\

+ + =
2 2 2
2
1 1 1
z y x
L L L
t o (2)
The pseudosteady state, numerically derived expression for shape factor (Kazemi, 1976, Kazemi and Gilman, 1988) has a
coefficient of 4 rather than t
2
. A transient state, analytically derived shape factor was given by Penuela, et. al. (2002), but is
not generally used in field scale simulations. Current commercial simulators allow this shape factor to vary on a cellbycell
basis and thus allow for complex characterization of fractured reservoirs.
A generalized shape factor based on water imbibition experiments (Kazemi, Gilman and ElSharkawy, 1992, Zhang, et. al.,
1996) is:
=
=
I
i i
i
d
A
V
1
1
o (3)
In the above equation, A designates the area of the exposed fracture surface i for the given matrix block with a volume
V, d is the distance from the center of the block to surface i, and I is the number of exposed fracture surfaces. This equation is
also the most general form for upscaling from complex fracture patterns (e.g., DFN network models) to equivalent dual
porosity media for both single and multiphase flow. Since flow between matrix and fractures is linearly dependent on o,
this equation shows how matrixfracture surface area is an important factor in fractured reservoirs.
Rather than shape factor, geologists often refer to fracture intensity using nomenclature such as P10, P32 or P33 (Dershowitz
et. al., 2000). P33 is equivalent to fracture porosity (pore volume per rock volume). DFN models are often built using a
volumetric measure of intensity, P32, which is the fracture surface area per volume of reservoir. This measure has many
desirable scaling properties that make it mathematically convenient. However, it is not possible to measure this parameter
directly. Fortunately, it can be linearly related to other measurable properties, and this relation can be computed through
simulation (Dershowitz et. al, 2000). Once this relation is established, it is possible to calculate the desired P32 value from
the number of fractures per meter (P10) or from fault traces per unit area (P21). In fact, for some ideal fracture distributions,
Wang (2005) has developed a correlation between P10 and P32. Eq. 4 shows how P32 has a similar form to the value for o,
thus they should be closely related. (Here A refers to the fracture area between two exposed matrix surfaces, so is lower by
a factor of two from A in the previous equation for shape factor (o). In our experience, and from theoretical calculations, o
is proportional to P32 squared. Additionally fracture porosity and fracture permeability are theoretically directly proportional
to P32; however because of wide variability in fracture width, length, roughness and connectivity, we do not generally see a
simple correlation.
CSUG/SPE 146580 5
=
=
I
i
i
A
V
P
1
1
32 (4)
In practice, shape factor is often considered a history matching parameter, but modern workflows start with the above
equations as the best estimate of the fracture intensity (which can be converted to shape factor). The shape factor can vary
over the computational grid and is only considered a constant over the REV (grid volume). Note that the numerical models
do not actually require any discrete representation of matrix sugar cubes. The idealization of sugar cubes is just a
convenience for illustration similar to the bundle of capillary tubes used for discussion of a singleporosity media. The dual
porosity models are simply two overlying porous media with a transfer term directly proportional to shape factor (i.e.
fracturematrix surface area). The shape factor can be calculated for any geometry (Heinemann and Mittermeir, 2006; Sarma,
and Aziz, 2003).
Theoretically the shape factor can be calculated from pressure buildup curves on individual wells. Ideally, such curves
develop two parallel straightline segments, indicative of dualporosity behavior (Warren and Root, 1963). The buildup time
at the inflection point, on the transition segment between the early and late straight line buildup segments, is related to shape
factor and depends on two dimensionless parameters, the storativity and the interporosity flow, e and , which are defined
as follows:
e

 
=
+
( )
( ) ( )
c
c c
t f
t f t m
(5)
e
w m
k
r k
2
o
= (6)
These relations are important for the classification plot presented in this paper. The inflection point on the transition between
early and late straightline flow periods depends on the shape factor according to the following relation (time is given in
hours):
e
w f t
k
r n c
t
e 
0002637 . 0
) ( ) (
2
*
= A (7)
In the above equations, k
e
, is the effective permeability of the fractured system as calculated from the slope of the buildup
curve. The vertical pressure separation (op) between the two parallel lines is related to e. Often the customary two straight
line segments do not appear on the pressure buildup curve because of well bore storage or boundary effects. The time
equation (Eq. 7) shows that the time frame for transition to occur in systems with moderate compressibility can be very short
(and thus is often masked by wellbore storage effects). Also nonuniform fracturing (variable o) complicates the transition
period. If the dualporosity response is not seen, shape factor will need to be inferred from other information (e.g. image logs)
and calibrated via DFN modeling.
In practice, the effective fracture intensity and degree of connection is difficult to predict. This may be because fractures are
often well connected on a large scale, but a few primary pathways may dominate flow and thus many of the fractures seen in
logs and cores do not interconnect on a large scale. Also, the apparent connectivity can depend on the well placement and
the recovery process. Thus, coarsegrid, dualporosity models for multiphase flow have historically been too well connected.
Secondly, the fracture network appears much more heterogeneous than first estimated and in fact the effective fracture
spacing can appear to be different for depletion versus waterflood. Under depletion, a nearby producer can deplete many of
the fractures, causing matrix fluid to expand through a large fracturematrix surface area. However, under waterflood, water
can imbibe only through the matrix surface that is contacted by water. Fractures that are not continuous along the water flow
path will not be flooded by injection water even if they are open and permeable. Recent industry activity has been directed at
using discrete stochastic distributions of fractures based on statistical and deterministic information gathered from logs, cores
and outcrops to improve understanding of the fracture connectivity. These stochastic realizations can provide a means to
estimate effective fracture network properties for upscaling into conventional dualporosity models.
Fluid Flow Behavior
In threephase systems, consisting of gas, oil, and water, different portions of a given reservoir can undergo different
6 CSUG/SPE 146580
displacement processes leading to complex flow behavior. This usually happens because of the differences in field operations
schemes (i.e., waterflooding vs. gas injection) and rock characteristics such as wettability, fracture intensity, permeability
variations, layering, structural effects, capillary pressure and relative permeability hysteresis. The oil recovery mechanisms
associated with the field characteristics as well as the production schemes in the field are best studied by laboratory
experiments and finegrid numerical simulation. The approximations made by the simulators for coarsegrid fullfield models
must be verified by these finegrid studies. In the following discussion, we present some of the basic ideas for flow behavior
under various recovery mechanisms. The idea is to illustrate the importance of various parameters on fracturematrix transfer.
Depletion
In purely singlephase depletion, the dominant recovery mechanism is fluid expansion. In practice, this occurs normally as
the first stage of reservoir operations, and fluid recoveries are typically quite low (except for gas systems). In singlephase
depletion, oil or gas recovery rate can be much greater in NFR compared to nonfractured reservoirs because the high
permeability fracture network undergoes rapid depletion and provides large surface area for reservoir fluid in the low
permeability matrix to expand into the fractures. This has been a means to obtain significant recovery from lowpermeability
unconventional reservoirs. In multiphase depletion, fracture flow can lead to free gas flow in the fractures, which, in turn,
can invoke gravity drainage of oil from the matrix. Efficient gravity drainage can lead to high oil recoveries if the gasoil
gravity drainage process is managed properly. The important parameters that affect the rate of matrixfracture fluid transfer
during depletion can be shown to be shape factor, o, matrix fluid mobility, (k
m
/), and fracturematrix pressure difference,
(p
m
p
f
) as can be seen from the equation for singlephase flow in the fractures. The second term in Eq. 8 represents the
matrixfracture transfer.
( )
t
p
c q p p
k
p
k
f
tf f f m f
m
f
f
c
c
= +


