cube model is not realistic in Type 4

SPE 104056

fractured reservoirs, since the fractures do not provide

additional porosity or permeability. These reservoirs are best
treated as anisotropic and analyzed accordingly.
4. Interporosity Flow coefficient
A characteristic minimum point, or trough, is typically
observed on the pressure derivative plot for naturally fractured
reservoirs, as shown in Figure 2. This minimum takes place at
the point where the second pressure derivative equals zero
(tDPD) = 0. The dimensionless time at which this minimum
point occurs is given by the following expression4, 5, 6
t D min =

1
ln ... (11)

On the semilog plot of well pressure versus test time, this

minimum point corresponds to the inflection point during the
transition portion of the curve. Therefore, Eq. 11 can be
rewritten as:
1
t D inf = ln . (12)

The dimensionless time is defined as:
t D inf =

0.0002637k tinf
(ct ) f + m rw 2

.(13)

Where tinf = tmin. Combining Eqs. 12 and 13 and solving

for , yields a new relationship for the interporosity flow
parameter:
3792(ct ) f +m rw
1
=
ln
k tinf

.. (14)

tinf can be directly read at the inflection point of the

pressure curve from a semilog plot of the flowing well
pressure versus test time. For a Miller-Dyes-Hutchinson
(MDH) semilog plot, i.e. shut-in well pressure (Pws) versus
shut-in time (t), tinf = tinf. When using a Horner plot, the
corresponding inflection (Horner) time, (HT)inf, is read and
converted to inflection time using the following equation:

tinf =

tp
( H T )inf 1

... (15)

Where (HT) is the Horner time (tp+t)/t or the effective

Horner time tpt/(tp+t).
The idea of estimating the interporosity flow parameter
from semilog plots is not new. Uldrich and Ershaghi7,
formulated a complex and cumbersome procedure for that
purpose. They introduced one equation for pressure drawdown
tests which uses the coordinates of the inflection point time,
the storage capacity ratio, the skin factor and a parameter read
from a plot which is a function of . They also introduced

another equation for pressure buildup tests which utilizes the

inflection point time, the storage capacity ratio, the
dimensionless Horner production time, tD, and two parameters
read from two different plots. These two graphically-obtained
parameters are also function of the value. These equations
have received limited applications. Bourdet and Gringarten8
suggest plotting a horizontal line through the approximate
middle of the transition portion of the curve, and then use the
time at which this horizontal line intersects the parallel straight
lines to calculate the storativity ratio, , and the interporosity
flow coefficient, . Eq. 14 offers a much simpler and
analytically sound procedure for calculating from the
conventional semilog analysis.
5. Short buildup Test Second Straight is not observed
The interpretation of a buildup test is similar to that of a
drawdown. Generally, the second straight line is more likely to
be observed than the first one, which often is masked by near
wellbore effects, such as wellbore storage. In Type 3 naturally
fractured system, where the matrix has a high enough
permeability for the fluid to enter the wellbore both from the
fracture (mostly) and the matrix, then the first straight line
should last a long time, and will not be masked by inner
wellbore effects. In this system, it is also possible for an
unsteady state flow regime to develop in the matrix. This flow
regime will appear during the transition period, i.e. after the
first semilog straight line.
However pressure buildup tests often give more reliable
value of the storage capacity ratio, , especially when the
second parallel straight line is not observed, such as when the
pressure test is too short, or the well is near a boundary. In
these cases it impossible to determine p, and consequently
Eq. 10 can not be used. The equation of the early time straight
line can be represented by9:
t p + t

+ log
Pws = Pi m log
..... (16)

1 +
t

Extrapolating the first straight line to a Horner time of

unity, i.e. (tp+t)/t = 1, where Pws=PFF1, then the storage
capacity ratio can be calculated from:

10 ( Pi PFF 1 ) / m
1 10 ( Pi PFF 1 ) / m

.....(17)

PFF1 stands for Fracture Flow pressure, since near the

wellbore, fluid flows into the well exclusively through the
fractures, particularly in Types 1 and 2 naturally fractured
reservoirs. PFF1 will always be greater than (by a value equal
to p) the average pressure, Pi and P*, since normally the
second parallel line is used to estimate these three pressure
values. If the initial reservoir pressure Pi is not available, use
the average reservoir pressure instead, or the false pressure P*
(if it is known from another source).
The vertical distance between the two parallel semilog
straight lines and passing through the inflection point is of
course identified as p. For uniformly distributed matrix

SPE 104056

blocks, the inflection point is at equal distance between the

two parallel lines. Therefore

= 10

2 P1 inf
m

.... (18)

Where:
P1inf (= 0.5P) is the pressure drop between the 1st
semilog straight line and the inflection point along a vertical
line parallel to the pressure axis.
Equation 18 is analogous to Eq. 10 for calculating the
storage capacity ratio, and therefore should yield the same
results as long as the first straight line is well defined and the
pressure test is run long enough to observe the trough on the
pressure derivative, and therefore the inflection point on the
semilog plot. The interporosity flow coefficient is then
calculated from Eq. 14.
If the inflection point is difficult to determine, then read
the end-time of the first or early time straight line, tEL1, and
use the following equation to estimate :
(ct ) f + m rw
(1 ) ........... (19)
=
0.013185kt EL1

If the buildup test is however too short to even observe the

trough (which provides the best evidence of a naturally
fractured system), then results obtained from the interpretation
of the test should at best be considered as an approximation.
The skin factor is then obtained from the following
equation:
(P ) ( P P )

k
1hr
i
FF 1
log
s = 1.1513
(c )

m
r 2
t f +m w

+ 3.23

(20)

or
(P ) 2P

k
1hr
1inf
log
s = 1.1513
(c )

m
r 2
t f +m w

+ 3.23 .

