26 Properties of Reservoir Rock

Chapter 26
Properties of Reservoir Rocks

Daniel M. Bass Jr., C~hado School of Mmea*
Introduction
This chapter deals with the fundamental properties of
reservoir rocks. The properties discussed are (1) pomsi-
8-a measure of the void space in a rock: (2)
pertneahi/it~-a measure of the fluid transmissivity of a
rock; (3) fluid saturutim-a measure of the gross void
space occupied by a fluid; (4) cupillaryprrssure rek-
rim--a measure of the surface forces existing between
the rock and the contained fluids; and (5) electrical cm-
clucrivity offluid-saturated rocks-a measure of the con-
ductivity of the rock and its contained fluids to electric
current. These properties constitute a set of fundamental
parameters by which the rock may be described
quantitatively.
Typical core-analysis data are presented to illustrate
the description of porous media by these fundamental
properties.
Porosity
Porosity is defined as the ratio of the void space in a rock
to the bulk volume (BV) of that rock, multiplied by 100
to express in percent. Porosity may be classified accord-
ing to the mode of origin as primary and secondary. An
original porosity is developed during the deposition of
the material, and later compaction and cementation
reduce it to the primary porosity. Secondary porosity is
that developed by some geologic process subsequent to
deposition of the rock. Primary porosity is typified by
the intergranular porosity of sandstones and the inter-
crystalline and oolitic porosity of some limestones.
Secondary porosity is typified by fracture development
as found in some shales and limestones and the vugs or
solution cavities commonly found in limestones. Rocks
having primary porosity are more uniform in their
Thts author also Wrote the orlglnal chapter on this topic m the 1962 edition
with coauthor James W. Amyx (deceased] D.M Bass Jr IS currently a petroleum
COnSUltant
characteristics than rocks in which a large part of the
porosity is induced. For direct quantitative measurement
of porosity, reliance must be placed on formation
samples obtained by coring.
Unit cells of two systematic packings of uniform
spheres are shown in Fig. 26.1. The porosity for cubical
packing (the least compact arrangement) is 47.6% and
for rhombohedral packing (the most compact arrange-
ment) is 25.96%. Considering cubical packing, the
porosity may be calculated as follows. The unit cell is a
cube with sides equal to 2r where r is the radius of the
sphere. Therefore, Vh=(2r)3=8r3, where V,] is the
bulk volume. Since there are 8% spheres in the unit cell.
the sand-grain volume, V,, , is given by
4ar
v, = -
3
The porosity, $, is given by
where L,, is PV. Therefore,
8r -413irr3
4=
8r3
X100=(1--&00
=47.6%.
Of particular interest is the fact that the radii cancel and
the porosity of uniform spheres is a function of packing
only.
26-2 PETROLEUM ENGINEERING HANDBOOK
POROSITY: 47.6 % POROSITY= 25.96%
CuB~~~KXE;IDE
RHOMBOHEDRAL OR
CLOSE PACKED
Fig. 26.1-Unit cells and groups of uniform spheres for cubic
and rhombohedral packing.
Tickell et al. has presented experimental data in-
dicating that, for packings of Ottawa sand, porosity was
a function of skewness of the grain-size distribution (see
Fig. 26.2). Skewness is a statistical measure of the
uniformity of distribution of a group of measurements.
Other investigators have measured the effects of distribu-
In dealing with reservoir rocks (usually consolidated
sediments), it is necessary to define total porosity and ef-
tion, grain size, and grain shape. In general, greater
fective porosity because cementing materials may seal
off a part of the PV. Totalporosity is the ratio of the total
angularity tends to increase the porosity, while an in-
void space in the rock to the BV of the rock, while effec-
tive porosity is the ratio of the interconnected void space
crease in range of particle size tends to decrease
in the rock to the BV of the rock, each expressed in per-
cent. From the reservoir-engineering standpoint, effec-
porosity.
tive porosity is the desired quantitative value because it
represents the space that is occupied by mobile fluids.
For intergranular materials, poorly to moderately well
cemented, the total porosity is approximately equal to
the effective porosity. For more highly cemented
materials and for limestones, significant differences in
total-porosity and effective-porosity values may occur.
-.08 -.04 0 +.04 +.08
SKEWNESS
Fig. 26.2-Variation of porosity with skewness of grain-size
distribution.
Photographs of oilwell cores are presented in Fig.
26.3. 3 The pore configuration of the sandstones shown
is complex, but the pores are distributed relatively
uniformly. Complex pore configurations arise from the
interaction of many factors in the geologic environment
Material having induced porosity, such as the car-
bonate rocks shown in Fig. 26.3, have even more com-
of the deposit. These factors include the packing and
plex pore configurations. In fact, two or more systems of
pore openings may occur in such rocks. The basic rock
particle-size distribution of the framework fraction, the
material is usually finely crystalline and is called the
matrix.
type of interstitial material, and the type and degree of
The matrix contains uniformly small pore
openings that comprise one system of pores. One or
cementation. The influence of these various factors may
more systems of larger openings usually occur in car-
bonate rocks as a result of leaching, fracturing, or
be evaluated as statistical trends.
dolomitization of the primary rock material. Vugular
pore openings are frequently as large as an ordinary lead
pencil and usually are attributed to leaching of the rock
subsequent to deposition. Fractures also may be quite
large and contribute substantially to the volume of pore
openings in the rock.
(b) (d)
Fig. 26.3-Oilwell cores. Consolidated sandstone: (a) wireline core, Lower Frio; (b) whole core, Seven
Rivers. Vugular, solution cavities, and crystalline limestone and dolomite: (c) whole core,
Devonian; (d) whole core, Hermosa.
PROPERTIES OF RESERVOIR ROCKS 26-3
Laboratory Measurement of Porosity
Numerous methods have been developed for the deter-
mination of the porosity of consolidated rocks having in-
tergranular porosity. Most of the methods developed
have been designed for small samples, roughly the size
of a walnut. Since the pores of intergranular material are
quite small, determining the porosity of such a sample
involves measuring the volume of literally thousands of
pores. The porosity of larger portions of the rock is
represented statistically from the results obtained on
numerous small samples.
In the laboratory measurement of porosity, it is
necessary to determine only two of the three basic
parameters (BV, PV, and grain volume). In general. all
methods of BV determination are applicable to determin-
ing both total and effective porosity.
BV. Usual procedures use observation of the volume of
fluid displaced by the sample. This procedure is par-
ticularly desirable because the BV of specially shaped
samples may be determined as rapidly as that of shaped
samples.
The fluid displaced by a sample may be observed
either volumetrically or gravimetrically. In either pm
cedure, it is necessary to prevent fluid penetration into
the pore space of the rock. This may be accomplished by
(1) coating the rock with paraffin or a similar substance,
(2) saturating the rock with the fluid into which it is to be
immersed, or (3) using mercury, which by virtue of its
surface tension and wetting characteristics does not tend
to enter the small pore spaces of most intergranular
materials.
Gravimetric determinations of BV may be accom-
plished by observing the samples weight loss when im-
mersed in a fluid or the difference in weight of a pyc-
nometer when filled with a fluid and when filled with
fluid and the core sample. The details of gravimetric
determinations of BV are best summarized by Example
Problems I through 3.
Example Problem l-Coated Sample Immersed in
Water. Given that the
weight of dry sample in air. A=20.0 g.
weight of dry sample coated with paraffin, B=20.9 g
(density of paraffin=0.9 gicmj), and
weight of coated sample immersed in water at 40F.
C= 10.0 g (density of water= 1 .OO g/cm3).
we can then calculate that
weight of paraffin=&A=20.9-20.0=0.9 g.
0.9
volume of paraffin = - = 1 cm3 ,
0.9
weight of water displaced=B-C=20.9- 10.0
= 10.9 g,
10.9
volume of water displaced = ~ = 10.9 cm3,
1 .o
volume of water displaced-volume of paraffin
=10.9-1.0=9.9 cm3, and
BV ofrock, Vb,=9.9 cm3.
Example Problem 2-Water-Saturated Sample Im-
mersed in Water. Given that
weight of saturated sample in air, D=22.5 g, and
weight of saturated sample in water at 40F,
E= 12.6 g.
we can calculate that
weight of water displaced=D-E=22.5- 12.6
x9.9 g, and
9.9
volume of water displaced= - =9.9 cm3.
1.0
Therefore,
BV rock, If,,, =9.9 cm3.
Example Problem 3-Dry Sample Immersed in Mer-
cury Pycnometer. Given A from Example Problem I
and the following values,
weight of pycnometer filled with mercury at 20C.
F=350.0 g and
weight of pycnometer filled with mercury and sample
at 20C. G=235.9 g (density of mercury=
13.546 g/cm3).
we can calculate that
weight of sample + weight of pycnometer filled with
mercury=A+F=20+350=370 g,
weight of mercury displaced =A+F-G=370-
235.9= 134.1 g. and
134.1
volume of mercury displaced = - =9.9 cm3
13.546
Therefore,
BV of rock=9.9 cm3.
Determinations of BV volumetrically use a variety of
specially constructed pycnometers or volumeters. An
electric pycnometer from which the BV may be read
directly is shown in Fig. 26.4. The sample is immersed
in the core chamber, causing a resulting rise in the level
of the connecting U tube. The change in mercury level is
measured by a micrometer screw connected to a low-
voltage circuit. The electric circuit is closed as long as
the measuring point is in contact with the mercury. The
travel of the measuring point is calibrated in volume
units such that the difference in the open-circuit readings
with and without the sample in the core chamber
represents the BV of the sample. Either dry or saturated
samples may be used in the device.
Sand-Grain Volume (GV). The various porosity
methods usually are distinguished by the means used to
determine the GV or PV. Several of the oldest methods
of porosity measurement arc based on the determination
of GV.
The GV may be determined from the dry weight of the
sample and the sand-grain density. For many purposes,
results of sufficient accuracy may be obtained by using
the density of quartz (2.65 g/cm3) as the sand-grain
density.
For more rigorous determinations either the A.F.
Melcher-Nutting4 or Russell method may be
employed. The BV of a sample is determined: then this
sample, or an adjacent sample, is reduced to grain size
and the GV is determined. In the M&her-Nutting
technique, all the measurements are determined
26-4
PETROLEUM ENGINEERING HANDBOOK
Type of sampling
Functions measured
Manner of measurement
Errors
TABLE 26.1~METHODS OF DETERMINING POROSITY
Effective Porosity Methods
Washburn-Bunting
Porosimeter
One to several pieces
per increment (usually
one).
Solvent extraction and
oven drying. Occa-
sionally use retort
samples.
PV and BV.
Reduction of pressure
on a confined sample
and measurement of air
evolved. BV from mer-
cury pycnometer.
Air from dirty mercury;
possible leaks in
system; incomplete
evacuation caused by
rapid operation or tight
sample.
Stevens
Porosimeter
one).
oven drying. Occa-
sionally use retort
samples.
Sand grain volume and
unconnected PV and
BV.
Difference in volume of
air evolved from a
constant-volume
chamber when empty
and when occupied by
sample. BV by Russell
tube.
Mercury does not
become dirty. Possible
leaks in system; in-
complete evacuation
caused by rapid opera-
tion or tight sample.
Kobe Porosimeter
one).
oven drying. Occa-
sionally use retort
samples.
unconnected PV and
BV.
air evolved from a
constant-volume
chamber when empty
tube.
Mercury does not
complete evacuation
Boyles Law
Porosimeter
one).
oven drying. Occa-
sionally use retort
samples.
unconnected PV and
BV.
air evolved from a
constant-volume
chamber when empty
tube.
Mercury does not
complete evacuation
MICROMETER SCALE
ADJUSTING SCREW
OH
INDICATOR
LIGHT
gravimetrically, using the principle of buoyancy (Exam-
ple Problem 2). The Russell method uses a specially
designed volumeter, and the BV and GV are determined
volumetrically. The porosity determined is total porosi-
ty, $r. Thus,
Vb - vs
+r=-
x loo.
vb
From the data of Example Problem 2 and using a sand-
grain density of 2.65 g/cm3,
vb =9.9 cm3,
20
v,=- =7.55 cm3,
2.65
and
9.9-1.55
+r=
x 100=23.8%.
