Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

136 PETROPHYSICS: RESERVOIR ROCK PROPERTIES

NMR Calibrated
NMR Prmeability Permeability
0.01 1 100 10000 0.01 1 100 10000

-NMR
Core

Figure 3.26. NMRpermeability before and after calibration [36].

EXAMPLE

An NMR log was run in a well, and indicated a porosity of 18% and
log mean of relaxation time 2.5 at the depth of 6,000 feet. Estimate the
permeability at this depth for this well, using SDR correlations.

SOLUTION

Using Equation 3.67:

k = 4(2.5)4(0.18)2= 5.06 mD

RELATIONSHIPS
PERMEABILITY-POROSITY ROCKS
IN CARBONATE

The relationship between permeability and porosity in carbonate rock


formations is related to the grain size of the rock matrix, the size of
PERMEABILITY-POROSITY RELATIONSHIPS 137

intergranular pore space, the amount of unconsolidated vugs (fractures


and solution cavities), and the presence or absence of connected
vugs [37]. Figure 3.27 is a log-log plot of the permeability-porosity
relationship for various particle size groups in the uniformly cemented
nonvuggy rocks. This plot indicates that there is a reasonably good
relation between three petrophysical parameters and, therefore, if the
particle size and the matrix porosity are known, the permeability (in
millidarcies) of the nonvuggy portion of the carbonate rock can be
estimated from:

where:

@ma= matrix porosity, fraction


A,, = grain size coefficient, dimensionless
Amcp = cementation-compaction coefficient, dimensionless.

The values of these coefficients are related to the average particle


diameter de as follows:

(1) for dWless than 20 pm, the values of Agr and Am, average 1.5x lo5
and 4.18, respectively;
(2) if d,, is n the range of 20 to 100, Agr = 2.60 x lo5 and Amcp = 5.68;
and
(3) for d,, greater than 100 pm, the values of A,, and Amcp are 8.25x 10'
and 8.18, respectively.

If the distribution of compaction or cementation is not uniform, the


constant Amcp will be affected. Patchy cementation tends to yield higher
values of Amcp, thus reducing the permeability.
To quantlfy the effect of unconsolidated fractures and cavities on the
inter-particleporosity, Lucia examined a large number of carbonate rocks
and measured visually the fraction of the total matrix porosity due to these
types of vugs [9].He found that their effect is to increase the interparticle
matrix porosity with little or no increase in the permeability of the matrix.
The following procedure is suggested for estimating the permeability in
carbonate rocks containing unconnected vugs:

1. Measure the total porosity (interparticle and unconnected vugs), Ot,


from well logs or core analysis.
2. Estimate visually unconnected vug porosity, 0".
138 PETROPHYSICS: RESERVOIR ROCK PROPERTIES

0.001 0.01 0.1 1 10 100 lo00


Pemability, mD

0.001 0.01 0.1 1 10 100 1000


Permeability, mD

100 Coan.Grainsd Samples ..-..-l- -

s
f
L
10
0
n
0 C O ~ I ~ ~no
~ Ioil
IS
0 contains Oil

1
0.001 0.01 0.1 1 10 100 Io00
Permeability. mD

Figwe 3.27. Effects of particle size on the pertneabiliq-pomstty rekattonship in


un@mzlycemented, nonvuggy carbonate rocks f37J.
PERMEABILITY-POROSITY RELATIONSHIPS 139

3. Calculate the intergranular porosity of the matrix ($ma) as

9t - 9 u (3.68b)
@ma = -
1 - 9"

4. Estimate the average particle size, dgr, using a compactor or


micrometer.
5. Calculate the permeability of the nonvuggy matrix, kma, using
Equation 3.68.

A negligibly small increase in the permeability of the matrix (with


unconsolidated vugs) will be observed if the total porosity cPt is used
in Equation 3.68 instead of @ma. Craze and Bagrintseva demonstrated
the influence of lithology on the relationship between porosity and
permeability [37,38]. On the basis of core data from cretaceous Edward
limestone (Figure 3.28), Craze noted that as the texture changes from
microgranular to coarse-grained, the permeability increases for a given
porosity [371. Bagrintseva investigated the interrelationships among
various rock properties of several carbonate reservoirs in the former
Soviet Union [38].
Chilingarian et al. used Bagrintseva's data and derived several useful
correlations between permeability and porosity by considering two
additional variables: irreducible fluid saturation and specific surface

1 10 100
Interparticle Porosity, %
Figure 3.28. Relationship between porosity and permeability for various types of
cretaceous Edwards limestone [371.
140 PETROPHYSICS: RESERVOIR ROCK PROPERTIES

area [6].The general form of the correlation is as follows:

(3.69)

where:
k = permeability to air, mD
S,, = residual water saturation, %
SvP = specific surface area
$o = open porosity, %

al, az, a3, a4, and a5 are constants for a given formation, determined
empirically.
Practically all permeability-porosity correlations should be used only
for qualitative purposes. To obtain an accurate correlation between the
porosity and permeability, one must include a large number of physical
factors that characterize a porous medium, including irreducible fluid
saturation, specific surface area, grain size distribution, grainshape,
packing and layering, lithology and mineralogy, degree and type of
cementing etc. Although some formations may show a correlation
between permeability and porosity, a large number of physical factors
influencing these two parameters differ widely in different formations.