.

\

V  V 
(8)
Solution of the singlephase equations for low compressibility flow shows that a double exponential flow rate decline can
develop for a closed radial reservoir, producing under constant bottomhole pressure depletion,. Chen, et. al., (1986) showed
the exponential decline characteristics of a number of Austin Chalk wells.
Satman (1985) derived equations for the initial flow rate and for the slope of the exponential recovery lines as a function of
the interporosity parameter, . Because of the small volume of fluid in the fractures, the initial period may be very short and
therefore difficult to measure. For large (> 16 r
w
2
/r
e
2
) only a single straight line will develop because of the rapid pressure
equalization between the fracture and matrix. The reservoir then responds like an nonfractured reservoir with permeability
equal to k
e
and porositycompressibility equal to that of the total system. For highly fractured systems with moderate to good
matrix permeability and close fracture spacing (e.g. ok
m
> 0.1 mdft
2
), only one straight line will develop.
In theory, Satmans equations could be used with decline curve analysis to determine reservoir properties for the fracture
matrix system (Chen, et al. 1986). In practice this is very difficult because of variations in bottomhole pressures, multiphase
flow, offset well interference effects, complex well geometries/completions and the long production times required to obtain
both exponential decline periods. Also, numerical simulation has shown that the first line is very short lived and the slope is
affected by matrixfracture flow. Therefore, fracture properties are very difficult to determine from decline curve analysis.
However, longterm rate and pressure decline can be used with conventional reservoir depletion analysis to determine the
effective permeability and total porevolume of the system (fracture and matrix). The magnitude of effective permeability
compared to only the matrix (core) permeability gives an indication of the relative importance of fractures. This discussion
shows how high fracture intensity (large o) can lead to efficient depletion of fractured systems with low matrix permeability.
Water Imbibition
Water imbibition has proven to be an effective recovery mechanism in some NFR. Imbibition in reservoir rock is the process
in which water is drawn into the rock by the action of capillary forces. Mattax and Kyte (1962), through experimental
investigations, found that recovery in such systems could be scaled through the following dimensionless time:
t
L
k
t
m
m
D (


.

\

=
2

(9)
This equation shows that recovery is inversely proportional to the matrix block size squared, L
2
. Capillary pressure is
indirectly incorporated in the above equation through the interfacial tension, , permeability and porosity terms. The above
equation can be rewritten in terms of shape factor, o, by replacing the 1/L
2
term as shown in Kazemi, et. al. (1992) and
CSUG/SPE 146580 7
Zhang, et al. (1996). The experimental data generated by Mattax and Kyte can be described by an exponential time function.
Kazemi, Gilman and ElSharkawy (1992) provide an analytical solution for onedimensional flow in a fractured systems
using such exponential relations.
The important parameters affecting oil recovery from water imbibition are shape factor, capillary pressure and oil mobility,
k
m
k
ro
/
o
. Oil rate is greater as permeability increases and matrix block size decreases (fracture intensity or shape factor
increases). Recovery rate also increases as oil relative permeability increases making it easier for water to imbibe and
displace oil from the matrix. Higher capillary pressure increases rate of oil recovery through increased imbibition force. NFR
that have undergone waterflood or strong aquifer drive include the Fahud Field (ONeill, 1988), Ekofisk Field (Hallenbeck et
al., 1991), Midale Field (Beliveau, et al., 1993) Ezzaouia Field (Gilman et. al., 1996), and Ghawar Field (Phelps and Strauss,
2002). The more than 30% recovery from the Ezzaouia Field is a case that illustrates the effectiveness of water imbibition
dominated recovery in some NFR.
From a largescale simulation perspective, some of the issues regarding recovery prediction uncertainty include incomplete
coverage of the matrix block as a result of nonuniform sweep through the fractures, variations in recovery as a result of non
uniform block size in a grid block and uncertainty in laboratory measurements.
GasOil Gravity Drainage
In some highly fractured reservoirs, gravity drainage can be a very effective recovery mechanism. If matrix drainage is the
limiting factor, then oil drainage can be approximated as a onedimensional solution. In onedimensional gasoil gravity
drainage in a porous media, the time rate of change in elevation of a constant saturation oil shock front (that is, the gasoil
frontal velocity, ft/d) is given by the following equation (Dykstra, 1978; Richardson, 1989):
c
c

c
c
z
t
x k
k
S
S
v og
o
ro
o
S o
o
=
7 83 10
6
. A
(10)
Assuming powerlaw oil relative permeability function with an exponent of n and coefficient of k
rom
, Eq. 10 can be solved
analytically for the initial oil rate as given by Eq. 11. This initial rate (q
oi
, STB/D) is sustained for a period of t
qmax
days:
0
5
6146 . 5
10 4 . 4
A
=
A k k x
q
rom v
oi
(11)
rom v
p
q
nk k x
h
t

A
=
5
0
max
10 4 . 4
(12)
For times greater than t
qmax
, analytical solution gives Eq. 13. This defines a hyperbolic decline equation given by the second
two equations. The hyperbolic b factor is equal to (n1)/n. The subscript p on porosity refers to the effective oil porosity
(1S
wir
S
gc
S
orw
).
( )
) 1 /( ) 1 /( 1 5
) 1 /( ) 1 /( 1
0
) 10 4 . 4 ( 6146 . 5
A
=
n n n
rom v
n n n
p p
o
t nk k x n
h A
q
 
(13)


.