9.9
Fig. 26.4-Electric pycnometer
The Stevens6 porosimeter is a means of determining the
effective GV. The porosimeter (Fig. 26.5) consists of
a core chamber that may be sealed from atmospheric
pressure and closed from the remaining parts of the
porosimeter by a needle valve. The accurate volume of
the core chamber is known. In operation, a core is placed
in the core chamber; a vacuum is established in the
PROPERTIES OF RESERVOIR ROCKS
TABLE 26.1-METHODS OF DETERMINING POROSITY (continued)
Effective Porositv Methods Total Porosity Method
26-5
Core Laboratories, Core Laboratories,
Dry Sample
one).
Saturation*
one).
Wet Sample Sand Density
Several pieces per
increment.
Several pieces for
retort; one for mercury
pump.
overy drying. Occa-
sionally use retort
samples.
PV and BV.
None. Solvent extraction and
oven drying. Occa-
sionally use retort
samples.
Solvent extraction;
then, in second step,
crush sample to grain
size.
Volumes of gas space,
oil, and water. BV.
Sand-grain volume and
unconnected PV, and
BV.
BV of sample and
volume of sand grains.
Weight of dry sample;
weight of saturated
sample in air: weight of
saturated sample im-
mersed in saturated
fluid.
Weight of retort sam-
ple; volume of oil and
water from retort sam-
ple; gas volume and
BV of mercury-pump
sample.
air evolved from a
constant-volume
chamber when empty
sample.
Weight of dry sample;
weight of saturated
sample immersed;
weight and volume of
sand grains.
Possible incomplete
saturation.
Obtain excess water
from shales. Loss 01
vapors through con-
densers.
Possible leaks in
system; incomplete
evacuation caused by
rapid operation or tight
sample.
Possible loss of sand
grains in crushing. Can
be reproduced most
accurately.
system by manipulating the mercury reservoir; the air in
the core chamber is expanded repeatedly into the
evacuated system and measured at atmospheric pressure
in the graduated tube. The Stevens method is an adapta-
tion of the Washburn-Bunting7 procedure, which will be
discussed with the measurement of PV.
Example Problem 4-Determination of Grain
Volume by Gas Expansion (Stevens Porosimeter).
Given that
volume of core chamber, H= 15 cm3
and
total reading, 1=7.00 cm3
where
volume of air (first reading)=6.970,
volume of air (second reading)=0.03, and
volume of air (third reading)=O,
we can calculate
effective grain volume=H-Z=8 cm3,
bulk volume of sample (from pycnometer)= 10 cm3,
and effective porosity, 4, =[(lO-8)110] X 100=20%.
PV. All methods of measuring PV yield effective
porosity. The methods are based on either the extraction
of a fluid from the rock or the introduction of a fluid into
the pore space of the rock.
The Washburn-Bunting 7 porosimeter measured the
volume of air extracted from the pore space by creating a
partial vacuum in the porosimeter by manipulation of the
attached mercury reservoir. In the process, the core is
exposed to contamination by mercury and, therefore, is
PULLEY.
MERCURY -
RESERVOIR
--P
GTOPCOCK
;RADUATED
TUBE
-CORE
:HAMBER
--CRANK
\
BRAKE
CONNECTING
HOSE -
Fig. 26.5-Stevens porosimeter
26-6
Sample
Number Type of Material
Approximate Gas
Permeability
(md)
1 LImestone 1
2 Fretted glass 2
3 Sandstone 20
4 Sandstone 1,000
BZE Semiquartzltic sandstone 0.2
0.x Semlquartzitic sandstone 0.8
61-A Alundum 1,000
722 Alundum 3
1123 Chalk 16
1141-A Sandstone 45
TABLE 26.2-CHARACTERISTICS OF SAMPLES USED IN POROSITY-MEASUREMENT COMPARISONS
Porosity (O/O)
Average From Average From Value From
Average
17.47
28.40
14.00
30.29
3.95
3 94
28.47
16.47
32.67
19.46
not suitable for additional tests. The previously de-
scribed Stevens method is a modification of the
Washburn-Bunting procedure especially designed to pre-
vent mercury contamination of the samples.
A number of other devices have been designed for
measuring PV; these include the Kobe8 porosimeter, the
Oilwell Research porosimeter, and the mercury-pump
porosimeter. Kobe and Oilwell Research porosimeters
are Boyles-law-type porosimeters designed for use with
nitrogen or helium with negligible adsorption on rock
surfaces at room temperature. The mercury-pump
porosimeter is designed so that the BV may be obtained
as well as the PV.
The saturation method of determining porosity con-
sists of saturating a clean dry sample with a fluid of
known density and determining the PV from the gain in
weight of the sample. The sample usually is evacuated in
a vacuum flask to which the saturation fluid may be ad-
mitted by means of a separatory funnel. If care is exer-
cised to achieve complete saturation, this procedure is
believed to be one of the best available techniques for in-
tergranular materials. Example Problem 5 illustrates the
saturation technique of measuring PV.
Example Problem G-Effective Porosity by the
Saturation Method. From the data of Example Prob-
lems 1 and 2, we can calculate
weight of water in pore space=D-A=22.5-20
=2.5 g.
EXPANSION CHAMBER
KNOWN VOLUME
COMPRESSED
GAS SOURCE
CORE CHAMBER
OF
KNOWN VOLUME
Fig. 26.6-Gas-expansion porosimeter for large cores.
Gas Saturation High
Methods Methods Observation
17.81 16.96 18.50
26.60 27.97 29.30
14.21 13.70 15 15
31.06 29.13 31.8
4.15 3.66 4.60
4.10 3.71 4.55
28.70 28.00 29.4
16.73 16.08 17.80
33 10 32.03 33.8
19.68 19.12 20.2
Value From
Low
Observation
16.72
27.56
13 50
26 8
3 50
3.48
27.8
16.00
31 7
100
2.5 g
volume of water in pore space = ~
1 g/cm3
=2.5 cm3,
effective PV=2.5 cm, and
BV (Example Problem 2)=9.9 cm3.
Therefore,
2.5
~,=~x100=25.3%.
Several methods of determining effective porosity are
compared in Table 26.1.
Precision of Porosity Measurements. A group of ma-
jor company laboratories conducted a series of tests to
determine the precision of porosity measurements. The
method used was either gas expansion or a saturation
technique. Table 26.2 summarizes the results of these
tests. Note that the gas-expansion method is consistently
higher than the saturation method. This undoubtedly
results from the fact that the errors inherent to each tend
to be in opposite directions. In the case of the gas-
expansion method, errors caused by gas adsorption
would cause high values to be obtained while, for the
saturation techniques, incomplete saturation of the sam-
ple would result in low values. The difference in the
average values obtained by the two methods is about
0.8% porosity, which is approximately a 5% error for a
16% porosity sample. However, it is felt that all the
methods commonly used to determine effective porosity
yield results with the desired degree of accuracy if
carefully performed.
Carbonate Rocks. Small samples, such as used in the
routine techniques already discussed, yield values of
porosity that do not include the effect of vugs, solution
cavities, etc. The saturation methods of determining PV
and BV are unsatisfactory because drainage will occur
from the larger pore spaces. Therefore, it is necessary to
use larger core samples and to determine the BV by
measurement of the core dimensions or after coating the
sample. The effective grain volume is obtained by using
a large gas-expansion porosimeter of the type shown in
Fig. 26.6. This porosimeter is based on Boyles law in
26-7
50
0
0 1000 2000 3000 4000 5000 6CCO
DEPTH OF BURIAL IN FEET
Fig. 26.7-Effect of natural compaction on porosity.
which high-pressure gas is equalized between two
chambers. The porosity may be calculated from the
measured pressures, the volume of either chamber, and
the bulk-sample volume by the use of Boyles law.
Keltont reported results of whole-core analysis. a
method utilizing large sections of the full-diameter core.
The following table of matrix vs. whole-core data sum-
marizes a part of Keltons work. lo
Grollp
1 2 3 4
Matrix porosity, % bulk 1.98 1.58 2.56 7.92
Total porosity, % bulk 2.21 2.62 3.17 8.40
Matrix porosity is that determined from small samples;
total porosity is that determined from the whole core.
Whole-core analysis satisfactorily evaluates most car-
bonate rocks. However, no satisfactory technique is
available for the analysis of extensively fractured
materials because the samples cannot be put together in
their natural state.
Compaction and Compressibility of Porous Rocks
The porosity of sedimentary rocks has been shown by
Krumbein and Sloss to be a function of the degree of
compaction of the rock. The compacting forces are a
function of the maximum depth of burial of the rock. The
effect of natural compaction on porosity is shown in Fig.
26.7. This effect is caused principally by the resulting
packing arrangement after compaction; thus, sediments
that have been buried deeply, even if subsequently
uplifted. exhibit lower porosity values than sediments
that have not been buried at great depth.
Geertsma12 states that three kinds of compressibility
must be distinguished in rocks: (I) rock-matrix com-
pressibility, (2) rock-bulk compressibility, and (3) pore
compressibility. Rock-bulk compressibility is a com-
bination of pore and rock matrix compressibility.
Rock-matrix compressibility is the fractional change in
volume of the solid rock material (grains) with a unit
change in pressure. Pore compressibility is the fractional
change in PV of the rock with a unit change in pressure.
POROSITY-PER CENT
Fig. 26.8-Effective reservoir-rock compressibilities
The depletion of fluids from the pore space of a reser-
voir rock results in a change in the internal stress in the
rock, thus causing the rock to be subjected to a different
resultant stress. This change in stress results in changes
in the GV, PV, and BV of the rock. Of principal interest
to the reservoir engineer is the change in the PV of the
rock. The change in rock-bulk compressibility may be of
importance in areas where surface subsidence could
cause appreciable property damage.
Hall I3 reported PV compressibility as a function of
porosity. These data are summarized in Fig. 26.8. The
effective rock compressibility in Fig. 26.8 results from
the change in porosity caused by grain expansion and
decrease in pore space because of compaction of the
matrix.
Fatt I4 indicates that the pore compressibility is a func-
tion of pressure. Within the range of his data, he was
unable to find a correlation with porosity.
725 Gradient (North
4 Gradient Sand
4 W.M.L(S.W.Tex.1
6
I I I I I I
35 30 25 20 15 IO 5
Formation Compressibility(Microsip1
1
Fig. 26.9-Depth vs. formation compressibility in abnormally
pressured segment of an abnormally pressured
reservoir.
/ I I
l
l \eet
-I
l
l
l
0
J
35
Fig. 26.10-PV compressibility at 75% lithostatic pressure vs.
initial sample porosity for limestones.
I I I I I I
0
initial sample porosity for consolidated
sandstones.
100
I I I I I
I
0
0 0
0 O 0 08 0 0
0 O
0
0 00 0 0
0
00
00
@ OQ
0
0
0
B
0
\ 0 0 O 0 0 0 0
l
CORRELPlTlON \
0
0
H&LLS -- 0 00
Ooo 0
0
@>&.
0 00:
0
0 0 I
0
0
initial sample porosity for friable sandstones.
The additional significance of changes in porosity with
the discovery of oil at deeper depths and new geological
areas resulted in the need for a better understanding of
the changes in porosity with the depletion of reservoir
fluid pressure.
Hammerlindl I5 developed a correlation from
measured field data that indicates the change in porosity
compressibility with depth of burial of an unconsolidated
sand (Fig. 26.9). Similar correlations have also been
presented by others in the technical literature.
Considerable laboratory work has been performed
recently in an attempt to understand better the effect of
formation compaction on porosity. Newman I6 per-
formed measurements on samples of limestone and con-
solidated, friable, and unconsolidated sandstones. He
compared his results with those of Hall I3 and van der
Knaap, as illustrated in Figs. 26.10 through 26.13. As
noted in his data, an approximate correlation may exist
between PV compressibility and porosity for limestones
and consolidated sandstones. Little or no correlation ex-
ists between PV compressibility and porosity for friable
and unconsolidated sandstones. By averaging the PV
compressibility in 5% ranges of porosity, Newman at-
tempted to correlate all four types of porous media. The
results of this averaging technique are presented in Fig.
26.14.