(21)

where (P)1hr is taken from the first straight line.

EXAMPLE 1
Given the build up test data in Table 1 and the following
formation and fluid properties, estimate formation
permeability, skin factor, , and from.
q = 125 STB/D
tp = 1200 hr
pwf = 211.20 psia
= 1.72 cp
ct =7.1910-6 psi-1

h = 17 ft
= 13.0%
rw = 0.30 ft
B=1.054 RB/STB

Solution
The following data are read from Figure 3:
tinf = 0.63 hr
P1inf = 33 psi
P1hr = 497 psi
m=35.67 psi/cycle
tEL1 = 0.012 hr

From Equation 7:
162.6(125)(1.054)(1.72)
k=
= 60.7 md
(35.67)(17)
From Equation 21 the storage capacity ratio is:

= 10

2(33)
35.67

= 0.014

Using equation 1, we can calculate (ct)f:

(c t ) f = (c t ) m

1
0.014
8
(c t ) f = (0.13)(7.19 10 6 )
= 1.3 10
1 0.014
From equation 21 the skin factor is:
(285.8 2 33)

60.7
s = 1.1513
log
(0.13 7.19 10 6 + 1.3 10 8 )(1.72)(0.3) 2
35.67

s = 0.89

+ 3.23

From Equation 14, the interporosity flow parameter is:

3792(0.13 7.19 10 6 + 1.3 10 8 )(1.72)(0.3) 2
1
0.014 ln

(60.7)(0.63)
0.014

= 8.7 10 7

From Equation 19:

(0.13 7.19 10 6 + 1.3 10 8 )(1.72)(0.3) 2
(1 0.014)0.014

(0.013185)(60.7)(0.012)

7
= 2.1 10

6. Long buildup Test First Straight is not Observed

Generally, the second straight line is more likely to be
observed than the first one, which often is masked by near
wellbore effects, such as wellbore storage. In Type 1 and Type
2 naturally fractured systems, where the matrix permeability is
negligible, the fluid flows into the wellbore exclusively
through the fractures. The first straight line will probably be
too short and easily masked by inner wellbore effects.
The permeability and skin factor are calculated from Eqs.
7 and 8 respectively. The following equation provides a direct
and accurate method for calculating , as long as the
inflection point and the second straight line are observed and
the matrix blocks are uniformly distributed:

= 10

2 P2 inf
m

.... (22)

P2inf (= 0.5p) is the pressure drop between the 2nd

semilog straight line and the inflection point along a vertical
line parallel to the pressure axis.
The interporosity flow parameters is then calculated from
Eq. 14.
If the inflection point is difficult to determine, then read
the starting-time of the second semilog straight line, tSL2, and
use the following equation to estimate :

SPE 104056

(ct ) f + m rw 2
(1 ) ..... (23)
5.27 10 5 kt
SL 2

EXAMPLE 2
Given the build up test data in Table 2 and the following
formation and fluid properties, estimate formation
permeability, skin factor, , and .
q = 125 STB/D
tp = 1200 hr
pwf = 211.20 psia
= 1.72 cp
ct =7.1910-6 psi-1

h = 17 ft
= 13.0%
rw = 0.30 ft
B=1.054 RB/STB.

Solution
The following data are read from Figure 4:
tinf = 3.05 hr
P2inf = 24 psi
P1hr = 419 psi
m=30 psi/cycle
tSL2 =55 hr

162.6(125)(1.054)(1.72)
= 72.25 md
(17)(30)

From Equation 22:

= 10

2 ( 24 )
30

k=

(c t ) f = (c t ) m

From equation 8:
= 1.1513

72.25
+ 2.4 10

)(1.72)( 0.3)

s = 1.69

From Equation 14:

3792(0.13 7.19 10 6 + 2.4 10 8 )(1.72)(0.3) 2
1
0.025 ln

(72.25)(3.05)
0.025

7
= 2.36 10

70.6qB
.... (25)
h(t P' ) R

where (tP')R is obtained by extrapolating the horizontal

line to the vertical axis. In order for the conventional semilog
analysis and the TDS technique to yield the same value of k,
the following equation must be true:

0.025
8
(c t ) f = (0.13)(7.19 10 6 )
= 2.4 10
1

0
.
025

70.6qB
.. (24)
kh

The subscript R stands for radial flow. The formation

permeability is therefore:

It is possible to calculate (ct)f by:

log
( 0.13 7.19 10 6

(B) TDS Technique

In 1993 Tiab introduced a technique10 for interpreting
loglog plots of the pressure and pressure derivative curves
without using type curve matching. This technique utilizes the
characteristic intersection points, slopes, and beginning and
ending times of various straight lines corresponding to flow
regimes strictly from loglog plots of pressure and pressure
derivative data. Values of these points and slopes are then
inserted directly in exact, analytical solutions to obtain
reservoir and well parameters. This procedure for interpreting
pressure tests, which is referred to as the Tiabs Direct
Synthesis (TDS) technique offers several advantages over the
conventional semilog analysis and type curve matching. It has
been applied to over fifty different reservoir systems11-18, and
hundreds of field cases.