\

=
t Ah
nq
b
D
p
oi
1 6146 . 5 1

(14)
( )
b
oi o
bDt q q
/ 1
1
+ = (15)
The important parameters affecting gravity drainage rate are vertical permeability, density difference, and oil mobility.
Fractured systems in which gas filled vertical fractures surround a vertically continuous matrix can be represented by this
equation. The above equation ignores capillary pressure. Capillary pressure has minimal effect on early rates, but can cause
holdup of oil above barriers or horizontal fractures. For lab centrifuge data in which the core is spun at a rate scaled to the
8 CSUG/SPE 146580
field oil column height, the resulting relative permeability will include the capillary holdup and can be used directly in the
above equations for scaleup. Centrifuge experiments (Kyte, 1970) are a common method of assessing effectiveness of
gravity drainage which can be directly upscaled to field behavior.
Permeability variations can affect early drainage rate, but a harmonic average permeability can give a reasonable
approximation to the drainage behavior over a limited permeability range. Simulation shows that for matrix block radii of
10s of ft or less, there is no difference in recovery, showing the insensitivity of this system to areal fracture spacing. Because
the flow is controlled by matrix vertical permeability, the areal spacing of the fractures has minimal impact for relative close
fracture spacing (<10 ft). Fracture intensity will however, significantly affect the rates and coning tendency of producing
wells with regard to the fracture network as well as controlling the ability to capture draining oil at wells. Adding vertical
barriers across a matrix column would appear to initially increase drainage rate, because each isolated matrix block can feed
into the fractures rather flowing vertically through the entire matrix column. However, this draining oil will resaturate lower
matrix blocks as drainage occurs, and thus the system with barriers will drain more slowly. Also, capillary holdup above the
barriers will cause longterm recovery to be reduced.
A commonly used finitedifference approximation to calculate the gravity drainage rate, in STB/D, for the dualporosity
idealization and negligible capillary pressure is based on the concept of vertical equilibrium as given by:
144
) ( 001127 . 0
og
cogth of om
o o
ro zm go
o
p h h
B
k Vk
o
t
A
= (16)
Where h
om
is the product of the mobile oil saturation in a given matrix grid cell and the matrix block height, L
z
. Similarly, h
of
is obtained by multiplying the mobile oil saturation in the given fracture cell and L
z
. The shape factor for gasoil gravity
drainage, o
go
, needs to reflect the fact that the process of drainage is primarily vertical. It is given by the following equation:
o
go
z
L
=
2
2
(17)
Because the oil height terms are proportional to L
z
and o
go
is inversely proportional to L
z
2
, the drainage rate becomes
inversely proportional to L
z
as with the analytical equation. L
z
is effectively the distance between major flow barriers. This
form generally over estimates the rate of gravity drainage because of the assumption of vertical equilibrium in the matrix.
A more realistic alternative to conventional dualporosity, pseudogravity simulation is to use fine grids and dual
permeability simulation only in the vertical direction (Gilman and Kazemi, 1983). This will provide for more accurate
simulation of the gravity drainage process. Dualpermeability in the vertical direction means that all matrix blocks are
allowed to be continuous in the vertical direction. This will allow oil to drain vertically in the matrix as well as laterally into
the fractures. Of course, there is added computational expense because of the dualpermeability approach and the requirement
of additional layers.
In gravitydominated recovery from NFR, fractures are used as gas flow conduits to invoke gravity drainage from the matrix
by controllably lowering the gasoil contact in the fractures. This procedure is the basis for the concept of contact
management in fractured reservoirs, which can lead to one of the most effective improvedoilrecovery techniques for
fractured reservoirs. NFR that have undergone gasoil gravity drainage include the Fahud Field (ONeill, 1988), the Yates
Field (Rothkopf and Wadleigh, 1994; Campanella, et. al., 2000), the Haft Kel Field (Saidi, 1996), and the Cantarell Field
(Arevalo, et al., 1996). In the Haft Kel Field (Saidi, 1996), the calculated water displacement efficiency, supported by field
measurements was only about 17%, whereas the calculated gasoil drainage efficiency, at a reservoir pressure of 1512 psi,
was about 32%.
Other Modeling Considerations
The multiphase flow equations require the use of relative permeability in both the matrix and fracture. Relative permeability
in the matrix is the same as obtained for conventional singleporosity cores (although there is an upscaling issue because the
flow rates are a function of the average saturation in large grid blocks). It is often assumed that fracture relative permeability
is a linear function of phase saturations. Laboratory experiments have shown this to be true for large fracture sizes (> 50 m).
However, for smaller fractures it has been demonstrated (Maloney, et al., 1997) that relative permeability in fractures is non
linear and depends on the fracture flow velocity, direction of flow, and density difference. Other fracture relative
permeability concepts have also been presented by a number of authors (e.g. MacDonald, et al., 1991; Kazemi and Gilman,
1993). There have also been a number of different methods proposed to handle the effect of relative permeability for matrix
fracture flow, although the most general method is to use conventional upstream weighting. Historical performance also
CSUG/SPE 146580 9
suggests that relative permeability in the fractures is not a linear function of phase saturations. This is because the relative
permeability for a fracture network is an average of a number of discrete fractures.
Previous discussions referred to depletion. water imbibition and gravity drainage, as the most common methods for
enhancing recovery in NFR; however, tertiary recovery, or enhanced oil recovery (EOR), are processes that produce further
additional oil economically over that which can be produced from primary and secondary recovery methods. In NFR, EOR
can be viewed as the methods that accelerate oil recovery by altering reservoir fluid and rock properties to better utilize the
reservoirs natural energy. The most promising EOR techniques in NFR (Christiansen, et al., 1989) include CO
2
(Beliveau
and Payne, 1993; Malik and Islam, 2000), heat (Reis and Miller, 1995), surfactants (Chen, et al., 2001), and polymers
(Sydansk and Moore, 1992). For some of these processes we must also be concerned about diffusion rates and /or heat
transfer rates between matrix and fractures. The effectiveness of these two transport phenomena is also strongly dependent on
fracturematrix surface area (shape factor).
A New Classification Plot
In this section we present our new classification plot (referred to here as the Gilman Plot) to highlight the relative
importance of fracturematrix transfer. This consideration is not directly incorporated in the Nelson plot (Figure 2). Effective
matrixfracture transfer provides a means for microdarcy rock to provide economic recovery of oil in unconventional plays.
This new classification concept is intended to be used to understand and illustrate differences in dynamic performance of
NFR for purposes such as analogue selection for decline forecasts, screening of potential recovery processes or making
decisions on modeling methodology. In this paper, several characterized NFR reservoirs (including unconventional systems)
are compared via the new classification plot and their differences and performance are discussed. This plot is certainly not all
encompassing with regard to fractured reservoir performance but it highlights the fact that matrixfracture surface area is a
major aspect of understanding performance of NFR.
The first ratio we will use in the plot is the storativity ratio; as defined by Warren and Root (1963). However because we
often dont know the fracture compressibility, we will remove compressibility (c
t
) from this equation.
m f
f
 