The methods used to measure PV compressibility have
come under discussion. PV compressibility can be
measured in the laboratory by the hydrostatic (same
pressure in all three directions) or the triaxial (different
pressure in the z direction than in the x and y directions)
techniques. The test samples also can be stress cycled
initial sample porosity for unconsolidated
sandstones.
until the strain resulting from a supplied stress is con-
stant, or the sample can be placed under stress and the
strain measured for purposes of calculating PV
compressibility.
Krug I8 and Graves I9 demonstrated that when a for-
mation sample was stress cycled to a stable strain condi-
tion, the sample gave repeatable values of PV com-
pressibility even when the sample was left in an un-
stressed condition for 30 days or more. The data reported
by Newman were for samples that were not stress
cycled.
Lachance* compared PV compressibilities obtained
by the hydraulic and triaxial methods. The reported
results (Fig. 26.15) indicate a large difference in the
magnitude of PV compressibilities obtained by the two
methods. The triaxial data indicate that PV com-
pressibility is essentially independent of the sample
porosity. Newmans I6 data were obtained by the
hydrostatic method, whereas Krugs I8 and Graves I9
data were obtained by the triaxial method.
In summary, rock compressibility is an important fac-
tor in reservoir evaluation. Oil reservoirs with high in-
itial pressures and low fluid bubblepoint pressures are
sensitive to the true value of PV compressibility. Gas
reservoirs with initial reservoir pressures in excess of
6,000 psi are also sensitive to the value of PV com-
pressibility. Newman, I6 Krug, I8 and Graves I9 all
recommend that PV compressibility be measured on
Fig. 26.14-Class averages of PV compressibility vs. initial
sample porosity.
Fig. 26.15-Calculated PV compressibility (porosity 2 8%).
26-10
-- . .
--- Lf
_--
-_-
6
3
II --
---
]
-_-
---I
- _-
---
--_
r/--i
a
t
h,
.--------_
;
h-h,
Fig. 26.16-Schematic of Henry Darcys experiment on flow of
water through sand.
samples from the reservoir in question when the PV
compressibility may be significant in reservoir
evaluation.
Permeability
Introductory Theory
It is the purpose of this section to discuss the ability of
the formation to conduct fluids. In the introduction to
API Code 27, it is stated that permeability is a proper-
ty of the porous medium and a measure of the mediums
capacity to transmit fluids. The measurement of
permeability, then, is a measure of the fluid conductivity
of the particular material. By analogy with electrical con-
ductors, the permeability represents the reciprocal of the
resistance that the porous medium offers to fluid flow.
The following equations for flow of fluids in circular
conduits are well known.
Poiseuilles equation for viscous flow:
d2Ap
v=-
321.LL ) . . . . . . . . . . . . . . . . ..I..........
(1)
Fannings equation for viscous and turbulent flow:
2dAp
$=-
fpL , . . . . . .
(2)
where
v = fluid velocity, cm/s,
d = diameter of conductor, cm,
Ap = pressure loss over length L, dynes/cm2,
L = length over which pressure loss is
measured, cm,
p = fluid viscosity, Pas,
p = fluid density, g/cm, and
f = friction factor, dimensionless.
A more convenient form of Poiseuilles equation is
ar4Ap
4=
, . . . . . . . . . 8clL (3)
where r is the radius of the conduit, cm, q is the volume
rate of flow, cm/s, and the other terms are as previously
defined.
If a porous medium is conceived to be a bundle of
capillary tubes, the flow rate q1 through the medium is
the sum of the flow rates through the individual tubes.
Thus,
qr=s i lljrj4, . . . .
P I 1
where n,, is the number of tubes of radius rj. If
(T/8) $ nj rj 4
j=l
(4)
is treated as a flow coefficient for the particular grouping
of tubes the equation reduces to
qr=c$; . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...(5)
where
k
The pore structure of rocks does not permit simple
classification of the flow channels. Therefore, empirical
data are required in most cases.
In 1856, Darcy investigated the flow of water through
sand filters for water purification. His experimental ap-
paratus is shown schematically in Fig. 26.16. Darcy
interpreted his observations to yield results essentially as
given in Eq. 7.
ht --h2
q=KA- . . . . . . . . . . . . . . . . . . . . . . L
(7)
q represents the volume rate of flow of water downward
through the cylindrical sandpack of cross-sectional area
A and length L. h 1 and h2 are the heights above the stan-
dard datum of the water in manometers located at the in-
put and output faces and represent the hydraulic head at
Points 1 and 2. K, a constant of proportionality. was
found to be characteristic of the sandpack. Darcys in-
vestigations were confined to flow of water through
sandpacks that were 100% saturated with water.
Later investigators found that Darcys law could be
extended to fluids other than water and that the constant
of proportionality K could be written as klp, where p is
the viscosity of the fluid and k is a proportionality con-
stant for the rock. The generalized form of Darcys law
as presented in API Code 27 is presented in Eq. 8.
u,yz-; ($-gp;). . .
where
s = distance in direction of flow, always
positive,
us =
volume flux across a unit area of the porous
medium in unit time along flow path S,
z = vertical coordinate, considered positive
downward, cm,
p = density of the fluid,
g = acceleration of gravity,
- = pressure gradient along s at the point to
ds
which u refers,
p = viscosity of the fluid,
k = permeability of the medium, and
dt
-=
sin 8, where fl is the angle between s and
ds
the horizontal.
u, may further be defined as q/A where q is the volume
rate of flow and A is the average cross-sectional area
perpendicular to the lines of flow.
The portion of Eq. 8 in parentheses may be interpreted
as the total pressure gradient minus the gradient caused
by a head of fluid. Thus, if the system is in hydrostatic
equilibrium, there is no flow and the quantity inside the
parentheses will be zero. Eq. 8 may be rewritten as
kd
u, =--@gz-p). . . . . . . . . .
fib
The quantity d(pgz-p)lds may be considered to be the
negative gradient of a potential function b, where
a=p-pgz. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(lO)
The potential function is defined such that flow will be
from higher to lower values.
The dimensions of penmeabilit
P
may be established
from an analysis of Eq. 8 as k=L . In the cgs system of
units, the unit of permeability would be cm*, a large unit
for common usage; therefore, the petroleum industry
adopted the darcy as the standard unit of permeability,
which is defined as follows.
A porous medium has a permeability of one darcy when a single
phase fluid of one centipoise viscosity that completely fills the
voids of the medium, will flow through it under conditions of
viscous flow at a rate of one cubic centimeter per second per
square centimeter cross sectional area under a pressure or
equivalent hydraulic gradient of one atmosphere per
centimeter.
Fig. 26.17-Sand model for rectilinear flow of fluids.
Conditions of viscous flow mean that the rate of flow
will be sufficiently low to be directly proportional to the
potential gradient.
Darcys law holds only for conditions of viscous flow
as defined. Further, for the permeability k to be a propor-
tionality constant of the porous medium, the medium
must be 100% saturated with the flowing fluid when the
determination of permeability is made. In addition, the
fluid and the porous medium must not react-i.e., by
chemical reaction, adsorption, or absorption. If a reac-
tive fluid flows through a porous medium, it alters the
porous medium and, therefore, changes the permeability
of the medium as flow continues.
Flow Systems of Simple Geometry
Horizontal Flow. Horizontal rectilinear steady-state
flow is common to virtually all measurements of
permeability. If a rock is 100% saturated with an incom-
pressible fluid and is horizontal (Fig. 26.17), then
dz/ds=O, dplds=dpi&, and Eq. 8 reduces to
4
-k dp
ux=-=--3
A
tth
which on integration becomes
kA(P I -P*)
9=
cLL , .,....................
(11)
where k is the specific permeability.
If a compressible fluid flows through a porous
medium, Darcys law, as expressed in Eq. 8, is still
valid. However, for steady flow, the mass rate of flow
rather than the volume rate of flow is constant through
the system. Therefore, the integrated form of the equa-
tion differs. Considering steady rectilinear flow of com-
pressible fluids, Eq. 8 becomes
b dp
pux=---, .,..........,,........... (12)
ttb
or, for steady-state flow,
pux=p: =constant.
I
FREE FLOW
II
FLOW UNDER
HEAD h
III
FLOW UNDER
HEAD h
Fig. 26.1 E-Sand model for vertical flow
The density-pressure relationship for isothermal condi-
tions of a slightly compressible fluid may be expressed
as
p=poe(p
and
ap=a, . . . . . .(13)
CP
where c is the fluid compressibility. Thus,
P 0 9 0 -b dp
-k dp
-z--c--
A
P dx
pc dx
where q. is the volume rate of flow of a fluid of density
PO.
On integration,
WP I -02)
PO90 =
. . . . . (14) cLcL
If terms of cp of second and higher order are neglected,
the density can be expressed as
such that Eq. 14 reduces to
k4P I -P2)
40=
d .
The density-pressure relationship for isothermal condi-
tions of an ideal gas may be expressed as
P P PPb
-=- orp=-. . ,(15)
Ph Pb Pb
Thus,
tbqb b dp
-=-_-
A
P dx
where pb and qb are the density and volume rate of flow,
respectively, at the base pressure, ph.
Substituting for p
Pbqb k dp
-=--p-, . . . . . . . . . . . .
A
cl&
which on integration yields
(16)
kA P,*-Pz~
q=G . . . . . . . _. __
(17)
Pb
Define p as (p, +p2)12, and yP as the volume rate of
flow at p such that pqp =pbqh; then
k4PI -P2)
41, =
uL .,.................... (18)
Therefore, flow rates of ideal gases may be computed
from the equations for incompressible liquids so long as
the volume rate of flow is defined at the algebraic mean
pressure.
Vertical Flow. Fig. 26.18 shows three sandpacks in
which linear flow occurs in the vertical direction.
First consider Case 1 (Fig. 26. IS)--when the pressure
at the inlet and outlet are equal (free flow), such that only
the gravitational forces are driving the fluids. Given
dz
dp
S=Z, -=l, and -=O,
ds ds
the flow is then defined by
q&g. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.(19)
P
Next consider Case 2-the case of downward flow
when the driving head (difference in hydraulic head of
inlet and outlet) is h (Fig. 26.18). We know that
dz, -ah
z=l and z=-
L
Therefore, from Eq. 8,
. . . . . . . . . . .(20)
26-13
When the flow is upward and the driving head is h,
Case 3 (Fig. 26.18), and z is defined as positive
downward,
and
k
u=+-
P
then
kApgh
4=
. . .._..........................
(21)
Radial Flow. A radial-flow system, analogous to flow
into a wellbore, is idealized in Fig. 26.19. If flow is con-
sidered to occur only in the horizontal plane under
steady-state conditions, an equation of flow may be
derived from Darcys law to be
2akMp, -P,, 1
4=
) . . . .
(22)
~ ln rplr,,
where rI, is the radius at the external boundary at which
pc (pressure at the external boundary) is measured, and
r,v is the radius of the wellbore at which pa, (pressure at
the wellbore) is measured. All other terms are as defined
for linear flow.
Eq. 22 may be modified appropriately for the flow of
compressible fluids. The details of modifying this equa-
tion are omitted because they are essentially the same as
the ones used in the horizontal rectilinear flow systems.
After modification for variations in flowing volumes
with changing pressures, Eq. 22 becomes for slightly
compressible fluids
2xk&, -P,,,)
h = . ,_..................... (23)
cp In r,,lr,,.
where w is the mass rate of flow, g/s, or
2mWpp -pw)
40 =
a~, In TJr,,.,
where q. is defined at the pressure p. where the density
is po.
For ideal gases, Eq. 22 becomes
7rkh(p<,-~,,.~)
4/J =
. (24)
Wb ln r,,r,,
or
2dd-0, -p w)
911 =
, . . . . cL ln r,,ra, (25)
Fig. 26.19-Sand model for radial flow of fluids to central
wellbore.
Conversion of Units in Darcys Law
It is convenient in many applications of Darcys law to
introduce commonly used oilfield units. The following is
a summary of the more common equations with the con-
version factors to convert to oilfield terminology.