(t P ' ) R =

= 0.025

207.8

30

(0.13 7.19 10 6 + 2.4 10 8 )(1.72)(0.3) 2

(1 0.025)

(5.27 10 5 )(72.25)(55)

7
= 6.91 10

1. Fracture Permeability
The pressure derivative portion corresponding to the
infinite acting radial flow line is a horizontal straight line. This
flow regime is given by10:

From Equation 7:
k=

From Equation 23:

+ 3.23

m = 2.303(t P' ) R ..... (26)

2. Skin Factor
The second radial flow line can also be used to calculate
the skin factor from10:
(P )

kt R 2
R2
ln
s = 0.5
2

(t P') R 2
(c t ) f + m rw

+ 7.43 (27)

Where tR2 is any convenient time during the systems

radial flow regime (as indicated by the horizontal line on the
pressure derivative curve, Figure 2) and (P)R2 is the value of
P on the pressure curve corresponding to tR2. If the test is too
short or the boundary is too close to the well to observe a well
defined second straight line, then the skin factor can be
estimated from the early-time horizontal straight line:

SPE 104056

(P )

kt R1
1
R1
(28)
ln
+
7
.
43
s = 0.5
(c t ) f + m rw2
(t P') R1

Where tR1 is any convenient time during the early-time

radial flow regime (as indicated by the horizontal line on the
pressure derivative curve, Figure 2) and (P)R1 is the value of
P on the pressure curve corresponding to tR1.

3. Interporosity Flow Coefficient

The interporosity flow parameter can also be obtained
from the loglog plot of the derivative function (txP) versus
test time4,5 by substituting the coordinates of the minimum
point of the trough, tmin and (txP)min:
42.5h(ct ) f + m rw 2 (t P ')
min

... (29)
qBo
t min

The advantage of Eq. 29 over Eq. 14 is that it is

independent of permeability and storage capacity ratio, and the
coordinates of the minimum points are easier to determine
than the inflection point on the semilog plot.

4. Storage Capacity Ratio

The coordinates of the minimum point of the trough can be
used to derive two equations to calculate accurately the
storage capacity ratio .
Pressure derivative Coordinate: Using the pressure
derivative coordinate of the minimum point and the radial
flow regime (horizontal) line, the following equation provides
a direct and accurate method for calculating :

= 10

( t P ') min
0.8684 1
( t P ') R

... (30)

Equation 30 is derived by observing that:

(t P ) R (t P ) min =

P
2

.... (30a)

Combining Equations 30a and 26 yields:

P
m

2((t P ) R (t P) min )
2.303(t P ) R

... (30b)
(t P ) min
= 0.86841

(t P ) R

Substituting Equation 30b into Equation 10 yields

Equation 30. Equation 30 assumes wellbore storage and
boundary effects do not influence the trough and the infinite
acting radial flow line is well defined.
In conventional analysis this ideal case displays two well
defined parallel lines with the inflection point equidistant of
those two lines, which means that the fractures are uniformly
distributed.

Minimum Time: Using the time coordinate of the minimum

point, a less direct but just as accurate value of the storage
capacity ratio can be obtained when wellbore storage is
present from the following equation:

= e t D min ..... (31)

Where the dimensionless time at the minimum point is
calculated from:
0.0002637 k
t D min =
2
(c )
t f + m rw

t ... (32)
min

Solving explicitly for Eq. 31 yields19:

= 2.9114

3.5688 6.5452

N S
ln( N S )

. (33)

Where the parameter NS is given by:

N S = e tD min (34)
Eq. 34 is obtained by assuming values of , from 0 to 0.5,
then values of = NS were plotted against . The resulting
curve was curve-fitted. Note that Eq. 33 can also be used in
the semilog analysis since tmin = tinf.
It is recommended that both methods be used for
comparison purposes. If the radial flow regime line on the
derivative curve is not well defined due to a combination of
inner and/or outer boundary effects or a short test, but the
minimum of the trough is well defined, then Eqs. 29 and 33
should be used to calculate, respectively, and .
EXAMPLE 3
Tiab Direct Synthesis technique is applied to Example 2.
Figure 5 is plotted with data from Table 2 and the
respective pressure derivative.
From Figure 5 the following data can be read:
PR = 274.51 psi
tmin= 3.05 hours
(tP)R = 13 psi
Pe = 13 psi

tR= 156.51 psi

(tP)min = 1.3 psi
te=0.018 hr

Wellbore storage coefficient is calculated by10:

C=

qB t
(125)(1.054) 0.018
3

= 7.60 10 bbl / psi

24 P e
24
13

From Equation 25:

k=

70.6(125)(1.72)(1.054)
= 72.39 md
(17)(13)