e

+
= (18)
The second ratio is the ratio of effective fracture permeability to matrix permeability; often referred to as excess permeability
ratio. However, as discussed earlier, the ratio does not need to be greater than one. For fracture permeability we generally use
a geometric average of the horizontal permeability diagonal tensor terms.
m
fe
exr
k
k
k = (19)
These two parameters are the basis of the Nelson plot and will remain the primary axes in our plot. These are relative easy to
comprehend in terms of magnitude as permeability and porosity are terms that we are all familiar with.
The previous discussion of depletion systems showed how Satman (1985) illustrated that the product of r
e
2
/r
w
2
is a
controlling factor in the longterm behavior of a radial, closed, singlephase, dualporosity system. He showed, for example,
that if this term is greater than 16, the system (under pseudosteady state singlephase depletion) will behave like a single
porosity system with permeability k
e
and porosity 
f
+
m
. Therefore, for use in our classification plot, a third ratio is a
redefined interporosity flow term,
A
, which indicates the relative contribution of fracturematrix flow compared to flow
through the fractures within the well drainage area, A (Eq. 20). The term is dimensionless as long as consistent units are
used. In this equation, there are different possible ways to define the well drainage area (which will change with time and
with well density). Also the drainage area is complicated for horizontal wells; however, a consistent method from fieldto
field is adequate for classification purposes. In our examples we use a uniform well spacing.
(20)
The magnitude of this interporosity flow term is not intuitive, but in a simple sense, it is a measure of the ability of the
matrix to transfer fluids to the fractures, relative to the ability of the fractures to transport fluids to the well over a drainage
volume. If we consider Satmans analysis, a value of
A
> 50 would suggest that matrixfracture transfer is rapid enough to
fe
m
A
k
A k o
=
10 CSUG/SPE 146580
keep up with depletion in the fractures throughout the drainage area. This is true only for singlephase depletion after pseudo
steady state is reached, but it does give a magnitude for relative comparison. Larger values mean that the matrixfracture
transfer can more likely keep up with fracture depletion. If we have competing mass transfer processes (e.g. pressure
depletion in fractures and diffusion from matrix to fractures), then the ratio may need to include other dimensionless
parameters such as a capillary or bond number to better estimate the effectiveness of matrixfracture transfer relative to
fracture transport.
We now present a few examples of the new plot. In the plot examples here, we suggest that the interporosity parameters data
be broken down into a number of equalized bins (e.g. ranges of interporosity flow ratio). For our examples we used 10%
probability ranges. This is similar to the idea for carbonate classification (Lucia, 1999), where the phik data is divided into
different flow indicator types. All of the examples presented here are from actual NFR with 1) rigorously defined matrix
reservoir characterization, 2) fractures networks defined via DFN and/or CFM methods calibrated to seismic, logs and/or
matrix characteristics, and 3) calibrated to dynamic performance. The comparisons are all made assuming a 160acre well
spacing to provide consistent interporosity flow ratio. The results shown are only for a limited subset of the main reservoir
around some of the most highly fractured areas. Figure 3 compares the first four examples.
Example 1: Carbonate, Light Oil, Pressure Depletion via Horizontal and Vertical Wells, Moderate Aquifer Influx
This reservoir has moderate matrix permeability (<10md) but is a highly productive oil reservoir as a result of natural
fracturing. Fractures were characterized via image logs to define mechanical units and controls on fracture intensity. Seismic
attributes such as distance to faults, brittleness, semblance, curvature and anisotropy were combined to provide either direct
correlation to P32 and thus permeability and porosity or calibrated via DFN models to match well P10 and account for
variability in fracture height, length orientation and transmissivity. The models were calibrated to continuous measures of
pressure, oil rate and water. The plot shows large variability in fracturematrix porosity and permeability ratios as a result of
high variability in fracture intensity. Fracturetomatrix permeability can vary by more than 3 orders of magnitude, typical of
many conventional fractured reservoirs. Fracturing here is strongly related to fault proximity. Where fracture permeability is
high we also see high fracture intensity and thus the interporosity flow ratio also shows a large range of values. Such a
system might require a dualpermeability representation to account for flow through the matrix in less fractured areas.
Example 2: Carbonate, Light Oil, Pressure Maintenance via Gas and Water Injection in Horizontal Wells
This is a low permeability carbonate (several md). Fracturing here is also strongly related to fault proximity; however, there
is significant vertical variation via stratigraphic mechanical units as well as structural relationships. Horizontal wells provide
high early productivity when intersecting open fractures. Water injection has provided significant recovery especially in the
less fractured peripheral areas. Gas injection in the crest provides additional oil recovery viagasoil gravity drainage.
Although excess fracture permeability is not as high as in Case 1, we still see 3 orders of magnitude variation, which has
resulted in high variability in well performance and early breakthrough of injection water and gas. This reservoir has been
modeled via dualpermeability approaches.
Example 3: Sandstone, Heavy Oil, Waterflood via Vertical Wells