Linear-Flow Liquids (or Gases with Volume at Mean
Pressure). Rate, BID, is given by
q=1.1271
kA(p I -p21
. . . . . . . LLL
(26)
Rate, cu ft/D, is given by
q=6.3230
Wp I -~2)
. . . . . . . . . AL
(27)
Radial-Flow Liquids (or Gases with Volume at Mean
Pressure). Rate, liters per day, is given by
q=92.349x 103
Wp, -PM)
. . . .
p In r,/r,
(28)
Rate, cu ft/D, is
q-92.349x 103
kh(p, -p,v)
cc ln r,/r,, . .
(29)
Gases at Base-Pressure, p,,, and Average Flowing
Temperature, Tf. Linear flow rate, cu ft/D. is given by
where q,, is the volume rate of flow at the algebraic mean
pressure (p, +p,,,)/2.
3.1615T,,kA(p, -pz *)
Yb =
, . . . . . . . .
TfW&PlJ
(30)
Pe
Fig. 26.20-Linear flow-parallel combination of beds. Fig. 26.21~Radial flow-parallel combination of beds.
and radial flow rate, cu ft/D, is given by*
19.88T~kh(p,2 -p,,,?)
q/J= -
.,.............
(31)
Tfq.4,~p~ In rplr,,,
where k is in darcies; A is in sq ft; h is in ft; p , , p2, pp,
p ,,.. and Ph are in psia; p is in cp; L is in ft; and re and
T,, are in consistent units.
Since the previous equations describe the flow in the
medium, appropriate volume factors must be introduced
to account for changes in the fluids caused by any
decrease in pressure and temperature from that of the
medium to standard or stock-tank conditions.
Permeability Conversion Factors. Following is a list
of various unit conversions from darcy units to other
systems of units.
kL!E-
A(p/L)
1 darcy (d)= 1,000 millidarcies (md)
(cm3/s)cp
=
cm (atm/cm)
=9.869x 10 -
(cm3/s) cp
,,* (dynIrm*)
=9.869x lop9 cm2
=1.062x IO- sq ft
=7.324x lop5
(cu ft/sec)cp
sq ft(psi/ft)
(cu ft/sec)cp
=9.697x10p4
cm(cm water/cm)
=1.127
WD)cp
sq ft (psi/ft)
= 1.424 x 10 -
(gal/min)cp
sq ft (ft water/ft)
Flow Systems of Combinations of Beds
Consider the case where the flow system comprises
layers of porous rock separated from each other by in-
finitely thin, impermeable barriers as shown in Fig.
26.20. The average permeability k may be evaluated by
Eq. 32.
(32)
C hj
j=l
Fig. 26.21 shows that the same terms appear in the
radial-flow network as in the linear system. The only dif-
ference in the two systems is the manner of expressing
the length over which the pressure drop occurs. Because
all these terms are the same in each of the parallel layers,
an evaluation of the parallel radial system yields the
same solution as obtained in the linear case.
Example Problem 6-Average Permeability of Beds
in Parallel. What is the equivalent linear permeability of
four parallel beds having equal widths and lengths under
the following conditions?
Pay
Horizontal
Thickness Permeability
Bed
(ft) (md)
1 20 100
2 15 200
3 10 300
4 5 400
i: kihj
j&
j=l .
ihi
j=I
k= (100x20)+(200x15)+(300x 10)+(400x5)
20+15+10+5
10,000
=--200 md.
50
AP, AP, AP,
h
Fig. 26.22-Linear flow-series combination of beds.
Fig. 26.23-Radial flow-series combination of beds
Another possible combination for flow systems is to
have the beds of different permeabilities to be arranged
in series as shown in Fig. 26.22. In the case of linear
flow, the average series permeability may be evaluated
by Eq. 33.
Therefore,
j&
250+2.50+500+ 1,000
2.50 250 500 1,000
z+50+100+- 200
L
2,000
k= . . . . . . .(33) = 10+5+5+5
=80 md.
LJ lkj
J ==I
For a radial system,
The same reasoning can be used in the evaluation of
the average permeability for the radial system (Fig.
k=
log 2.000/0.5
26.23) so as to yield
log 25010.5 log 500/2so lop I .000/500 log 2.ooo11 .ooo
+ + +
25 SO loo 200
kc
In r,,/r,,.
In rjlr, -1
= h,
j&l
.I
Example Problem 7-Average Permeability of Beds
in Series. What is the equivalent permeability of four
beds in series, having equal formation thicknesses under
the following conditions?
Assume Bed 1 adjacent to the wellbore (1) for a linear
system and (2) for a radial system if the radius of the
penetrating wellbore is 6 in. and the radius of effective
drainage is 2,000 ft.
=30.4 md.
Permeability of Channels and Fractures
Only the matrix permeability has been discussed in the
analysis to this point. In some sand and carbonate reser-
voirs, the formation frequently contains solution chan-
nels and natural or artificial fractures. These channels
and fractures do not change the permeability of the
matrix but do change the effective permeability of the
flow network.
Circular Channel. Equating Darcys and Poiseuilles
equations for fluid flow in a tube, the permeability may
be expressed as a function of the tube radius.
k=;, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(35)
Length Horizontal
of Bed Permeability
_ @I
Bed
(md)
1 250 25
2 250 50
3 500 100
4 1,000 200
where k and r are in consistent units.
If r is in centimeters, then k in darcies is given by
r2
k=
8(9.869x 10 -9)
I 12.50X 106r,
where 9.869~ 1O-9 is a conversion factor from the
previous list. Then, if r is in inches,
For a linear system,
k= 12.50~ 106(2.54)*r2
k=
k Ljlkj
j=l
=80~10~r*=20~10~d*,
where d is the diameter of the opening in inches.
(PI -P2YL
(0)
~~~~~~
01 03
(p: -Op5:),2L
07 09
fb)
Fig. 26.24-Plol of experlmental results for calculation of permeability-(a) from k/F =qL/
VW, -P~)I; @I from ~~P=~~,P&L~P, 2 -P~*)I.
Therefore, the permeability of a circular opening
0.005 in. in radius is 2,000,OOO md.
Fracture. For flow throu
4
h slots of fine clearances and
unit width, Buckingham- reports that
12pL
Ap=-
h?
such that the permeability of the slot is given by
k=;. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(36)
When h is in centimeters and k in darcies. the
permeability of the slot is given by
k=
12(9.869x 10-9)
=84.4x 10h2,
and when h is in inches and k is in darcies, permeability
is given by
k=54.4 x 106h2,
so that the permeability of a fracture 0.01 in. in thickness
would be 5,440 darcies or 5,440,OOO md.
Physical Analogies to Darcys Law
Ohms law as commonly written is
where
I = current, A,
E = voltage drop, V, and
r = resistance of the circuit, Q,
but
L L
r=p- or r= -,
A OA
where
p = resistivity, Q-cm,
u = llp=conductivity,
L = length of flow path, cm, and
A = cross-sectional area of conductor, cmZ.
Therefore,
I&
PL
Comparing with Darcys law for a linear system,
,=!A%.
P L
Note that
4 = 1,
k 1
and
I* P
4 E
-
=-
L L
The Fourier heat equation may be written as
d=k,aF,
where
Q= rate of heat flow, Btulhr,
A = cross-sectional area, sq ft,
AT = temperature drop, F,
L = length of the conductor, ft, and
kh = thermal conductivity, Btu/(ft-hr-F).
A comparison with Darcys law indicates that
s=Q,
k
-=kl,, and
P
Ap AT
-=-
L L
Electrical and heat models (based on these analogies) of
rock and well systems frequently are used to solve fluid-
flow problems involving complex geometry.
Measurement of Permeability
The permeability of a porous medium may be deter-
mined from samples extracted from the formation or by
in-place testing. The procedures discussed in this section
pertain to the permeability determinations on small
samples of extracted media.
Two methods are used to evaluate the permeability of
cores. The method most used on clean, fairly uniform
formations uses small cylindrical samples called perm
plugs that are approximately % in. in diameter and 1 in.
in length. The second method uses full-diameter core
samples in lengths of 1 to 1% ft. The fluids used with
either method may be gas or any nonreactive liquid.
Perm-Plug Method. As core samples ordinarily contain
residual oil, water, and gas, it is necessary that the sam-
ple be subjected to preparation before the determination
of the permeability. The residual fluids normally are ex-
tracted by retorting or solvent extraction. The core is
dried before permeability measurements are taken. Air
commonly is used as the fluid in permeability
measurements. The requirement that the permeability be
determined for conditions of viscous flow is best
satisfied by obtaining data at several flow rates and plot-
ting results, as shown in Fig. 26.24, based on either Eq.
17 or 18. For conditions of viscous flow, the data should
plot a straight line, passing through the origin. Tur-
bulence is indicated by curvature of the plotted points.
The slope of the straight-line portion of the curve is equal
to k/p, from which the permeability may be computed.
To obtain k in darcies, 4 must be in cm3/s, A in cm, p 1
and p2 in atm. L in cm, and ~1 in cp.
A permeameter designed for the determination of the
permeability of rocks with either gas or liquid is il-
lustrated in Fig. 26.25. Data ordinarily are taken from
this device at only one flow rate. To assure conditions of
viscous flow, the lowest possible rate that can be
measured accurately is used.
Example Problem S-Permeability Measurement.
The following data were obtained during a routine
penneability test. Compute the permeability of this core.
1. Flow rate= 1,000 cm3 of air at 1 atm absolute and
70F in 500 seconds.
2. Pressure, downstream side of core= 1 atm absolute;
flowing temperature, 70F.
3. Viscosity of air at test temperature=0.02 cp.
4. Cross-sectional area of core=2.0 cm*.
5. Length of core=2 cm.
6. Pressure, upstream side of core= 1.45 atm
absolute.
PI VI fP?V2 =Pv
where 1 is upstream conditions and 2 is downstream con-
ditions, and
P +Pz
1.45+1
P=-
E-=1.225,
2 2
Fig. 26.25-Ruska universal permeameter.
and
1 x 1,000=1.225 V.
V=816 cm3,
_ v 816
-=1.63,
4=;=500
1.63x2x0.02
=
x 1,000
2 x0.45
=72.5 md.
Assuming that the data indicated were obtained, but
water was used as the flowing medium, compute the
permeability of the core. The viscosity of water at test
temperature was 1 .O cp.
v 1,000
4-50040
t
and
.-
k=!!kx,,)oO=
2X1X2
~ x 1,000=4,450 md.
A AP
2x0.45
Whole-Core Measurement. The core must be prepared
in the same manner as perm plugs. The core is then
mounted in special holding devices as shown in Fig.
26.26. The measurements required are the same as the
perm plugs but the calculations are slightly different.
AIR PRESSURE
-GASKET OPENING
END VIEW
TOP VIEW
SPLT -
TELOWMETER
GASKET1
Fig. 26.26-Clamp-type permeameter for large cores.
In the case of the clamp-type permeameter, the
geometry of the flow paths is complex, and an ap-
propriate shape factor must be applied to the data to com-
pute the permeability of the sample. The shape factor is a
function of the core diameter and the size of the gasket
opening. The shape factor affects the quantity L/A in the
previous equations.
Factors Affecting Permeability Measurements
In the techniques of permeability measurement previous-
ly discussed, certain precautions must be exercised to
obtain accurate results. When gas is being used as the
measuring fluid, corrections must be made for gas slip-
page. When liquid is the testing fluid, care must be taken
that it does not react with the solids in the core sample.
Also, corrections may be applied for the change in
permeability because of the reduction in confining
pressure on the sample.
Effect of Gas Slippage on
Permeability Measurements
Klinkenberg I4 has reported variations in permeability
determined by using gases as the flowing fluid from that
determined by using nonreactive liquids. These varia-
tions were ascribed to slippage, a phenomenon well
known with respect to gas flow in capillary tubes. The
phenomenon of gas slippage occurs when the diameter of
the capillary openings approaches the mean free path of
the gas.
Fig. 26.27 is a plot of the permeability of a porous
medium as determined at various mean pressures using
hydrogen, nitrogen, and carbon dioxide as the flowing
fluids. Note that for each gas a straight line is obtained
for the observed permeability as a function of the
reciprocal of the mean pressure of the test. The data ob-
tained with the lowest-molecular-weight gas yield the
straight line with the greatest slope, indicative of a
greater slippage effect. All the lines when extrapolated to
infinite mean pressure (l/j=O) intercept the permeabili-
ty axis at a common point. This point is designated as the
equivalent liquid permeability, kL. Klinkenberg and
b CARBON DIOXIDE
I / I I I I 1
0 2 0.4 0.6 0.8 1.0 12 1.4
1
J
RECIPROCAL MEAN PRESSURE (ATM)-
k!