From Equation 27:

274.51

(72.39)(156.51)
+ 7.43
s = 0.5
ln
(0.13 7.19 10 6 + 2.4 10 8 )(1.72)(0.3) 2
13

s = 1.74

SPE 104056

From Equation 29:

42.5(17)(0.13 7.19 10 6 + 2.4 10 8 )(0.3) 2 1.3
=

(125)(1.054)

3.05
= 2.02 10 7

From Equation 33 the storage capacity ratio can be

calculated in the presence of wellbore storage:

3.5688
6.5452

= 2.9114

ln(
0
.
924
)
0.924

= 0.024

Table 3 is a comparison of the TDS results with that of

conventional method.
FRACTURE POROSITY AND COMPRESSIBILTY
Once is estimated, the fracture porosity can be estimated
if matrix porosity, m, total matrix compressibility, ctm, and
total fracture compressibility, ctf, are known, as follows:
ctm
m .. (35)

1 ctf
Fracture compressibility may be different from matrix
compressibility by an order of magnitude. Naturally fractured
reservoirs in Kirkuk field (Iraq) and Asmari field (Iran) have
fracture compressibility ranging from 4x10-4 to 4x10-5 psi-1. In
Grozni field (Russia) ctf ranges from 7x10-4 to 7x10-5. In all
these reservoirs ctf is 10 to 100 folds higher than ctm. Therefore
the practice of assuming ctf = ctm is not acceptable.
The fracture compressibility can be estimated from the
following expression9:

f =

ctf =

1 k f / k fi

1/ 3

1 k f / k fi

) .. (36)

3P
P
k fi = Fracture permeability at the initial reservoir pressure

pi
k f = Fracture permeability at the current average reservoir

pressure.
19

Combining Equations 35 and 36 yields :

P

.. (37)
f = m ctm

1/ 3

1
(
k

f / k fi )
In deep naturally fractured reservoirs, fractures and the
stress axis on the formation generally are vertically oriented.
Thus when the pressure drops due to reservoir depletion, the
fracture permeability reduces at a lower rate than one would
expect. In Type-2 naturally fractured reservoirs, where matrix
porosity is much greater than fracture porosity, as the reservoir
pressure drops the matrix porosity decreases in favor of
fracture porosity9. This not the case in Type-1 naturally
fractured reservoirs, particularly if the matrix porosity is very
low or negligible.
For fractured reservoirs and, indeed, all highly anisotropic
reservoirs, the geometric mean is currently considered the
most appropriate of the three most common averaging
techniques (arithmetic, harmonic and geometric). Therefore, a
representative average value of the effective permeability of a

naturally fractured reservoir may be obtained from the

geometric mean of kmax and kmin as illustrated in Figure 6.
k = k max k min .. (38)

where
kmax = maximum permeability measured in the direction
parallel to the fracture plane (Figure 6), thus
kmax kfracture
kmin = minimum permeability measured in the direction
perpendicular to the fracture plane (Figure 6), thus
kmin kmatrix
Substituting kf and km for, respectively, kmax and kmin,
Equation 38 becomes:
k = k f k m .. (39)

The fracture permeability can therefore be estimated from:

k2
.. (40)
km
Where km is the matrix permeability, which is measured
from representative cores and k is the mean permeability
obtained from pressure transient tests. Combining equations
36 and 40 yields:
kf =

ctf =

1 (k / k i )
P

2/3

.. (41)

Where
ki = average permeability obtained from a transient test run
when the reservoir pressure was at or near initial conditions
Pi and
k = average permeability obtained from a transient test at
the current average reservoir pressure.
P = Pi P
Combining Equations 41 and 35 yields19:
P

........ (42)
f = m ctm

2/3
1 1 (k / k i )
Matrix permeability is assumed to remain constant
between the two tests. Note that equations 37 and 42 are also
valid for calculating fracture porosity change between two
consecutive pressure transient tests, and therefore

P = P1 P2 . The time between the two tests must be long

enough for the fractures to deform significantly in order to
determine an accurate value of ctf. Table 5 shows pressure
transient analysis in Cupiaga field, a naturally fractured
reservoir in Colombia22. The reduction in permeability for
well 1 is about 13% and the change in pressure is 344 psi from
1996 to 1997. This type of data can be used in order to
estimate f from Eq. 42. Eq. 37 should yield a more accurate
value of fracture porosity than Eq. 42, as the latter assumes
Eq. 39 is always applicable.
Substituting the values of k m , k f , and f into the
following equation should yield approximately the same value
of the effective permeability obtained from well testing20:
k k m + f k f .. (43)

SPE 104056

Eq. 43 should only be used for verification purposes. The

fracture width or aperture may be estimated20 from
wf =

kf

Using the TDS technique, the value of k is obtained from

Equation 25:
k=

........ (44)

70.6(3000)(1)(1.25)
= 72.53 mD
(25)(146)