This third example also shows high variability in fracture excess permeability. Matrix permeability for this sandstone
reservoir is quite good (~100 md), but oil viscosity is high resulting in similar oil mobility to the previous two cases. This is
an older field that is primarily developed on vertical wells and thus productivity is lower than the horizontal well cases. High
dip has allowed for economic oil recovery via down dip water injection. Early water breakthrough and high water cuts are
common. Again, excess permeability varies by three orders of magnitude a common theme with conventional naturally
fractured reservoirs. This reservoir has been modeled via dualpermeability approaches in order to account for slow matrix
fracture water transfer.
Example 4: Quartzite Sand, Low Matrix Permeability, Gas Reservoir with Vertical and Horizontal Wells
This reservoir could be considered a tight gas reservoir (microdarcy perm) if it was not for the extensive fracture system. The
fracture network leads to high potential rates in both vertical and horizontal wells. The excess permeability does not appear
to be as heterogeneous as the previous examples, but this could be because the system has only undergone depletion. The true
measure of heterogeneity often reveals itself during displacement processes. The high fracture intensity and singlephase
depletion means that the system essentially acts like a high permeability singleporosity system although it has been modeled
as a dualporosity reservoir.
Example 5: Unconventional, Light Oil, Pressure Depletion via Hydraulically Fractured Horizontal Wells
This is an unconventional oil reservoir developed via hydraulically fractured horizontal wells. The industry is only beginning
to characterize these reservoirs. We believe that many of the same conventional characterization methods can be applied to
unconventional reservoirs. This includes calibration to image logs via DFN; correlation to seismic attributes and calibration
of these attributes to well performance. Such systems can show extreme variability over a play, but looking within local
sweet spots, we see the type of characteristics shown in Figure 4(A): lower range in excess permeability. Surprisingly, the
CSUG/SPE 146580 11
interporosity flow values are quite high, because of the large matrixfracture surface area. Dualporosity or dualpermeability
models are required to match early pressure depletion in the nearwell shattered rock volume, but later time declines are
indicative of a total system response with effective permeability one to two orders of magnitude greater than matrix
permeability. Of course, this is for one specific example only and is not a general rule.
Example 6: Unconventional Light Oil/Condensate with Free Gas Phase, Pressure Depletion via Hydraulically
Fractured Horizontal Wells (singlewell characterization)
A recent singlewell calibration in a different unconventional play is also shown in Figure 4(B). At first glance it looks to
have similar characteristics to the previous plot. However, here productivities are quite low. This is the difficulty in simply
using dimensionless parameters for comparison. There are many other factors besides excess permeability which dictate the
well performance. Other concerns include PVT characteristics, initial pressure, thickness, number of frac stages, and
horizontal well length. Thus such comparisons as provided by these dimensionless plots are only one aspect of an overall
integrated characterization and simulation. The dimensionless plots are most useful when trying to make development
decisions such as well spacing within one specific unconventional play.
Example 7: Synthetic Example of Conventional Reservoir with Strong FaultRelated Fracturing
This final short example shows the importance of shape factor (or fracture intensity). This is a synthetic example based on
concepts from the previous examples. It has moderate API oil and limited aquifer drive. It has a strong fault related fracturing
along several large faults. The Gilman plot is shown in Figure 5(A). The band running through the plot is a result of several
more highly fractured zones. The excess permeability ratio is generally higher than the other cases. A 10year, multiwell
simulation of depletion is shown in Figure 5(B). All cases have the same properties except that shape factor is increased by a
factor of 10 or decreased by a factor of 10 compared to the base case relations. The main differences in decline are a result of
differences in water production. Of note here is the large longterm difference in oil production; however, the first 3years
show very minimal differences. This is typical of the types of responses we see in NFR. The short term impact of fracture
spacing differences are difficult to determine from production response, but have a significant impact on longterm behavior.
Differences are seen more quickly in injection processes.
The final comparison of the average properties for the different cases is shown in Figure 6. Table 1 provides the data. These
figures are a plot of the average values from Table 1. Averages themselves are not sufficient for estimating reservoir
response; however, they do more clearly show the relative differences between different reservoirs. There is generally an
increasing trend of excess permeability with increasing storage ratio as expected. Case 4 is an outlier because of the very low
matrix porosity. Figure 6(B) shows lower interporosity flow ratio as excess permeability increases. This trend is not strong
because as excess permeability increases, matrix shape factor may also increase. For all these cases, the average inter
porosity flow parameter is greater than the value of 50 that might indicate rapid matrixfracture equilibration for singlephase
flow. However, as shown in Figure 6(B), the longterm behavior is still very sensitive to the value of
A
. This is because of
limited fluid transfer rate for capillary and gravity dominated flow relative to transport through the fracture network.
Summary