Fig. 26.27-Permeability constant of core sample L to hydro-
gen, nitrogen, and COP at different pressures
(permeability constant to iso-octane, 2.55 md).
others established that the permeability of a porous
medium to a nonreactive homogeneous single-phase liq-
uid was equal to the equivalent liquid permeability.
The linear relationship between the observed
permeability and the reciprocal of mean pressure may be
expressed as follows.
kg
kL=-
1 +blp
=k, -ml, . . . . . . (37)
P
where
kL = permeability of the medium to a single
liquid phase completely filling the pores
of the medium,
k, = permeability of the medium to a gas
completely filling the pores of the
medium,
j = mean flowing pressure of the gas at which
k, was observed,
b = constant for a given gas in a given medium,
and
m = slope of the curve.
Reactive Fluids. While water commonly is considered
to be nonreactive in the ordinary sense, the occurrence of
swelling clays in many reservoir rock materials results in
waters being the most frequently occurring reactive liq-
uid in connection with permeability determinations.
Reactive liquids alter the internal geometry of the porous
medium. This phenomenon does not vitiate Darcys law
but rather results in a new porous medium, the
permeability of which is determined by the new internal
geometry.
While fresh water may cause the cementation material
in a core to swell because of hydration, it is a reversible
process. A highly saline water may be flowed through
the core and return the permeability to its original value.
The effect of water salinity on permeability is shown in
Table 24.3. 25
TABLE 26.3-EFFECT OF WATER SALINITY ON PERMEABILITY OF NATURAL
CORES (Grains/gal Chloride ion)
Field Zone
~ -
S 34
: 34 34
S 34
S 34
s 34
S 34
S 34
S 34
S 34
S 34
S 34
s 34
S 34
T 36
T 36
T
T ;:
T 36
T 36
k
a
4,080
24,800
40,100
39,700
12,000
4.850
22,800
34.800
27,000
12,500
13,600
7,640
11,100
6,500
2,630
3,340
2,640
3,360
4,020
3,090
k
1,000
k
500
k
300
1,445 1,380 1,290
11,800 10,600 10,000
23,000 18.600 15,300
20,400 17,600 17,300
5,450 4,550 4,600
1,910 1,430 925
13,600 6,150 4,010
23,600 7,800 5,460
21,000 15,400 13,100
4,750 2,800 1,680
5,160 4,640 4,200
1,788 1,840 2,010
4,250 2,520 1,500
2,380 2,080 1,585
2,180 2,140 2,080
2,820 2,730 2,700
2,040 1,920 1,860
2,500 2,400 2,340
3,180 2,900 2,860
2,080 1,900 1,750
k
200
1,190
9,000
13,800
17,100
4,510
736
3,490
5,220
12,900
973
4,150
2,540
866
1,230
2,150
2,690
1,860
2,340
2,820
1,630
k
100
k
w
885 17.2
7,400 147
8,200 270
14,300 1,680
3,280 167
326 5.0
1,970 19.5
3,860 9.9
10,900 1,030
157 2.4
2,790 197
2,020 119
180 6.2
794 4.1
2,010 1,960
2,490 2,460
1,860 1,550
2,280 2,060
2,650 2,460
1,490 1,040
Far example. ks means permeablhty lo air, hSoo
means permeablhty 10 iresh water
means permeablllty 10 500.gram/gal chloride solul~on, and kw
Care must be taken that laboratory permeability values
are corrected to liquid values obtained with water whose
salinity corresponds to formation water.
Overburden Pressure. When the core is removed from
the formation, all the confining forces are removed. The
rock matrix is permitted to expand in all directions, par-
tially changing the shapes of the fluid-flow paths inside
the core.
Compaction of the core caused by overburden pressure
may cause as much as a 25 % reduction in the permeabil i-
ty of various formations, as observed in Fig. 26.28.2h
Note that some formations are much more compressible
than others; thus, more data are required to develop em-
pirical correlations that will permit the correction of sur-
face permeability for overburden pressures.
Factors in Evaluation of Permeability From
Other Parameters
,Permeability. like porosity, is a variable that can be
measured for each rock sample. To aid in understanding
fluid flow in rocks and possibly to reduce the number of
measurements required on rocks, correlations among
porosity, permeability, surface area, pore size, and other
variables have been made. The reasoning behind some of
the correlations among porosity, permeability, and sur-
face area is presented here to enable the reader to gain
some understanding of the interrelation of the physical
properties of rocks. Although these relations are not
quantitative, they are indicative of the interdependence
of rock characteristics.
Use of Capillary Tubes for Flow Network. The
permeability of a tube derived from Darcys and
Poiseuilles equation is
k=f
x
If a porous system is conceived to be a bundle of
capillary tubes, then it can be shown that the permeabili-
ty of the medium depends on the pore-size distribution
and porosity. A flow network of tubes would be similar
to layers of different permeability in parallel, such that
the average permeability could be calculated by adapting
Eq. 32 to read
m
C kjAj
J =I
WI
. . . . . . . . ..I..............
(38)
J =l
where kj is the permeability of one capillary tube and Aj
is the area of flow represented by a bundle of tubes of
permeability ki.
lCURVE SOURCE PERMEABILITY POROSITY
-- nr rr.n.- . .^
--l?rruT
ti d 60
zt I t
3 ~WXIERN CALIFORNIA
335 25
:z
4
LOS ANGELES BASIN
2 CALIFORNIA
I10 22
y 50
I I
0 1000
5
2000 3000 4000 5000
OVERBURDEN PRESSURE-PSIG
Fig. 26.28-Change in permeability with overburden pressure.
26-20
The quantities k, and A, can be defined in terms of the Wyllie* derived the Kozeny equation from Poi-
radius of capillary tubes.
seuilles law by using a specified flow network. The
resulting permeability for this flow network is given by
A,=sn r
I I
and
Eq. 43.
, . . . . . . . . . . . . . . . . . . . . . .
(43)
7
k;+,
where
k = permeability of the porous medium,
F,Y = shape factor,
where n, is the number of tubes of radius r,.
5 A,=$A,,
J=l
L = length of the sample, and
L, = actual length of the flow path.
If
Lo 2
where 4 is the porosity of the flow network and A, is the
( > T
=T= tortuosity of the porous medium
total cross-sectional area of the flow network. By
substitution, Eq. 38 reduced to
and K, =F,Y7=Kozeny constant, then Wyllies equation
will reduce to the same form as Eq. 42.
j&f
,,1
c n/r, 3
j= I
111
( . . . . . . c ,l/r; *
j=l
(39)
where li is the average permeability of the tube bundle.
Note that the permeability of a bundle of tubes is a
function not only of the pore size but of the arrangement
or porosity of the system.
Consider a system that comprises a bundle of capillary
tubes of the same radii and length; k, the permeability,
may be written from Eq. 39 as
7
k=$. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(40)
The internal surface area per unit PV, A,,, may be de-
fined in terms of the tube radius by
A,,=?. ,.,.....,.,,.,,_...,..........,
.(41)
r
Combining Eqs. 40 and 41 gives the permeability as a
function of porosity and internal surface area. This func-
tion is
If l/K; is substituted for the constant l/z, the resulting
expression is Kozenys equation wherein K; is defined
as the Kozenys constant.
4
k=-
K,A, ~~~~..........,......___,,,,. (42)
Fluid Saturations
In the previous sections of this chapter, the storage and
the conduction capacity of a porous rock were discussed.
To the engineer there is yet another important factor to
be determined-i.e., the fluid content of the rock. In
most oil-bearing formations, it is believed that the rock
was completely saturated with water before the invasion
of the rock by petroleum. The oil will not displace all the
water from the pore space. Therefore, to determine the
quantity of hydrocarbons accumulated in a porous rock
formation, it is necessary to determine the fluid satura-
tion (oil, water, and gas) of the rock.
There are two approaches to the problem of determin-
ing the fluid saturations within a reservoir rock. The
direct approach is to measure. in the laboratory, the
saturations of selected rock samples recovered from the
parent formation. The indirect approach is to determine
the fluid saturation by measuring some related physical
property of the rock.
Factors Affecting Fluid Saturations of Cores
The core samples delivered to the laboratory for fluid-
saturation determinations are obtained from the ground
by either rotary, sidewall, or cable-tool coring. In all
cases, the fluid content of these samples has been altered
by two processes. First, in the case of rotary drilling, the
mud column exerts a greater pressure at the formation
wellbore surface than the fluid in the formation. The dif-
ferential pressure between the mud column and the for-
mation fluids causes mud and mud filtrate to invade the
formation, thus flushing the formation with mud and its
fyltrate. The filtrate displaces some of the original fluids.
This displacement process changes the original fluid
contents of the in-place rock. Second. as the sample is
brought to the surface, the confining pressure of the fluid
column is constantly decreasing. The reduction of
pressure permits the expansion of the entrapped water,
ail, and gas. Gas. having the greater coefficient of ex-
pansion, expels oil and water from the core. Thus, the
content of the core at the surface has been changed from
that which existed in the formation. Because the invasion
26-2 1
of the filtrate precedes the core bit, it is not possible to
use pressurized core barrels to obtain undisturbed
samples.
Drill cuttings, chips, or cores from cable-tool drilling
also have undergone definite physical changes. If little or
no fluid is maintained in the wellbore, the formation ad-
jacent to the well surface is depleted because of pressure
reduction. As chips are formed in the well, they may or
may not be invaded, depending on the fluids in the
wellbore and the physical properties of the rock. In all
probability, fluid will permeate this depleted sample,
resulting in flushing. Thus, even cable-tool cores
undergo the same two processes as was noted in the case
of rotary coring, although in reverse order.
Sidewall cores from either rotary- or cable-tool-drilled
holes are subjected to these same processes.
In an attempt to understand better the overall effect of
the physical changes that occur in the core because of
flushing and fluid expansion, Kennedy et al. 28 under-
took a study to simulate rotary-coring techniques. The
effects of both invasion and expansion because of
pressure reduction were measured.
Schematics of the changes in saturation resulting from
these two processes for oil- and water-based muds are
shown in Fig. 26.29. For the water-based mud, the
original displacing action of the water filtrate reduced the
oil saturation by approximately 14%. The expansion to
surface pressure displaced water and additional oil. The
final water saturation was greater than the water satura-
tion before coring. With oil-based mud, wherein the
filtrate is oil, the displacing action did not alter the initial
water saturations but did result in replacement of approx-
imately 20% of the initial oil. On pressure depletion, a
small fraction of the water was expelled, reducing the
water saturation from 49.1 to 47.7 % The oil saturation
was reduced by both processes from 50.9 to 26.7%.
Thus, even when high water saturations are involved, up
to approximately 50%, the water-saturation values ob-
tained with oil-based muds may be considered to be
representative of the initial water saturations in the reser-
voir. Hence, it is possible to obtain fairly representative
values of in-place water saturations by using oil-based
muds.
Attempts have been made to use tracers in the drilling
fluid to determine the amount of water in the core that is
caused by mud-filtrate invasion. The theory was that
mud filtrate displaced only oil. Thus. upon recovering
the core to the surface, the salt concentration of the core
water could be determined. Thus, if the salt concentra-
tion in the reservoir water and the tracer concentration in
the drilling fluid were known, the volume of filtrate and
reservoir water in the core could be calculated. Because a
large fraction of the initial reservoir water may have been
displaced by the invading filtrate, the tracer method
results in incorrect values for reservoir-water saturation.
To obtain realistic values of fluid saturation, it is
necessary to choose the proper drilling fluid or to use in-
direct methods of saturation determination.