33t

where: fracture width = microns, permeability = mD,

porosity = fraction, and storage capacity = fraction.
EXAMPLE 4
Pressure tests in the first few wells located in a naturally
fractured reservoir yielded a similar average permeability of
the system of 82.5 mD. An interference test also yielded the
same average reservoir permeability, which implies that
fractures are uniformly distributed. The total storativity,
(ct)m+f = 1x10-5 psi-1 was obtained from this interference test.
Only the porosity, permeability and compressibility of the
matrix could be determined from the recovered cores.
The pressure data for the well are given in Table 4. The
pressure drop from the initial reservoir pressure to the current
average reservoir pressure is 300 psi. The characteristics of the
rock, fluid and well are given below:
q = 3000 STB/D
m = 10%
= 1 cp
ctm=1.3510-5 psi-1

The inter-porosity fluid transfer coefficient is given by

Equation 29:
(42.5)(25)(1 10 5 )(0.4) 2 (70.5)
5

2.5 = 1.28 10
(3000)(1.25)

Since the two parallel lines are well defined the storage
coefficient is calculated from Equation 30

= 10

= 0.35

The conventional semilog analysis yields similar values of

k, and as the TDS technique. The main reason for this
match is that both parallel straight lines are well defined.
2 Current properties of the fracture
(a) The fracture permeability is calculated from Equation.
40:
kf =

h = 25 ft
rw = 0.4 ft
B=1.25 RB/STB.
km=0.10 mD

1 - Using conventional semilog analysis and TDS

technique, calculate the current formation permeability,
storage capacity ratio, and fluid transfer coefficient
2 Estimate the three fracture properties: permeability,
porosity and width.

70.5
0.8684 1

146

k 2 72.532
=
= 52,606 mD
km
0.10

The fracture permeability at initial reservoir pressure is:

k fi =

k i2 82.5 2
=
= 68,062 mD
km
0.10

(b) The fracture porosity

In fractured reservoirs with deformable fractures, the
fracture compressibility changes with declining pressure. The
fracture compressibility can be estimated from Equation 41:
1 (52,606 / 68,062)
300

2/3

Solution
1(a) Conventional method
From Figure 7:
P = 130 psi m=325 psi/cycle tinf =2.5 hrs
The average permeability of the formation is estimated
from the slope of the semilog straight line. Using Equation 7
yields:
162.6(3000)(1.25)(1)
k=
= 75.05 mD
(325)(25)

ctf =

= 5.2 10 4 psi 1

The compressibility ratio is:

ctf
ctm

5.2 104
= 38.5
1.35 10 5

Thus, the fracture compressibility is more than 38.5 folds

higher than the matrix compressibility, or c tf = 38.5c tm .

The fracture porosity from Equation 42 is:

Fluid storage coefficient is estimated using Equation 10:

= 10 ( 130 / 325) = 0.39

0.35 0.1
= 0.00139 0.14%

1 0.35 38.5

f =

The storage coefficient of 0.39 indicates that the fractures

occupy 39% of the total reservoir pore volume.

The total porosity of this naturally fractured reservoir is:

The inter-porosity fluid transfer coefficient is given by

Equation 14:

Substituting the values of k m , k f , and

3792 1 10 5 (1)(0.4 2 )
1
5
0.39 ln
= 1.19 10
(75.05)(2.5)
0.39

1(b) TDS technique

From Figure 8, the following characteristic points are read:
(tP)R = 146 psi
tmin = 2.5 hrs
(tP)min = 70.5 psi

t = m + f = 0.10 + 0.0014 = 0.1014

43:

into Equation

k k m + f k f = 0.1 + 0.0014 52,606 = 73.7 mD

This value is approximately the same value of the effective

permeability obtained from well testing (72.53 mD). The
fracture width or aperture may be estimated from Equation 44:
wf =

52,606
= 212 microns = 0.212 mm
33 0.35 0.1014

SPE 104056

The fracture width is a useful parameter for identifying the

nature of fracturing in the reservoir.
Conclusions
1. The inflection point on the semilog plot of well pressure
versus test time and the corresponding minimum point on
the trough of the pressure derivative curve are unique
points that can be used to characterize a naturally
fractured reservoir.
2. The interporosity flow parameter can be accurately
obtained from the conventional semilog analysis if the
inflection point is well defined and the new proposed
equation is utilized. The equation is valid for both
pressure drawdown and pressure buildup tests.
3. Two new equations are introduced for accurately
calculating the storage capacity ratio from the coordinates
of the minimum point of the trough on the pressure
derivative curve.
4. For a short test, in which the late-time straight line is not
observed, the storage capacity ratio and the interporosity
flow coefficient can both be calculated from the inflection
point.
5. For a long test, in which the early-time straight line is not
observed, due to near-wellbore effects, the storage
capacity ratio can also be calculated from the inflection
point.
6. A new equation is proposed for calculating fracture
porosity, as a function of reservoir compressibility.
7. The practice of assuming the total compressibility of the
matrix (ctm) is equal to the total compressibility of the
fracture (ctf) should be avoided. From field observations,
ctf is several folds higher than ctm.

Subscripts
i
initial
o
oil
D
dimensionless
f
fracture, fissure
m
matrix
t
total
inf
inflection point
min
minimum
1
1st semilog straight line
2
2nd semilog straight line
1hr
1 hour
References
1.
2.
3.
4.
5.