Estimating Permeability in Carbonate Rocks

Although the absolute porosity provided by natural fractures is


negligible (< 3%), the effective porosity is considerably enhanced
because fractures connect the available pore volume. Consequently, the
reservoir permeability and petroleum recovery are greatly enhanced. The
net impact of fracture connectivity may a decisive factor in exploiting a
particular reservoir. Many methods have been proposed for estimating
fracture permeability, including parallel plate models, electric analog
systems, core analysis, we11 logging, and pressure transient testing.
The equation for volumetric flow rate between the two smooth plates,
combined with Darcy’s law, provide the basic approach for estimating
fracture permeability and its influence on fluid flow in naturally fractured
rocks. Parsons used this approach to express the total permeability
of the fracture-matrix system in which vertical fractures occur in
sets of specified spacing and orientation relative to overall pressure
gradient [39]. Murray used a parallel-plates model and a geometric
approach applicable to folded rocks to demonstrate that, in folded beds
with extension fractures normal to the bedding and parallel to the fold
axis, the fracture porosity and permeability are functions of bed thickness
and curvatures [40]. He assumed that extension fractures form primarily
PERMEABILITY-POROSITY RELATIONSHIPS 141

in the outer layers of curved beds. Murray applied this approach to the
Spanish pool in McKenzie County, North Dakota, and demonstrated a
good coincidence between areas of maximum curvature and areas of
best productivity.
The flow of fluid through porous media is directly analogous to the
flow of electricity. McGuire and Sikora used this analogy and showed
that the width of artificial fractures is much more important than their
length in affecting communication among natural fractures [41]. Stearns
and Friedman summarized that the permeability of a naturally fractured
formation can be expected to be greatest where the reservoir bed
contains wide, closely spaced, smooth fractures oriented parallel to the
fluid pressure gradient [ 4 2 ] .
Fracture permeability cannot be estimated directly from well logs.
The modern trend is to combine core-derived parameters with
computer-processed log data to establish a statistical relationship
between the permeability of the matrix-fracture system and various
parameters, such as porosity and irreducible water saturation. With such
a relationship established, the formation’s petrophysical parameters,
including permeability distribution, can be deduced from log data alone
in wells or zones without core data. In carbonate formations, however,
where structural heterogeneity and textural changes are common, and
only a small number of wells are cored because of the difficulty and
cost of the coring, the application of statistically derived correlations
is extremely limited. Watfa and Youssef developed a sound theoretical
model that relates directly to the flow of path length (tortuosity), pore
radius changes, porosity, and cementation factor m [ 4 3 ] . This model
assumes that:

(1) a porous medium can be represented by a bundle of tubes, as shown


in Figure 3.29;

Cube Length = L
k J

Figure 3.29. A bundle-oftubes model [43J.


142 PETROPHYSICS: RESERVOIR ROCK PROPERTIES

(2) the cross-sectional area of each tube, Aa, is constant; and


(3) the fluid path and the electric current path are same and the
true conductivity, i.e., the reciprocal of resistivity, of the bundle
of tubes is:

where C, is the water conductivity and m is the cementation factor.


Because the apparent conductivity Ca of a block having a cross-sectional
area A and length L is related to the true conductivity C& by the following
expression:

(3.71)

where A is the cross-sectional area of all tubes, then

(3.72)

assuming that the bundle-of-tubes model contains n tubes, the


conductivity of the ith tube (Ci) can be defined as:

(3.73)

where is the length of the ith tube.


The porosity of the tube is unity. The apparent conductivity of the
block is the sum of the individual conductivities of all the tubes. Thus:

(3.74)

From the assumption that A2 is constant:

(3.75)

By definition:

(3.76)
PERMEABILITY-POROSITY RELATIONSHIPS 143

Combining Equations 3.70 and 3.75 yields:

(3.77)

and the tortuosity is:

Using the same approach, the effects of flow path on the permeability
can be evaluated.Applying Poiseiulle’sequation to the ith tube, the flow
rate in the ith tube, qi, is equal to:

(3.79)

where rpai, 1, and AP are, respectively, the apparent radius of the


ith tube, fluid viscosity, and pressure differential across the unit block.
For n tubes, the total flow rate q is:

(3.80)

and, assuming A2 is constant, the flow rate is:

(3.81)

Applying Darcy’s law to the unit block, the flow rate is equal to:

(3.82)

Combining Equations 3.76, 3.77, 3.79, 3.80, and 3.81, and solving for
the apparent permeability of the block, ka:

(3.83)
144 PETROPHYSICS: RESERVOIR ROCK PROPERTIES

This equation is similar to Equation 3.14 for m = 1 . Combining


Equations 3.83 and 3.78 gives:

ka = (")? (3.84)
8 z

Assuming z = QFR, where FR is the formation resistivity factor,


Equation 3.59 becomes:

(3.85)

Expressing tortuosity as z = (QFR)~,


Equation 3.84 results in:

ka=(%)g1

and, for z = @(FR)~,


Equation 3.84 gives

These equations clearly indicate that no single correlation can be used


to determine the formation permeability from logs alone.
If ka is expressed in mD, rPain pm, Equation 3.83 becomes:
ka = 126.7rpaQ
2 m (3.88)

Figure 3.30 is a semilog plot of this relationship. The Cartesian axis on


this plot is Qm instead of the conventional +.The importanceof including
dimensions of the flow channels in developing k-Q relationships for
carbonates is clearly demonstrated by this plot. FQuation 3.88, which
also is applicable to sandstones, is derived on the basis that the average
pore radius of the flow channels remains constant along the length of the
unit block. As shown in Figure 3.31(A), however, the true pore radius
changes along the flow path length. The effect of changing cross-sectional
area along the flow path can be evaluated by considering the system
of Figure 3.31(B) as two resistors in series. The total conductivity C
of this system is related to the two conductivities C1 and C2 by the
paralIel-conductivity equation:

(3.89)
PERMEABILITY-POROSITY RELATIONSHIPS 145

100,000

10.000

1.000

22

100

10
0 0.05 0.1 0.15 0.2 0.25
tJm
Figure 3.30. Variations of ka,Qm, and rpa for an ideal system of tube bundles [43].

4 L
b
(B)

Figure 3.31. A and B: Variations inflowpath length andpore radius with variations
in grain size [43].

Substituting Equation 3.73 into the above expression, it can be shown


that the change in conductivity caused by the change in pore radius is as
follows:

(3.90)

where = L1/L (Figure 3.31B), Ar = 1 - A1 /Aa, and Ca is the apparent


conductivity of the block such that Equation 3.52 is true. Combining
the Darcy and Poiseuille equations, it can be shown that the effect of
146 PETROPHYSICS: RESERVOIR ROCK PROPERTIES

pore-radius changes on the true permeability k is:

(3.91)

where kR is the ratio of apparent permeability to absolute permeability.


+ +
Assuming AP = AP1 AP2 and AaL = AIL1 A2L2, Equation 3.88
becomes:
2
r ~ am
k = 126.7-$ (3.92)
kR

Letting rpe, the effective pore radius, be equal to rpa/&, Equation 3.92
becomes similar to Equation 3.88:

k = 126.71-:~$~ (3.93)

The value of rpe can vary considerably from the average radius value
E, depending on the texture and heterogeneity present in the system.
Consider two systems with different grain sizes and with no vugs or
fractures (Figure 3.32). Because:

(a) the path of the current represents the true inter-matrix tortuosity,
(b) the tortuosity is a function of grain size and usually decreases with
decrease in the grain size, and
(c) the value of rpe varies with the variation in grain size,
(d) a relationship between z and rpe must exist for a particular formation.
Using experimental data of rpa, m, $, and k, Wafta and Youssef
showed that rpe and z are related as follows [43]:

logrp, = a1& + a2 (3.94)

.-
I....

-"...+
Figure 3.32. Two systems with dqferent grain sizes and pore radii [43J
PERMEABILITY-POROSITY RELATIONSHIPS 147

4 ,-...- . - .-
.. - .~~
. ..
. . .. . . .-.... . .. .....
.
. ...
I
I
3.5
I
E I

‘ ~ , , , , ,

Potential Fracture
Z
‘one
1 2 3 4 5 6 7 8 9 10

Secondary Porosity Index @PI%)


Figure 3.33. Estimation of the value of the inter-math Archie factor (mm) from
Archiefactor (m) and seconday porosity index (i32) cross-plot [43J

where the coefficients a1 and a2 can be determined according to the


following procedure:

(a) Obtain values of m and 0 from well logs and k from core analysis.
(b) Determine the cementation factor of the matrix mm from:

(3.95)

where Is2 is the secondary porosity index, SPI, i.e., - @SL, where Qt
and QSL are, respectively, the total porosity and the sonic log porosity.
Figure 3.33 shows how to obtain mm from a plot of the cementation
factor m versus SPI. To compensate for the effects of fractures, data
points for Is2 < 1% are not used to obtain m,. Inasmuch as the curve
is not linear, one needs to be careful when extrapolating the curve
to Is2 = 1 to obtain mm on the m axis.
(c) Determine the value of the effective pore radius rpe from
Equation 3.93.
(d) Calculate the tortuosity from Equation 3.78.
(e) Establish a data bank for rpe and z, and plot log rpe versus ,h.
(0 Draw the best-fit straight line. The general form of this line is given
by Equation 3.95.
(g) Determine the correlation coefficient a2 from the rpe log-axis at
, h = O a n d a l fromthe,haxisatrp,= 1.

DIRECTIONAL
PERMEABILITY
In homogeneous reservoirs, permeability is assumed to be the same
in all directions. However, in heterogeneous reservoirs, permeability in

You might also like