Determination of Fluid Saturations from
Rock Samples
One of the most popular means of measuring fluid
saturations of cores is the retort method. This method
takes a small rock sample and heats the sample to
SATURATION AFTER
MUD FLUSHING BEFORE
PRE :SSURE REDUC TION
RESIDUAL
SbTUkATlON SATURATION
WATER BASE MUD
(a)
ORIGINAL
SATURATION
T---l
SATURATION AFTER
RESIDUAL
SATURATION
MUD FLUSHING BEFORE
PRESSURE REDUCTION
OIL BASE MUD
(b)
Fig. 26.29-Typical changes in saturations of cores flushed
with water-based and oil-based muds.
vaporize the water and the oil, which is condensed and
collected in a small receiving vessel. The retort method
has several disadvantages as far as commercial work is
concerned. The water of crystallization within the rock is
driven off, causing the water-recovery values to be too
high. The second error that occurs from retorting
samples is that the oil, when heated to high
temperatures, has a tendency to crack and coke. This
change of hydrocarbon molecules tends to decrease the
liquid volume. The fluid wetting characteristics of the
sample surface may be altered during the retorting proc-
ess as a result of the two previous factors. Before retorts
can be used, calibration curves must be prepared for
water and oils of various gravities to correct for losses
and other errors. These curves can be obtained by run-
ning blank runs (retorting known volumes of fluids of
known properties). The retort is a rapid method for deter-
mining fluid saturations and, if the corrections are used,
yields satisfactory results. It gives both water and oil
volumes such that the oil and water saturations can be
calculated from the following equations.
S&,
P
S, +,
P
26-22
and
S,=l-S,-S,,
where
s, =
s, =
s, =
v, =
VP =
v, =
water saturation,
oil saturation,
gas saturation,
The other
extraction
Fig. 26.30Laboratory layout for performlng routine core analysis
water volume, cm
3
,
pore volume, cm
3
, and
oil volume, cm
3
.
method of determining fluid saturation is by
with a solvent. Extraction may be accom-
plished by a modified ASTM distillation method or a
centrifuge method. In the standard distillation test, the
core is placed such that a vapor of either toluene, pen-
tane, octane, or naphtha rises through the core. This
process leaches out the oil and water in the core. The
water and extracting fluid are condensed and collected in
a graduated receiving tube. The water settles to the bot-
tom of the receiving tube because of its greater density,
and the extracting fluid refluxes over the core and into
the main heating vessel. The process is continued until
no more water is collected in the receiving tube. The
water saturation may be determined directly by
S,,=v,;.
VP
The oil saturation is an indirect determination. The oil
saturation as a fraction of PV is given by
so=
w,, - w,, - w,
VP-P* )
where
W II =
weight of wet core, g,
Wtd = weight of dry core, g,
W, = weight of water, g,
VP = PV, cm
3
, and
PO
= density of oil, g/cm
3
.
The gas saturation is obtained in the same manner as the
retort.
Another method of determining water saturation is to
use a centrifuge. A solvent is injected into the centrifuge
just off center. Because of centrifugal force, it is thrown
to the outer radii and forced to pass through the core
sample. The outflow fluid is trapped and the quantity of
water in the core is determined. The use of the centrifuge
provides a very rapid method because of the high forces
that can be applied. In both extraction methods, at the
same time that the water content is determined, the core
is cleaned in preparation for the other measurements
such as porosity and permeability.
There is another procedure for saturation determina-
tion that is used in conjunction with either of the extrac-
tion methods. The core as received from the well is
placed in a modified mercury porosimeter in which the
BV and gas volume are measured. The volume of water
is determined by one of the extraction methods. The
fluid saturations can be calculated from these data.
In connection with all procedures for determination of
fluid content, a value of PV must be established in order
that fluid saturations may be expressed as percent of PV.
Any of the porosity procedures previously described may
be used. Also, the BV and gas volume determined from
the mercury porosimeter may be combined with the oil
and water volumes obtained from the retort to calculate
PV, porosity, and fluid saturations.
Porosity, permeability, and fluid-saturation determina-
tions are the measurements commonly reported in
routine core analysis. A laboratory equipped for such
determinations is shown in Fig. 26.30.
Interstitial Water Saturations
Essentially, three methods are available to the reservoir
engineer for the determination of interstitial water satura-
tions. These methods are (1) determination from cores
cut with oil-based muds, (2) determination from
capillary-pressure data, and (3) calculation from electric-
log analysis (see Chap. 49).
Oil-Based Mud. The obtaining of water saturations by
using oil-based muds has been discussed. A correlation
between water saturation and air permeability for cores
obtained with oil-based muds is shown in Fig. 26.31.
29
26-23
A general trend of increasing water saturation with
decreasing permeability is indicated. It is accepted from
field and experimental evidence that the water content
determined from cores cut with oil-based mud reflects
closely the water saturation as it exists in a reservoir, ex-
cept in transition zones where some of the interstitial
water is replaced by filtrate or displaced by gas
expansion.
Fig. 26.32 shows permeability/interstitial-water rela-
tionships reported in the literature for a number of fields
and areas. There is no general correlation applicable to
all fields; however, an approximately linear correlation
between interstitial water and the logarithm of
permeability exists for each individual field. The general
trend of the correlation is decreasing interstitial water
with increasing permeability.
22 4 6 4 6
II t tic: t t ttn
I 20 40 60 100 200 400
AIR PERMEABILITY, MILLIDARCYS
Fig. 26.31--Relation of air permeability to the water content of
the South Coles Levee cores.
Capillary Pressure. Capillary pressure may be thought
of as a force per unit area resulting from the interaction
of surface forces and the geometry of the medium in
which they exist. Capillary pressure for a capillary tube
is defined in terms of the inter-facial tension between the
fluids, a. the angle of contact of the interface of these
two fluids and the tube. 0(., and the radius of the tube,
VI.
vature (these radii touch at only one point), r, and r7 _
and the interfacial tension of the fluids. This relationship
is given in Eq. 4.5.
(45)
This relationship is expressed in Eq. 44.
P,. =
20 cos 8,.
, ,,..,.....................
(44)
rt
Comparing Eq. 4.5 with the equation for capillary
pressure as determined by the capillary-tube method, it is
found that the mean radius U is defined by
where the angle 0,. is measured through the more dense
fluid.
I 1 I 2 cos 0
_=-+-=-.--.-------'
r
rt r2 rf
In a packing of spheres, the capillary pressure is ex-
pressed in terms of any two perpendicular radii of cur-
rir2 rf
y=-=-.---
rl +r2
2cos*,.. . . . . . . . . . . . . . .
(46)
0 IO 20 30 40 50 60 70 80 90 100
LEGEND
I = HAWKINS
2: MAGNOLIA
3= WASHINGTON
4=ELK BASIN
5= RANGELY
6: CREOLE
7= SYNTHETIC ALUNDUM
8: LAKE ST. JOHN
C
9= LOUISIANA GULF COAST
MIOCENE AGE-WELL A
IO=DITTO-WELLS BAND C
fil = NORTH BELRIDGE-CALIFORNIA
CORE ANALYSIS DATA
% INTERSTITIAL WATER
Fig. 26.32-Interstitial water vs. permeability relationships.
MEASURING
APPARATUS
Fig. 26.33-Schematic of porous-diaphragm method of
capillary pressure.
It is practically impossible to measure the values of rl
and t-2 ; hence, they generally are referred to by the mean
radius of curvature and empirically determined from
other measurements on a porous medium.
The distribution of the liquid in a porous system
depends on the wetting characteristics. It is necessary to
determine which is the wetting fluid so as to ascertain
which fluid occupies the small pore spaces. From pack-
ings of spheres, the wetting-phase distribution within a
porous system has been described as either funicular or
pendular in nature. In funicular distribution, the wetting
phase is continuous, completely covering the surface of
the solid. The pendular ring is a state of saturation in
which the wetting phase is not continuous and the
nonwetting phase is in contact with some of the solid sur-
face. The wetting phase occupies the smaller interstices.
As the wetting-phase saturation progresses from the
funicular to the pendular-ring distribution, the volume of
the wetting phase decreases and the mean radius of cur-
vature or the values of rl and r2 tend to decrease in
magnitude. Referring to Eq. 46, we see that if r I and r2
decrease in size, the magnitude of the capillary pressure
would have to increase in value. Since r , and r2 can be
related to the wetting-phase saturation, it is possible to
express the capillary pressure as a function of fluid
saturation when two immiscible phases are within the
porous matrix.
Laboratory Measurements of Capillary Pressure
Essentially, five methods of measuring capillary
pressure on small core samples are used. These methods
are (1) the desatumtion or displacement process, through
a porous diaphragm or membrane (restored-state method
of Welge3), (2) the mercury-injection method, (3) the
centrifuge or centrifugal method, (4) the dynamic-
capillary-pressure method, and (5) the evaporation
method. *
Porous Diaphragm. The first of these, illustrated in
Fig. 26.33, is the displacement or diaphragm method.
The essential requirement of the diaphragm method is a
Since the method IS seldom used today, it WIII not be dwzussed The procedure con-
slsts ot continuously monitoring the decrease m weight caused by evaporabon of a
core sample lnitlally 100% saturated wlh a wetltng fluld. See Messner, E S :
Intetsbtial Water Determmation By An Evaporatmn Method. Trans, AIME (1951)
192,269-74
permeable membrane of uniform pore-size distribution
containing pores of such size that the selected displacing
fluid will not penetrate the diaphragm when the pressures
applied to the displacing phase are below some selected
maximum pressure of investigation. Various materials
including fritted glass, porcelain, and cellophane have
been used successfully as diaphragms. Pressure applied
to the assembly is increased by small increments. The
core is allowed to approach a state of static equilibrium
at each pressure level. The saturation of the core is
calculated at each point defining the capillary-pressure
curve. Any combination of fluids may be used: gas, oil,
and/or water. Although most determinations of capillary
pressure by the diaphragm method are drainage tests, by
suitable modifications imbibition curves similar to
Leveretts may be obtained.
Mercury Injection. The mercury-capillary-pressure ap-
paratus was developed to accelerate the determination of
the capillary-pressure/saturation relationship. Mercury is
normally a nonwetting fluid. The core sample is inserted
in the mercury chamber and evacuated. Mercury is
forced into the core under pressure. The volume of mer-
cury
injected at each pressure determines the
nonwetting-phase saturation. This procedure is con-
tinued until the core sample is filled with mercury or the
injection pressure reaches some predetermined value.
Two important advantages are gained by this method: (1)
the time for determination is reduced to a few minutes,
and (2) the range of pressure investigation is increased
because the limitation of the diaphragms properties is
removed. Disadvantages are the difference in wetting
properties and permanent loss of the core sample.
Centrifuge Method. A third method for determining
capillary properties of reservoir rocks is the centrifuge
method.3 The high accelerations in the centrifuge in-
crease the field of force on the fluids, in effect subjecting
the core to an increased gravitational force. By rotating
the sample at various constant speeds, a complete
capillary-pressure curve may be obtained. The speed of
rotation is converted into force units in the center of the
core sample, and the fluid saturation is read visually by
the operator. The advantage of the method is the in-
creased speed of obtaining the data. A complete curve
may be established in a few hours, while the diaphragm
method requires days.
Dynamic Method. Brown3* reported the results of
determining capillary-pressure/saturation curves by a
dynamic method. Simultaneous steady-state flow of two
fluids is established in the core. By the use of special
wetted disks that permitted hydraulic pressure transmis-
sion of only the selected fluid phase, the difference in the
resulting measured pressures of the two fluids in the core
is the capillary pressure. The saturation is varied by
regulating the quantity of each fluid entering the core.
Thus, it is possible to obtain a complete capillary-
pressure curve.
Comparison of Methods of Measurement
Intuitively, it appears that the diaphragm method
(restored state) is superior in that oil and water are used;
therefore, actual wetting conditions are more nearly ap-
proached. Hence, the diaphragm method is used as the
standard to which all other methods are compared. The
mercury-injection test data must be corrected for wetting
conditions before they can be compared with results
from the restored-state method. If it is assumed that the
mean curvature of an interface in rock is a unique func-
tion of fluid saturation, then the ratio of mercury to water
capillary-pressure data is given by
+-+f+6.57. . . . . (47)
M 0,.
DEPTH FEET BELOW SEA-LEVEL
where urn is the surface tension of mercury and u, is the
surface tension of water. Experimentation has shown the
ratio to vary between 5.8 (for limestones) and 7.5 (for
sandstones). Thus, no conversion factor can be defined
Fig. 26.34-Comparison of water saturation from capillary
pressure and electric log.
that will apply to all rocks.