Performance evaluation, laboratory studies and modeling of fields with long histories has provided much information about
the nature of the fractured networks, fluid transfer mechanisms and indications of how the effective network properties relate
to directly measurable properties like image and production logs, pressure transients and tracer tests. However, the
characterization and modeling process still requires extensive calibration with dynamic data. Therefore we still must often
rely on analogue reservoirs and prior experiences to estimate the flow characteristics and future behavior in NFR for fields
with less production history. Much emphasis in modern workflows is being placed on improvements in characterization of
the fracture network; however, reservoir flow behavior is also very dependent on the effectiveness of matrixfracture transfer
mechanisms.
As a general recommendation for NFR simulation, is it important to identify intrinsic (e.g. lithologic) and extrinsic (e.g.
structural) controls on fracture distribution and intensity and relate these to well performance (e.g. productivity and water
breakthrough) in order to develop algorithms for distributing the fracture network properties in a 3D static geomodel. The
simulation model must honor the variability and should not be an overly simplified model in order to capture the
heterogeneous nature of the reservoirs. Fracturematrix transfer must be studied via laboratory experimentation, finegrid
modeling and analogous reservoirs.
The new classification plot illustrated here provides a means to compare characteristics of different reservoirs including the
very important aspect of matrixfracture transfer. Interporosity fluid transfer is not directly incorporated in the Nelson plot.
Our new classification concept can be used to understand differences in dynamic performance of NFR for purposes of
analogue selection or screening of potential recovery processes. We compared several characterized NFR reservoirs
(including unconventional systems) via the new classification plot and their differences and performance were summarized.
12 CSUG/SPE 146580
Acknowledgements
As mentioned in the Introduction, many of the background concepts were a result of unpublished collaborative efforts with
Dr. H. Kazemi, Colorado School of Mines. We would like to acknowledge his many contributions to the ideas presented
here. We would also like to acknowledge Ron Nelson, PhD, Broken N Consulting for providing a modified schematic of his
wellknown distribution of fracture reservoir types.
Nomenclature
A area, ft
2
B formation volume factor (FVF), RB/STB
b harmonic decline factor
c compressibility, psia
1
d distance, ft
D depth, ft
D decline factor, 1/day
h height or thickness, ft
k permeability, md
L matrix block dimension, ft
ln natural logarithm
n exponent, powerlaw equation
p pressure, psia
P10 linear fracture intensity measure, 1/L
P32 volumetric fracture intensity factor, 1/L
P33 volume of fractures per unit volume
q flow rate, STB/D
r radius, ft
S saturation, fraction
t time, days
V volume, ft
3
w width, ft
x distance in xdirection, ft
y distance in ydirection, ft
z distance in zdirection, ft
A difference or change in values
 porosity, fraction
interfacial tension, dyne/cm
o difference in value
interporosity flow parameter, dimensionless
mobility, md/cp
viscosity, cp
t Pi, constant
density, lb/ft
3
o shape factor, ft
2
t fracturematrix flow rate, STB/D
e storativity ratio, dimensionless
c partial derivative
V gradient operator
V divergence operator for a vector
Subscripts
A area
c capillary
D dimensionless
e effective, external
exr excess ratio
f fracture
g gas
i initial
CSUG/SPE 146580 13
i,j,k indices
m matrix
o oil
p effective porosity
qmax maximum rate
r relative
t total
th threshold
v vertical
w water
x,y,z directions
References
Araujo, H., Gilman, J. R., and Kazemi, H.: "Analysis of Interference and Pulse Tests in Heterogeneous Naturally Fractured Reservoirs",
SPE 49234 presented at the 1998 SPE Annual Technical Conference and Exhibition, New Orleans, LA (Sep. 2730, 1998).
ArevaloV., J. A., SamaniegoV., F., LopezC., F. F., UrquietaS., E.: On the Exploitation Conditions of the Akal Reservoir Considering
Gas Cap Nitrogen Injection SPE 35319, Intl. Petroleum Conference & Exhibition of Mexico, Villahermosa, Mexico, (March 57.
1996).
Ata, E. and Michelena, R. J.: Mapping distribution of fractures in a reservoir with PS converted waves: The Leading Edge, 14, p.664673
(1995)
Beliveau, D., Payne, D. A. and Mundry, M.: Waterflood and CO2 flood of the Fractured Midale Field, J. Pet. Tech, 881887,
(September, 1993).
Campanella, J. D., Wadleigh, E. E., and Gilman, J. R.: Flow CharacterizationCritical for Efficiency of Field Operations and IOR, SPE
58996 presented at the 2000 SPE International Petroleum Conference and Exhibition in Mexico, Villahermosa, Mexico (Feb. 13,
2000).
Chang, MM: Deriving the Shape Factor of a Fractured Rock Matrix, U.S. Department of Energy (National Institute for Petroleum and
Energy Research, Bartlesville, OK, NIPER696 (DE93000170) (Sept. 1993).
Chen, H. Y., Poston, S. W., and Wu, C. H.: "Characterization of the Austin Chalk Producing Trend," SPE 15533 presented at the 61st SPE
Annual Technical Conference, New Orleans, LA (Oct. 1986).
Chen, H. L., Lucas, L. R., Nogaret, L. A. D., Yang, H. D., and Kenyon, D. E.: Laboratory Monitoring of Surfactant Imbibition With
Computerized Tomography, SPE Res. Engr. and Eval., (February 2001).
Christiansen, R. L., Gilman, J. R., Jargon, J. R., Kazemi, H., and Smith, R. E.: "Enhanced Oil Recovery in Naturally Fractured Reservoirs,"
AIChE 51d presented at the 1989 AIChE Summer National Meeting Philadelphia, PA (August 1989).
Dershowitz, W., LaPointe, P., Eiben, T., and Wei, L.: Integration of Discrete Fracture Network Methods with Conventional Simulator
Approaches, SPE Res. Eval. & Eng., 165170, (April 2000).
Dershowitz, W., Hermanson, J., Follin, S. and Mauldon, M.: "Fracture intensity measures in 1D, 2D, and 3D and sp, Sweden,Pacific