6.

7.

Nomenclature
B
oil volumetric factor, rb/STB
c
system compressibility, psi-1
h
formation thickness, ft
HT
Horner time, dimensionless
k
permeability, md
m
semilog slope, psi/log cycle
Pws
well shutin pressure, psi
Pwf
well flowing pressure, psi
q
oil flow rate, BPD
rw
wellbore radius, ft
s
skin factor
tp
producing time before shut-in, hrs
wf
Fracture width in microns
Greek Symbols
P
vertical distance between the two semilog straight
lines, psi

Geometry parameter, 1/L2

Porosity, dimensionless
P1inf Pressure drop between the 1st semilog strigth line and
the inflection point, psi
P2inf Pressure drop between the 2nd semilog strigth line and
the inflection point, psi
t
shut-in time, hrs

Interporosity flow parameter, dimensionless

Viscosity, cp
Storage capacity ratio, dimensionless

8.

9.
10.

11.

12.

13.

Nelson, R.: Geologic Analysis of Naturally Fractured

Reservoirs. Gulf Professional Publishing, 2nd Edition. 2001
Ershaghi, I.: Evaluation of Naturally Fractured Reservoirs.
IHRDC, PE 509, 1995.
Warren, J.E. and Root, P.J.: The Behavior of Naturally
Fractured Reservoirs. Soc. Pet. Eng. J. (Sept. 1963): 245-255.
Trans. AIME, 228.
Engler, T. and Tiab, D.: Analysis of Pressure and Pressure
Derivative without Type Curve Matching, 2. Naturally Fractured
Reservoirs. Journal of Petr. Sci. and Eng. 15 (1996):127-138.
Engler, T. and Tiab, D.: Analysis of Pressure and Pressure
Derivative without Type Curve Matching, 5. Horizontal Well
Tests in Naturally Fractured Reservoirs. Journal of Petr. Sci.
and Eng. 15 (1996); 139-151.
Engler, T. and Tiab, D.: Analysis of Pressure and Pressure
Derivative without Type Curve Matching - 6. Horizontal Well
Tests in Anisotropic Media. Journal of Petroleum Science and
Engineering, Vol. 15 (Aug. 1996) N0. 2-4, 153-168.
Uldrich, D.O. and Ershaghi, I.: A Method for Estimating the
Interporosity Flow Parameter in Naturally Fractured
Reservoirs: Paper SPE 7142, Proceedings, 48th SPE-AIME
Annual California Regional Meeting held in San Francisco, CA,
Apr. 12-14, 1978.
Bourdet, D. and Gringarten. AC.: Determination of fissured
volume and block size in fractured reservoirs by type-curve
analysis. Paper SPE 9293. Soc. Pet. Eng., Annu. Tech. Conf.,
Dallas, TX, Sept. 21-24, 1980,
Saidi, M. A.: Reservoir Engineering of Fractured Reservoirs.
Total Edition Presse, 1987.
Tiab, D.: "Analysis of Pressure and Pressure Derivative without
Type-Curve Matching - 1. Skin and Wellbore Storage". Journal
of Petroleum Science and Engr., Vol. 12, No. 3 (January, 1995)
171-181.
Jongkittinarukorn, K. and Tiab, D.: Analysis of Pressure and
Pressure Derivative without Type Curve Matching - 6. Vertical
Well in Multi-boundary Systems. Proceedings, CIM 96-52, 47th
Annual Tech. Meeting, Calgary, Canada, June 10-12, 1996.
Jongkittinarukorn, K. and Tiab, D.: Analysis of Pressure and
Pressure Derivatives without Type Curve Matching - 7.
Horizontal Well in a Closed Boundary Systems, Proceedings,
CIM 96-53, 47th Annual Tech. Meeting, Calgary, Canada, June
10-12, 1996.
Tiab, D., Azzougen, A., F.H., Escobar, and S. Berumen:
Analysis of Pressure Derivative Data of Finite-Conductivity
Fractures by the Tiabs Direct Synthesis Technique. Paper SPE
52201.
Proceedings,
SPE
Mid-Continent
Operations
Symposium, Oklahoma City, 28 31 March 1999; Proceedings

10

14.

15.
16.

17.

18.

19.
20.

21.
22.