Good agreement of centrifuge data with those from the
diaphragm method was reported by Slobod. 31 Unlike the
mercury-injection method, there is no need of conver-
sion factors to correct for wetting properties. The same
fluids are used in the centrifugal and diaphragm
methods.
Excellent correlation was obtained by Brown3 be-
tween the diaphragm and dynamic methods. The
dynamic data were obtained by simultaneous steady flow
of oil and gas through the porous sample at a predeter-
mined level of pressure difference between the fluids.
Care was taken to maintain uniform saturations
throughout the core as well as to conduct the test such
that a close correspondence to drainage conditions
existed.
If capillary-pressure data are to be used for determin-
ing fluid saturations, the values obtained should be com-
parable with those of other methods. Water distributions,
as determined from electric logs, and capillary-pressure
data are normally in good agreement. A comparison of
these methods is shown in Fig. 26.34. 33 Shown also is
the approximate position of the gas/oil contact as deter-
mined from other test data. In the gas-bearing portion of
the formation there is no significant variation in water
saturation with depth or method of determination.
However, in a thin oil zone, such as that shown in Fig.
26.34. there is a significant variation in the water satura-
tion with depth. Variations in water saturations with
depth within an oil zone must be taken into account to
determine accurately average reservoir interstitial-water
saturations.
Water Saturation from Capillary-Pressure Data. In
oilfield terms, the capillary pressure may be stated as
gas/water or an oil/water system that normally does not
have the same physical properties as the reservoir water,
oil, and gas.
Essentially two techniques, differing only in the initial
assumptions, are available for correcting laboratory
capillary-pressure data to reservoir conditions.
p,..R=
uw><, cos 0,.,,
uwg cos 0,.,
p,., L
or
P,..R=~P& . . . . . . (49)
(JL
where
fJ wo
= interfacial tension water/oil,
UWR
= interfacial tension water/gas,
Ocwo = water/oil contact angle,
e,.,,
= water/gas contact angle,
subscript
R = reservoir conditions, and
subscript
L = laboratory conditions.
Since the interfacial tensions enter as a ratio, pressure in
any consistent units may be used together with the inter-
facial tension in dynes/cm.
Averaging Capillary-Pressure Data. Two methods
have been proposed for correlating capillary-pressure
data of similar geologic formations. The first correlating
procedure is a dimensionless grouping of the physical
properties of the rock and the saturating fluids. This
function is called a J function34 and is expressed as
P,.=+(p, -p2), ....................... (48)
, ...................... .(SO)
where h is in feet and p, and p2 are the densities of
Fluids 1 and 2, respectively, in lbmicu ft at the condi- where
tions of the capillary pressure. S,,, = water saturation, fraction of PV,
P,. = capillary pressure, dyne/cm*,
Converting Laboratory Data. To use laboratory u = interfacial tension, dyne/cm,
capillary-pressure data, it is necessary to convert them to k = permeability, cm*, and
reservoir conditions. Laboratory data are obtained with a
C#I = fractional porosity.
90
The results of the statistical correlation previously
81 z
discussed applied to the capillary-pressure data presented
72 I+
in Fig. 26.35 are shown in Fig. 26.36. The reader should
2
note the linearity of the curves for each value of capillary
63 3
g
pressure and the tendency of all capillary-pressure curves
54 a i:
to converge at high permeability values. This behavior is
3;
what normally would be expected because of the larger
45 2 ,o capillaries associated with high permeabilities.
732
36
gs
To convert capillary-pressure saturation data to height
2, 2 2
saturation, it is necessary only to rearrange the terms in
3-
&
Eq. 48 so as to solve for the height instead of the
I8 z
capillary pressure-i.e.,
3
P,. x 144
20 30 40 50 60 70 80 90 IO8
h,,=-, . . . . . . . . (52)
PW -PO
WATER SATURATION,%
RESERVOIR FLUID DISTRIBUTION CURVES
where
Fig. 26.35~Series of capillary-pressure curves as a function of
hh = height above the free-water surface, ft,
permeability.
P II
= density of water at reservoir conditions,
lbm/cu ft,
PO
= density of oil at reservoir conditions,
lbm/cu ft, and
Some authors alter the above expression by including the
cos f3(. (where 8,. is the contact angle) as follows.
P,. = capillary pressure at some particular satura-
tion for reservoir conditions (it must be
converted from laboratory data first, psi).
Example Problem 9-Calculation of Saturation Plane
From Laboratory Capillary-Pressure Data. If
The J function originally was proposed as a means of
converting all capillary-pressure data to a universal
curve. There exist significant differences in correlation
of the J function with water saturation from formation to
formation such that no universal curve may be obtained,
but the / function may be used to correlate the data from
one formation.
The second method of evaluating capillary-pressure
data is to analyze a number of representative samples and
treat the data statistically to derive correlations that,
together with the porosity and permeability distribution
data, may be used to compute the interstitial-water
saturations for a field. A first approximation for the cor-
relation of capillary-pressure data is to plot water satura-
tion against the logarithm of permeability for constant
values of capillary pressure. A straight line may be fitted
to the data for each value of capillary pressure, and
average-capillary-pressure curves may be computed
from permeability-distribution data for the field. The
resulting straight-line equation takes the general form of
Fluid-distribution curves are reported for several values
of permeability, ranging from 10 to 900 md in Fig.
26.35. These data also may be considered to be
capillary-pressure curves. The ordinate on the right
reflects values of capillary pressure determined by
displacing water with air in the laboratory. The ordinates
on the left include the corresponding oil/water capillary
pressure that would exist at reservoir conditions and the
fluid distribution with height above the free-water
surface.
P<.,L = 18 psi for S,.=O.35,
CJ 110
= 24 dynes.
P M
= 68 lbmlcu ft,
is 8
= 72 dynes, and
PO
= 53 lbm/cu ft,
then, from Eq. 49,
P(.,R = 18(24/72)= 18/3=6 psi,
and
Pc,R x 144 6 x 144
h= =-=58 ft.
P II -PO
68-53
Thus, a water saturation of 35 % exists at a height of 58 ft
above the free-water surface.
To calculate the fluid saturation in the gas zone, it is
necessary to consider all three phases: oil, water, and
gas. If all three phases are continuous, it can be shown
that
where
PC., R
= capillary pressure at a given height above
the free-water surface determined by
using water and gas,
PC, H%)
the free-water surface, using water and
oil, and
pC.O,q
the free-oil surface, using oil and gas.
If the wetting phase becomes discontinuous, then the
wetting-phase saturation takes on a minimum value, and,
at all heights above the point of discontinuity, the
wetting-phase saturation cannot be less than this
minimum value. It is then possible to calculate the fluid
saturations above the free-oil surface by the following
relations.
1. S,. at h is calculated using oil and water as the con-
tinuous phases.
2. SL at h is calculated using oil and gas as the con-
tinuous phases and height denoted by the free-oil
surface.
3. S,=l-SL andS,=SL-S,,
where SL is the total liquid saturation, oil plus water,
fraction.
Example Problem lo-Calculation of Water and Oil
Saturation in Gas Zone From Capillary-Pressure
Data. Let oil-zone thickness, h,, equal 70 ft and
u R
= 72 dynes,
(JOf
= 50 dynes,
(J I,0
= 25 dynes,
P II
= 68 lbm/cu ft
Pn
= 7 Ibm/cu ft, and
PO
= 53 lbm/cu ft.
From Fig. 26.28 for a 900-md sample, let
PI,,L = 54 psi by the method illustrated in
Example Problem 9,
Pc,R = 18 psi,
h fi = height above free-water level= 120 ft, and
S,,. = 16% at a height of 70 ft or greater (read
from curve).
Since the oil zone is only 70 ft thick, the height of 120
ft above the free-water surface must be at least 50 ft into
the gas-saturated zone. The first step is to calculate the
total fluid saturation, SL, using gas and oil as the con-
tinuous phases.
h,,,=h,,-h,=120-70=50 ft;
hj,
Pc,.R = IJ~(PI, -P,h
where h&, is the height above free-oil level, ft.
P<,.R =
50/144x(53-7)=501144x46.
P<,.R
= 15.97,
and
P,.,[, = Pr-,R~=15.97~72!50=23 psi.
0 og
From Fig. 26.35, for a laboratory capillary pressure of
23 psi and a permeability of 900 md, the total wetting
saturation, SL, equals 18 % Therefore.
S,,=SL-S,,,=18-16=2X
)- IOOC
700
500
400
300
__
u-l
$ 200
Q:
8
3
2 IOC
z
2 7c
s
2 50
% 40
3c
2c
I-
)-
I-
I-
IO-
0
-
-
-
-
-
-
-
-
-
-
-
20 40 60 80
WATER SATURATION (%)
Fig. 26.36-Correlation of water saturation with permeability
for various capillary pressures.
and
It must be understood that the relationships used in Ex-
ample Problem 10 for calculating the fluid saturations in
the gas zone were based on continuity of all three phases.
Since this is not normally the case, it might be expected
that saturations somewhat different from the calculated
values exist. Because the capillary pressure for a discon-
tinuous phase could vary from pore to pore, it is impossi-
ble to ascertain the exact relationships that should exist.
Hence, the preceding method of calculating fluid
distributions is not exact but is usually as accurate as the
data available for making the computation.
Electrical Conductivity of
Fluid-Saturated Rocks
Porous rocks comprise an aggregate of minerals, rock
fragments, and void space. The solids, with the excep-
tion of certain clay minerals, are nonconductors of elec-
tricity. The electrical properties of a rock depend on the
geometry of the voids and the fluids that fill the voids.
The fluids of interest in petroleum reservoirs are oil, gas,
and water. Oil and gas are nonconductors. Water is a
26-28
Fig. 26.37-Core-sample resistivity cell
The second fundamental notion of electrical properties
of porous rock is the resistivity index, IK, which is de-
fined as
(55)
where R, is the true resistivity of the rock system at some
particular value of water saturation and Ra is as
previously defined.
Three idealized representations have been introduced
in the literature from which the formation resistivity fac-
tor, FR, and the resistivity index , f~, have been related
to the porosity, 4, and the rock tortuosity, 7.
From Wyllies2 analysis the relations are
F&
4
and
IR =7(,--r-,
SW
conductor when it contains dissolved salts. Current is
conducted in water by movement of ions and therefore
may be termed electrolytic conduction. The resistivity of
a material is the reciprocal of conductivity and common-
ly is used to define the ability of a material to conduct
current. The resistivity of a material is defined by the
F,=d-
following equation.
4
and
where T,, is the effective rock tortuosity at some water
saturation.
Cornell and Katz3 presented an analysis of a slightly
different model. The relationships developed are as
follows.
p=*A. ._.,.,.,,...,...,,.,,_,....,,.,, (53)
L
JR=&&
se
where
p = resistivity.
I = resistance,
A = cross-sectional area of the conductor, and
L = length of the conductor,
Wyllie and Gardner 37 later presented an analysis
based on a probability theory from which the following
relationships were obtained.
FK=;
For electrolytes, p is commonly reported in Q-cm, and
r is expressed in ohms, A in cm, and L in cm. In the
study of the resistivity of soils and rocks. it has been
found that the resistivity may be expressed more conve-
niently in n-m. To convert to 9-m from O-cm, divide the
resistivity in Q-cm by 100. In oilfield practice. the
resistivity in Q-m commonly is represented by the sym-
bol R with an appropriate subscript to define the condi-
tion to which R applies.
Fundamental Concepts
The definition of electrical formation resistivity factor is
perhaps the most fundamental concept in considering
electrical properties of rocks. The formation resistivity
factor as defined by Archie is
4
and
I
IR=--
s,,.
From the analysis of the electrical properties of the
foregoing idealized pore models, general relationships
between electrical properties and other physical proper-
ties of the rock may be deduced. The formation resistivi-
ty factor has been shown to be some function of the
porosity and the internal geometry of the rock system. In
particular, the formation resistivity factor may be ex-
pressed in the following form.