Rocks 2000, Girard, Liebman, Breeds & Doe (eds) c 2000 Balkema, Rotterdam, ISBN 90 5809 155 4. pg 849853.
Dykstra, H.: "The Prediction of Oil Recovery by Gravity Drainage", J. Pet. Tech. 818830, (May 1978).
Gale, Julia F. W.: Specifying Lengths of Horizontal Wells in Fractured Reservoirs, SPE Res. Eval. & Eng., 266272, (June 2002).
Gauthier, B. D. M., Garcia, M., and Daniel J. M.: Integrated Fractured Reservoir Characterization: A Case Study in a North Africa
Field, SPE Res. Eval. and Engr., 284294 (Aug. 2002).
Gilman, J. R. and Kazemi, H.: "Improvements in Simulation of Naturally Fractured Reservoirs," Soc. Pet. Engr. J. 695707 (Aug. 1983).
Gilman, J. R. and Kazemi, H.: "Improved Calculations for Viscous and Gravity Displacement in Matrix Blocks in DualPorosity
Simulators," J. Pet. Tech, 6070 (Jan. 1988).
Gilman, J. R., Tiss, M. Jargon, J. R., and Hinchman, S. B.: DualPorosity Simulation of the Zebbag Reservoir, Ezzaouia Field, Presented
at the Second ECLIPSE International Forum, Houston, TX, (April 1519, 1996).
Gilman, J. R.: "Practical Aspects of Simulation of Fractured Reservoirs ", presented at the Seventh International Forum on Reservoir
Simulation, Schlosshotel Bhlerhhe, Bhl/BadenBaden, Germany (June 23rd  27th, 2003).
14 CSUG/SPE 146580
Hallenbeck, L.D., Sylte, J.E., Ebbs, D.J., Thomas, L.K.: Implementation of the Ekofisk Field Waterflood SPE Form. Eval., , p. 284,
(1991).
Haws, G. W. and Hurley, N. F.: "Applications of PressureInterference Data in Reservoir Characterization Studies, Big Horn Basin,
Wyoming", SPE 24668 presented at the 67th SPE Annual Technical Conference and Exhibition, Washington, DC, (Oct. 47, 1992).
Heinemann, Z.E. and Mittermeir, G.M. Rigorous Derivation of the KazemiGilmanElsharkawy Generalized Dual Porosity Shape Factor
Paper B044 presented at the 10th European Conference on the Mathematics of Oil Recovery  Amsterdam, The Netherlands, (Sept. 4
7, 2006)
Hossain, M. M., Rahman, M. K. and Rahman, S. S.: A Shear Dilation Stimulation Model for Production Enhancement from Naturally
Fractured Reservoirs, SPE Journal, 183195, (June 2002).
Iwere, F.O., Moreno, J. E., Apaydin, O. G., Len Ventura, R., Garcia, J. L. Vug Characterization and Pore volume Compressibility for
Numerical Simulation of Vuggy and Fractured Carbonate Reservoirs SPE 74341, SPE International Petroleum Conference and
Exhibition, Villahermosa, Mexico, (Feb. 2002).
Kazemi, H., Merrill, L. S., Porterfield, K. L., and Zeman, P. R.: Numerical simulation of WaterOil flow in Naturally Fractured Reservoirs,
Soc. Pet. Eng. J. 31726, (Dec. 1976).
Kazemi, H., Gilman, J. R., and ElSharkawy, A. M.: "Analytical and Numerical Solution of Oil Recovery from Fractured Reservoirs Using
Empirical Transfer Functions," Soc. Pet. Eng. Res. Engr., 219227, (May 1992).
Kazemi, H. and Gilman, J. R.: "Chapter 6. Multiphase Flow in Fractured Petroleum Reservoirs", in Flow and Contaminant Transport in
Fractured Rock, Edited by J. Bear, CF. Tsang, and G. de Marsily, Academic Press, San Diego, CA, 267323 (1993).
Kazemi, H. and Shinta, A. A.: Determining Orientation and Conductivity of High Permeability Channels in Naturally Fractured
Reservoirs, Reservoir Characterization IV, Edited by Bill Linville, Pennwell Books, (1993).
Kyte, J. R.: A Centrifuge Method To Predict MatrixBlock Recovery in Fractured Reservoirs, Soc. Pet. Eng. J., p. 164, (1970)
Lake, Larry W.: Enhanced Oil Recovery, Prentice Hall, p. 18, (1989).
Lim, K. T. and Aziz, K.: MatrixFracture Transfer Shape Factor for DualPorosity Simulation, J. of Petroleum Science and Engineering,
169178, (1995).
Lucia, F. Jerry, Carbonate Reservoir Characterization, SpringerVerlag Berlin Heidelberg, (1999)
MacDonald, A. E., Beckner, B. L., Chan, H. M., Jones, T. A. and Wooten, S. A.: Some Important Considerations in the Simulation of
Naturally Fractured Reservoirs, SPE 21814 presented at the Rocky Mountain Regional Meeting and LowPermeability Reservoir
Symposium, Denver, CO (April 1517, 1991).
Malik, Q. M., and Islam, M. R.: CO2 Injection in the Weyburn Field of Canada: Optimization of Enhanced Oil Recovery and Greenhouse
Gas Storage With Horizontal Wells SPE 59327, SPE/DOE Improved Oil Recovery Symposium, Tulsa, Oklahoma, (April 35, 2000)
Maloney, D. and Doggett, K.: Multiphase Flow in Fractures DOE Contract DEAC2294PC91008, Report SCA 9730, (1977).
Mattax, C. C., and Kyte, J. R.: "Imbibition Oil Recovery from Fractured, WaterDrive Reservoirs", Soc. Pet. Eng. J., 177184, (June 1962).
Nelson, R. A.: Geologic Analysis of Naturally Fractured Reservoirs, Gulf Professional Publishing Co., (2
nd
Ed. 2001).
Oda, M.: Permeability Tensor for Discontinuous Rock Masses, Geotechniques, V. 35, 483495, (1985).
O'Neill, Niel: Fahud Field Review: A Switch from Water to Gas Injection J. Pet. Tech, p. 609, (1988).
Pereira, C, Kazemi, H., Ozkan, E.: Combined Effect of NonDarcy Flow and Formation Damage on Gas Well Performance In Dual
Porosity/Dual Permeability Reservoirs SPE 90623 presented at SPE ATCE, September 28. (2004)
Penuela, G., Civan, F., Hughes, R. G. and Wiggins, M.L.: TimeDependent Shape Factor for Interporosity Flow in Naturally Fractured
GasCondensate Reservoirs, SPE 75524, SPE Gas Technology Symposium, Calgary, (April 30May 2, 2002).
Phelps, R. E. and Strauss, J. P.: Capturing Reservoir Behavior by Simulating Vertical Fracture and SuperK Zones in the Ghawar Field,
SPE Res. Eval. & Eng, 333340, (Aug. 2002).
CSUG/SPE 146580 15
Reis, J. C., Miller, M. A.: Oil Recovery from Naturally Fractured Reservoirs by Steam Injection Methods, DOE Contract DOE/BC
1466110, DE95000140, (May 1995).
Richardson, J. G., Sangree, J. B., and Sneider, R. M.: "Oil Recovery by Gravity Segregation", J. Pet. Tech., 581582, (June 1989).
Rothkopf, B., and Wadleigh, E.: Retooling Improves Field Efficiency, The American Oil and Gas Reporter, 9195, (Sept., 1994).
Saidi, Ali M.: Twenty Years of Gas Injection History into WellFractured Haft Kel Field (Iran) SPE 35309, Intl. Petroleum Conference
& Exhibition of Mexico, Villahermosa, Mexico, (March 57. 1996).