SPE 104056

SPE Latin American & Caribbean Petr. Engr. Conf., Caracas,

Venezuela, 2123 April 1999, 17 pages.
Mongi, A. and Tiab, D.: Application of Tiabs Direct Synthesis
Technique to Multi-rate Tests, SPE/AAPG 62607, Proceedings,
Western Regional Meeting, Bartlesville, California, 19-23 June
2000.
Benaouda, A. and Tiab, D.: Application of Tiabs Direct
Synthesis Technique to Gas Condensate Wells. Proceedings,
SPE Permian Basin Conference, Texas, May 2001
Jokhio, S.A., Hadjaz, A. and Tiab, D.: Pressure falloff Analysis
in Water Injection Wells Using the Tiabs Direct Synthesis
Technique. Paper SPE 70035, Proceedings, SPE Permian Basin
Conference, Midland, Texas, May 15-16, 2001.
Bensadok A. and Tiab, D.: Interpretation of Pressure Behavior
of a Well between Two Intersecting Leaky Faults Using Tiabs
Direct Synthesis (TDS) Technique. CIP2004-123, Proceedings,
Canadian International Petroleum Conference, 7 10 June 2004
Chacon, A., Djebrouni, A. and Tiab, D.: Determining the
Average Reservoir Pressure from Vertical and Horizontal Well
Test Analysis Using Tiabs Direct Synthesis Technique. Paper
SPE 88619, Proceedings, Asia Pacific Oil and Gas Conference
and Exhibition, Perth, Australia, Oct. 18-20, 2004.
Tiab, D. and E.C. Donaldson: Petrophysics: theory and
practice of measuring reservoir rock and fluid transport
properties. Gulf professional Publications, 2nd Edition, 2004.
Bona, N., Radaelli, F., Ortenzi, A., De Poli, A., Pedduzi, C. and
Giorgioni, M: Integrated Core Analysis for Fractured
Reservoirs: Quantification of the Storage and Flow Capacity of
Matrix, Vugs, and Fractures. SPERE, Aug. 2003, Vol.6,
pp.226-233.
Stewart G. Ascharsobbi F. Well test interpretation for
Naturally Fractured Reservoirs. Paper SPE 18173.
Giraldo L. A., Chen Her-Yuan, Teufel L. W. Field Case Study
of Geomachanical Impact of Pressure Depletion in the LowPermeability Cupiaga Gas-Condensate Reservoir. SPE 60297.
SPE Rocky Mountain Regional/Low Permability Reservoirs
Symposium, Denve, CO, March 12-15, 200.

SI Metric Conversion Factors

bbl x 1.589873
E-01 = m3
*
cp x 1.0
E-03 = Pa-s
ft x 3.048*
E-01 = m
ft2 x 9.290304
E-02 = m2
psi x 6.894757
E+00 = kPa
*Conversion factor is excat.

Table 1. Pressure data for Example 1

Time

Pressure

hours

psi

psi

Horner Time

0.0000

211.20

0.00

0.0010

390.73

179.53

1200001.00

0.0023

404.32

193.12

521740.13

0.0040

413.00

201.80

300001.00

0.0062

419.73

208.53

193549.39

0.0090

425.39

214.19

133334.33

0.0128

430.36

219.16

93751.00

0.0176

434.81

223.61

68182.82

0.0239

438.82

227.62

50210.21

0.0320

442.43

231.23

37501.00

0.0426

445.66

234.46

28170.01

0.0564

448.48

237.28

21277.60

0.0743

450.87

239.67

16151.74

0.0976

452.84

241.64

12296.08

0.1279

454.36

243.16

9383.33

0.1673

455.46

244.26

7173.74

0.2190

456.20

245.00

5480.45

0.2850

456.65

245.45

4211.53

0.3720

456.90

245.70

3226.81

0.4840

457.03

245.83

2480.34

0.6300

457.11

245.91

1905.76

0.8200

457.18

245.98

1464.41

1.0670

457.27

246.07

1125.65

1.3890

457.39

246.19

864.93

1.8060

457.55

246.35

665.45

2.3500

457.75

246.55

511.64

3.0500

458.01

246.81

394.44

SPE 104056

11

Table 2 Pressure data for Example 2

Time
hours

Pressure
psi

Time
hours

Table 4. Pressure data for Example 4

Pressure
psi

0.0000

211.20

1.8060

456.85

0.0010

212.07

2.3500

457.47

0.0023

213.19

3.0500

457.80

0.0040

214.64

3.9700

458.15

0.0062

216.50

5.1600

458.58

0.0090

218.90

6.7100

459.14

0.0128

221.98

8.7300

459.84

0.0176

225.91

11.3500

460.73

0.0239

230.92

14.7600

461.85

0.0320

237.26

19.1800

463.23

0.0426

245.22

24.9400

464.92

0.0564

255.11

32.4200

466.95

0.0743

267.26

42.1500

469.35

0.0976

281.94

54.8000

0.1279

299.31

0.1673

Time
hours

Pws
psi

P
psi

(txP')
psi

4473.0

0.093

4373.4

99.60

84.473

0.177

4299.1

173.90

133.483

0.260

4246.1

226.90

146.776

0.343

4203.6

269.40

151.595

0.427

4173.8

299.20

157.618

0.510

4139.7

333.30

150.295

0.593

4118.5

354.50

141.355

0.677

4103.5

369.50

111.676

0.760

4086.4

386.60

99.694

0.927

4075.4

397.60

95.720

1.093

4060.3

412.70

87.234

1.260

4043.1

429.90

84.384

472.11

1.427

4032.2

440.80

76.719

71.2400

475.21

2.427

3997.0

476.00

70.469

319.31

92.6100

478.57

3.427

3971.3

501.70

77.268

0.2190

341.53

120.3900

482.11

4.427

3948.3

524.70

87.168

0.2850

365.13

156.5100

485.71

5.427

3931.6

541.40

95.595

0.3720

388.74

203.5000

489.29

6.427

3917.1

555.90

108.303

0.4840

410.60

264.5000

492.77

7.427

3898.4

574.60

122.336

0.6300

428.91

343.9000

496.09

9.427

3865.3

607.70

142.426

0.8200

442.40

447.0000

499.23

12.43

3824.2

648.80

137.651

1.0670

450.83

581.0000

502.15

14.43

3804.1

668.90

136.857

1.3890

455.12

720.0000

504.36

20.43

3758.7

714.30

Parameter

MDH

TDS

k, md

72.25

72.39

1.69

2.3610

0.025

Pressure and pressure derivative, psi

Table 3. Results for examples 2 and 3

1.74
-7

2.0210

-7

0.024

6,500

Shut-in Pressure, psia

6,400
P

6,300

Inflection point

6,000
5,900
5,800
5,700
100,000

t H-inf

10,000

1,000

Horner time,

(P)R2

(P)R1

(txP')R1

(txP')R2

100

(txP')min
tmin

tR1

Minimum
point

tR2

10
10

100

1000

10000

Time, hr
Figure 2 P and pressure derivative plot for a naturally fractured
reservoir

6,200
6,100

1000

100

10

tH= (tp + t)/t

Figure 1- Semilog pressure behavior of a naturally fractured

reservoir

12

SPE 104056
W

520

Inflection point
t inf= 0.63 hr

Shut in pressure, psi

500

P1hr =497 psi

480

P1inf = 33 psi

460

m = 35.67 psi/cycle

kmax

h = hf

440

kmin

420
400

Wf

tEL1 = 0.012 hr

380
0.001

0.01

0.1

10

Figure 6- Maximum and Minimum Permeability

Time, hr
4400

Figure 3 Conventional MDH plot for Example 1

4300
550
Inflection point
t inf= 3.05 hr

m = 325 psi/cycle
Pressure, psi

Shut in pressure, psi

500
450
400

P2inf = 24 psi

P1hr = 419 psi

4100

3900
3800

300

3700
3600
0.01

tSL2 = 55 hr

200
0.001

Inflection point
tinf = 2.5 hr

4000

350

250

P = 130 psi

4200

m = 30 psi/cycle

0.1

10

100

Time, hr

0.01

0.1

10

100

1000

Figure 7- Conventional MDH plot for Example 4

Time, hr
1000

Figure 4 -Conventional MDH plot for Example 2

PR = 669 psi

P and (txP'), psi

PR = 274.51 psi
100

te = 0.018 hr
(P)e = 13 psi
(tP')R = 13 psi

10

P and (txP'), psi

1000
(tP')R = 146psi
100
tmin = 2.5 hr
(tP')min = 70.5 psi
tR = 14.43 hr

tmin = 3.05 hr
(tP')min = 1.3 psi

10

tR = 156.51 hr

10
Time, hours

0.1
0.001

Figure 8- Pressure Derivative plot for Example 4

0.01

0.1

Time, hr
Figure 5- Log-log plot for Example 3

10

100

1000

100

SPE 104056

13

Table 5. Pressure Transient Analysis of Selected Cupiaga wells22

Test Type

P*

Date

BHFP

Global

Particular

K (md)

(psi)

Skin

Types

h(ft)

Comments/Remarks

of Skin
Well 1

6004

3000

91.3

Pre-Frac Test

Mechanical

16.4

48

171

Mechanical

16.4

18

171

Homogenous reservoir model. The turbulence factor is quite large

due to non-darcy flow (high rates) combined with the condensate
banking

1996
Well 1

6004

3200

38

Post-Frac Test

Homogeneous reservoir model. Rate dependent Skin was observed.

The effects of the condensate banking are observed in the GOR
response, at higher drawdowns the GOR increased.

1996
Well 1

5660

3410

20

14.2
171

Post-Frac Test

Homogeneous reservoir model. Some drainage area is still above

the dew point.

1997
Well 1

5150

3000

4.3

Drainage area

171

Below dew-point

Homogeneous reservoir model. Derivative curve indicates radial

flow with a low value of gas effective permeability out to a radius
of 800 ft followed by an increase in effective permeability further
out. This is interpreted as being due to liquid condensate drop-out.

1998
Well 1

5050

5380

Injector PFO

6267

3100

1995
Well 2

171

Homogeneous reservoir model. A small negative mechanical skin is

suggested possibly due to activation of fractures by injecting
pressure/temperature. The well is in an under-injecting situation.
Three layer model.

4600

2772

Total

Total Mechanical

8.35

19

429

Includes Turbulent
and Condensate
effects

0.6

1998

Well 2

11.4

than non -Darcy

1998
Well 2

Zero skin for

other

5500?

6947

1(@100 ft)

993

2.9

Injector PFO

993

1999

(@100 ft)

Homogeneous reservoir model. Entire drainage area is below dew

point pressure. The explanation for the reduced skin is that due to
rate dependent relative permeability and pressure dependent
saturation, the condensate impact on relative permeability is less
close to the wellbore than deeper in the reservoir.
Homogeneous reservoir model. P*is difficult to estimate because of
the variation of Kh with radius. Given that the pressure at 100 ft
radius is well above dew point, it is of some concern that Kh has
not been fully restored. At a radius of 350 ft the Kh reduces below
2000 md-ft.