Sarma, P., Aziz, K.: New Transfer Functions for Simulation of Naturally Fractured Reservoirs with DualPorosity Models paper SPE
980231 presented at the SPE Annual Technical Conference and Exhibition, Houston, Texas, (2629 September, 2003).
Satman, A.: "Decline Curve Analysis for Naturally Fractured Reservoirs: A Comparison of Models" unsolicited paper SPE 14473, (April
19, 1985).
Sydansk, R. D. and Moore, P. E.: Gel Conformance Treatments Increase Oil Production in Wyoming, Oil and Gas J. V. 90, N. 3, 4045,
(Jan. 20, 1992).
Wang, X.: "Stereological Interpretation of Rock Fracture Traces on Borehole Walls and Other Cylindrical Surfaces" PhD Dissertation,
Virginia Polytechnic Institute and State University September 16, 2005, Blacksburg, Virginia
Warren, J. E. and Root, P. J.: "The Behavior of Naturally Fractured Reservoirs," Soc. Pet. Eng. J. 245255, (Sept. 1963).
Zhang, X., Morrow, N. R., and Ma, S.: Experimental Verification of a Modified Scaling Group for Spontaneous Imbibition, SPE Res.
Eng., V. 11, 280285, (1996).
Zimmerman, Chen, Hadgu and Bodvarsson: A Numerical DualPorosity Model with SemiAnalytical Treatment of Fracture/ Matrix
Flow, Water Resources Research, V. 29, N. 7, 21272137, (July 1993).
Table 1. Average properties for all cases
Case kfe f
km m o, 1/ft
2
L=
(8/o)
0.5
ft
f/(f+m) kfe/km kmo, md/ft
2
A=
(Akmo)/kfe
+
Case 1 1.97E+01 0.0004 2.53E+00 0.1149 0.0152 40.4 0.0033 29.73 3.52E02 13653.30
Case 2 6.59E+01 0.0038 3.51E+00 0.2364 0.0284 21.4 0.0177 49.70 8.56E02 10541.45
Case 3 4.64E+02 0.0054 5.48E+01 0.1250 0.0407 16.2 0.0456 66.60 2.19E+00 33423.98
Case 4 1.99E+00 0.0037 1.16E02 0.0280 0.0050 92.5 0.0872 411.51 3.00E05 202.67
Case 5 1.08E01 0.0033 5.06E03 0.0909 0.0013 82.4 0.0362 29.26 6.95E06 413.15
Case 6 4.01E04 0.0010 3.52E05 0.0770 0.0032 50.0 0.0131 14.04 1.13E07 1957.22
Case 7 5.26E+01 0.0058 1.38E+00 0.1449 0.0070 47.5 0.0387 926.13 6.75E03 1276.83
+A=160acres for all cases
16 CSUG/SPE 146580
Figure 1. Idealization of a fractured system
with fracture spacing Lx and fracture width wf
Figure 3. Gilman Plot for conventional reservoirs, Cases 1 to 4. A) Carbonate, light oil, pressure depletion via horizontal and vertical
wells, moderate aquifer influx, B)Carbonate, light oil, pressure maintenance via gas and water injection in horizontal wells, C)
Sandstone, heavy oil, waterflood via vertical wells; D)Quartzite, low matrix permeability, gas reservoir with vertical and horizontal
wells
L
x
L
x
w
f
Gilman Plot
1.00
10.00
100.00
1000.00
0.0000 0.0200 0.0400 0.0600 0.0800 0.1000
f
f
/(f
f
+f
m
)
k
f
e
/
k
m
>0. to 1478.
>1478. to 4441.
>4441. to 9510.
>9510. to 15581.
>15581. to 22050.
>22050. to 29327.
>29327. to 39029.
>39029. to 51233.
>51233. to 74122.
>74122. to 229912.
lA=(Adkms)/kfe
Intellectual Property of iReservoir.com, Inc
0.10
1.00
10.00
100.00
1000.00
0.0000 0.0100 0.0200 0.0300 0.0400
k
f
e
/
k
m
f
f
/(f
f
+f
m
)
Gilman Plot
>0. to 2163.
>2163. to 5056.
>5056. to 28388.
>28388. to 68730.
>68730. to 135408.
>135408. to 213857.
>213857. to 341834.
>341834. to 622681.
>622681. to 6452860.
>6452860. to 793982524.
lA=(Adkms)/kfe
Intellectual Property of iReservoir.com, Inc
Figure 2. Modified schematic distribution of fracture
reservoir types IIII relative to reservoir porosity and
permeability. Courtesy Ron Nelson, PhD, Broken N
Consulting (after Nelson, 2001).
1.00
10.00
100.00
1000.00
0.0000 0.0200 0.0400 0.0600 0.0800 0.1000
k
f
e
/
k
m
f
f
/(f
f
+f
m
)
Gilman Plot
>0. to 16.
>16. to 19.
>19. to 22.
>22. to 25.
>25. to 30.
>30. to 37.
>37. to 47.
>47. to 58.
>58. to 91.
>91. to 33739.
lA=(Adkms)/kfe
Intellectual Property of iReservoir.com, Inc
0.10
1.00
10.00
100.00
1000.00
0.0000 0.0050 0.0100 0.0150 0.0200
k
f
e
/
k
m
f
f
/(f
f
+f
m
)
Gilman Plot
>0. to 735.
>735. to 1349.
>1349. to 2389.
>2389. to 4118.
>4118. to 6993.
>6993. to 11847.
>11847. to 20576.
>20576. to 39851.
>39851. to 90546.
>90546. to 872404.
l
A
=(A
d
k
m
s)/k
fe
Intellectual Property of iReservoir.com, Inc
I
Type II
III
%of Total Porosity
%
o
f
T
o
t
a
l
P
e
r
m
e
a
b
i
l
i
t
y
100 % k
f
100% k
m
100% f
m
100% f
f
M
All Fractures
Little or No Fractures
A B
C
D
CSUG/SPE 146580 17
Figure 4. Gilman Plot for unconventional oil reservoirs, Cases 5 and 6: A) Unconventional, light oil, pressure depletion via
hydraulically fractured horizontal wells; B) Unconventional, light oil/condensate with free gas phase, pressure depletion via
hydraulically fractured horizontal wells (singlewell characterization)
Figure 5. A) Gilman Plot for synthetic example and B) Normalized production for 10yr forecast with shape factor increased by a
factor of 10 or decreased by a factor of 10 compared to base simulation
Figure 6. Comparison of A) average values for all cases in Nelson Plot and B) plot of interporosity flow ratios
Gilman Plot
1.00
10.00
100.00
1000.00
0.0000 0.0200 0.0400 0.0600 0.0800 0.1000
f
f/(f
f+f
m)
k
f
e
/
k
m
>0. to 152.
>152. to 202.
>202. to 248.
>248. to 296.
>296. to 358.
>358. to 449.
>449. to 575.
>575. to 755.
>755. to 1007.
>1007. to 4969.
lA=(Adkms)/kfe
Intellectual Property of iReservoir.com, Inc
1.00
10.00
100.00
1000.00
0.000 0.005 0.010 0.015 0.020
k
f
e
/
k
m
f
f
/(f
f
+f
m
)
Gilman Plot
>0. to 1001.
>1001. to 1112.
>1112. to 1224.
>1224. to 1390.
>1390. to 1613.
>1613. to 1891.
>1891. to 2225.
>2225. to 2781.
>2781. to 3615.
>3615. to 4283.
l
A
=(A
d
k
m
s)/k
fe
Intellectual Property of iReservoir.com, Inc
1.00
10.00
100.00
1000.00
10000.00
0.0000 0.0200 0.0400 0.0600 0.0800 0.1000
k
f
e
/
k
m
f
f
/(f
f
+f
m
)
Gilman Plot
>0. to 14.
>14. to 34.
>34. to 71.
>71. to 139.
>139. to 265.
>265. to 464.
>464. to 855.
>855. to 1521.
>1521. to 3286.
>3286. to 47463.
lA=(Adkms)/kfe
Intellectual Property of iReservoir.com, Inc
Case 1
Case 2
Case 3
Case 4
Case 5
Case 6
Case 7
1.00
10.00
100.00
1000.00
0.0000 0.0200 0.0400 0.0600 0.0800 0.1000
k
f
e
/
k
m
f
f
/(f
f
+f
m
)
Case 1
Case 2
Case 3
Case 4
Case 5
Case 6
Case 7
10.00
100.00
1000.00
10000.00
100000.00
1.00 10.00 100.00 1000.00
l
A
=
(
A
k
m
s
)
/
k
f
e
k
fe
/k
m
A B
A B