FK=I@ -I, (56)
F/$5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(54)
where K is some function of the tortuosity and VI is a
R \I
function of the number of reductions in pore-opening
sizes or closed channels. It is suggested that K should be
where R. is the resistivity of the rock when saturated I or greater. The value of rn has been shown from theory
with water having a rcsistivity of R,, to range from I to 2.
Both the formation resistivity factor, F,, and the
resistivity index. IR, depend on ratios of path length or
tortuosities. Therefore, to compute the formation
resistivity factor or resistivity index from the relations
derived from models, it is necessary to determine the
electrical tortuosity. Since direct measurement of the
path length is impossible, reliance has been placed
primarily on empirical correlations based on laboratory
measurements. Winsauer et al. 3X devised a method of
determining tortuosity by transit time of ions flowing
through a rock under a potential difference. The data ob-
tained were correlated with the product FR+. The
resulting correlation of formation resistivity factor,
porosity, and tortuosity is given in Eq. 57.
1.67
=F,&
where L, is the actual length of flow path, and L is the
length of the sample, or
$ IO
II: 0
k-
6
T.~? =F&. . . . . . .(57)
4 1 HUMBLE RELATION F= 5 ,t-j&tt-
The deviation from the theory is believed to be an indica-
tion of the greater complexity of the actual pore system
than that of the models on which the theory was based.
Archie suggested that the formation resistivity factor
could be correlated with porosity and have the form
FR=c#-, . . . ., ._.. . . .(58)
where 4 is the fractional porosity and m is the cementa-
tion factor. Archie further reported that the cementation
factor probably ranged from 1.8 to 2.0 for consolidated
sandstones and, for clean unconsolidated sands, was
about 1.3.
Measurement of Electrical Resistivity of Rocks
Laboratory measurements of electrical properties of
rocks have been made with a variety of devices. The
measurements require a knowledge of the dimensions of
the rock, the fluid saturation of the rock, the resistivity of
the water contained in the rock, and a suitable resistivity
cell in which to test the samples. A simple cell is shown
in Fig. 26.37. 39 A sample cut to suitable size is placed
in the cell and clamped between electrodes A. Current is
then passed through the sample and the potential drop
observed. The resistance of the sample is computed from
Ohms law,
E
r=-.
I
and R (the resistivity) is computed from
R,A
L
where A is the cross-sectional area of the sample and L is
the length of the sample. The saturation conditions of the
test may be established at known values before measure-
ment or determined by an extraction procedure after
measurement.
--
5 60
5
2 40
zi
5 20
F
TIXIER S RELATION
FOR UNCONSOLIDATED
2p-M
CONSOLIDATED FORMATIONS
1,
2 4 6 8 IO 20 40 60 80
POROSITY, =/a
Fig. 26.38-Comparison of various formation resistivity factor
correlations.
Empirical Correlation of
Electrical Properties
Archie, as previously mentioned, reported the results of
correlating laboratory measurements of formation factor
with porosity. He expressed his results in the form
FRz4-m.
Winsauer et al. reported a similar relationship based
on correlations of data from a large number of sandstone
cores. This equation, commonly referred to as the Hum-
ble relation, is
FR=0.62q!-2.5. . .(59)
In discussing the theory of the formation resistivity
factor, it was stated that K should be greater than 1 and
that m should be 2 or less. At this time, the discrepancy
between theory and experiment must be attributed to the
possible effect of conducting solids.
Improved correlations should result from considering
other parameters, such as permeability, as variables in
the relations.
A comparison of suggested relationships between
porosity and the formation resistivity factor is shown in
Fig. 26.38.
A COMPARABLE CLEAN SAND I
4
!,
1.0 IO 100
R,-WATER RESISTIVITY, OHM-METERS
Fig. 26.39-Effect of Interstitial clay on formation resistivity
factors.
Effect of Conductive Solids
Investigations by Wyllie4 Indicate that clays contribute
substantially to the conductivity of a rock when the rock
is saturated with a low-conductivity water. The effect of
water resistivity on the formation resistivity factor for
sands containing clay minerals is shown in Fig. 26.39.
The formation resistivity factor for a comparable clean
(clay-free) sand is a constant. The formation resistivity
factor for the clayey sand increases with decreasing
water resistivity and approaches a constant value at a
water resistivity of about 0.1 Q-m. Wyllie proposed that
the observed effect of clay minerals was similar to hav-
ing two electric circuits in parallel-the conducting clay
minerals and the water-filled pores. Thus,
Ro.\/z
1 1 1
FRO=- and -==++
FRR,,, \ . (60)
R, Roslr R,./
where
FRO
= apparent formation resistivity factor,
Rosh
= resistivity of a shaly sand when 100%
saturated with water of resistivity R,,.,
R,., = resistivity caused by the clay minerals,
R,,. = resistivity caused by the distributed water,
and
FR = true formation resistivity factor of the rock
(i.e., the constant value of formation fac-
tor approached when the rock contains
low-resistivity water).
The data presented in Fig. 26.40 represent graphically
the confirmation of the relationship expressed in Eq. 60.
The graphs were plotted by deWitte@ from data
presented by Hill and Milburn.3 The plots are linear
and are of the general form
1 1
-=m-+b, . . . . . . . . . . . . . . . . . . . (61)
Rosh R,i
where m is the slope of the line and h is the intercept.
Comparing Eq. 60 with Eq. 61, note that m= ~/FR and
b= l/R,.[. The curve labeled Suite 1 No. 40 indicates a
clean sand because the line passed through the origin,
thus having a zero intercept b= l/R,.,=O. Then
l/Rosh =m(l/R,,)=(l/FRR,,), or Ro=F,R,,.. The re-
maining samples are from shaly sands, which have a
finite conductivity of the clay minerals, as indicated by
the intercepts of the lines. The linearity of the plots in-
dicates that l/R,, is a constant independent of R,,. This
phenomenon may be explained in terms of the ions ab-
sorbed on the clay. When the clay is hydrated, the ab-
sorbed ions form an ionic conducting path, which is
closely bound to the clay. The number of absorbed ions
is apparently little changed by the salt concentration of
the interstitial water.
Eq. 60 may be rearranged to express the apparent for-
mation resistivity factor in terms of R,., and FRR, .
K
O = R,. +R,.,/FR
and
FRY =
&I
R,, +(R,.,IF,)
As R,,-+O,
I
-+
Rw
Fig. 26.40-Water-saturated rock conductivity plotted against
water conductivity yields these measurements:
(A) Suite 1 No. 40; (X) Suite 1 No. 21; (0) Suite 1
No. 4; (0) Suite 2 No. 13: (II!) Suite 6 No. 2.
RCl
lim FRY =-
R,IIFR
=FR.
Therefore, FRY approaches FR as a limit as R, becomes
small. This behavior was observed in Fig. 26.39.
Hill and Milbum43 evaluated 450 samples from both
sandstone and limestone formations. The formation
resistivity factor was determined at a water resistivity of
0.01 R-m, a value at which the apparent formation
resistivity factor, FRa, approaches the formation
resistivity factor, FR. They designated the formation
resistivity factor as FR,O,oI The data were fitted by the
method of least squares to yield
F R.0,0,=1.4+-.78. . . . . ._ .(62)
This equation conforms to the theory previously dis-
cussed. They also fitted the data with K in Eq. 56
restricted to a value of 1. This yielded FR,P,o, =$ - 93,
which corresponds closely to Archies onginal expres-
sion F=$ -,
In summary, Eqs. 58 (with m=2.0) and 59 have been
used widely to represent the relation between formation
resistivity factor and porosity. Both equations yield
results satisfactory for most engineering purposes.
However, we propose that Eq. 62 be considered as more
valid because the data were taken to minimize the effect
of clays. The selection of a particular relation should be
based on independent observations on the formations or
formations of interest in a given geologic province.
Resistivity of Partially Water-Saturated Rocks
A rock containing both water and hydrocarbon has a
higher resistivity than when fully saturated with water.
The resistivity of partially water-saturated rocks has been
shown to be a function of the water saturation, S,,. . From
theoretical developments, the following generalization
may be drawn.
IR=KS,,/, . . . (63)
where 1~ = R,/Ro, the resistivity index; K is some func-
tion of tortuosity; and n is the saturation exponent.
Archie compiled and correlated experimental data
from various sources from which he suggested that the
data could be represented by
IR =S,,. G. . . . (64)
Wyllie confirmed the suggested relationship for clean
sands but found that the presence of clays [conductive
solids) altered the relationship. A comparison of Ar-
chics relationship and that for a core containing conduc-
tive solids is shown in Fig. 26.41. The change in the
relationship depends on both the amount of clays and the
water resistivity. Therefore. a general correlation for
sands containing conductive solids is not available,
although dewitte has proposed a method of using Eq.
64 for evaluation of shaly sands.
Use of Electrical Parameters in
Characterizing Porous Media
In the section on permeability, the Kozeny equation was
developed as follows.
4
k=- F,A,7, . . . ..~~~~....
(65)
loo
\
60
\
40
I \ I
z 20
2- WITHOUT CONDUCTIVE
k
; IO
m
56
ii
4
SOLIDS
2
I.01 .02 .04 08 .20 40 .80
WATER SATURATION
Fig. 26.41-Effect of conductive solids on the reslstivity-index-
vs.-saturation relationship in Stevens sandstone
core.
where
k = permeability,
C#I = porosity, fraction,
F, = shape factor,
A,v = internal surface area/unit PV, and
r = Kozeny tortuosity.
7 has been shown to be a function of FR+i.e.,
7= (FR#)-, where x ranges between 1 and 2. If the inter-
nal surface area is expressed in terms of the mean
hydraulic radius, rH, by
A,& . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(66)
rH
the general form of the relationship may be stated as
or
@H) 2
k
=F. (FR)X
& . . . . . . . . . . . . . . . . . .._.
(67)
~
Eq. 67 perhaps provides an improved basis for the cor-
relation of characteristic physical data of porous media
with the electrical formation resistivity factor.
The electrical properties of porous rocks, as discussed
in the foregoing sections, form the basis of quantitative
evaluation of electrical-logging records. In particular,
the Humble relation (Eq. 59) has been used widely by
service companies in estimating porosity from
measurements made with contact resistivity devices
such as the Microlog. Eq. 64 forms the basis of inter-
pretation of water saturations from deeper-penetration
resistivity devices such as the conventional resistivity
curves of the standard electrical log or from the various
focused electrical-resistivity devices.
Improvements in the statistical correlations of elec-
trical properties of rocks will improve the results of such
analyses.
Nomenclature
A = cross-sectional area
A., = area of flow represented by a bundle of
tubes of permeability k,
A, = internal surface area/unit PV
h = constant for a given gas in a given medium
c = fluid compressibility
C = flow coefficient
d = diameter
E = voltage drop
f = friction factor
F, = shape factor
FR = formation resistivity factor
F~ri
= apparent formation resistivity factor
K = acceleration of gravity
h = driving head
I?
to
= height above free-oil level
IT,,~ = height above free-water level
/f, = thickness ofjth layer
I = current
IH = resistivity index
J = J function, Eq. 48
k = permeability of the medium
k,, = permeability to air
k,y = permeability of the medium to a gas com-
pletely filling the pores of the medium
kl, = thermal conductivity
k, = permeability of one capillary tube
kL = permeability of the medium to a single
liquid phase completely filling the pores
of the medium
k,,. = permeability to fresh water
km
= permeability to 500-grain/gal chloride
solution
K = constant of proportionality, Eq. 7
K = some function of the tortuosity. Eq. 63
K; = Kozenys constant
f. = length of flow path or length of the sample
L,, = actual length of the flow path
L.; = length ofjth layer
UI = slope of the curve or cementation factor
n/ -
~ number of tubes of radius r,
p = mean flowing pressure
A~YI = pressure loss over length L
~1~ = base pressure
Pe
= pressure at the external boundary
p,\. = pressure at the wellborc
P,. = capillary pressure
Pc.,l, = capillary pressure. laboratory conditions
P,.,,, = capillary pressure of mercury
pc~.i,,y =
capillary pressure at height above free-oil
surface, using oil and gas
P
c.R
= capillary pressure, reservoir conditions
P,.,,. = capillary pressure of water
P
c I,(,
the free-water surface determined by
using water and gas
P,,,,,, = capillary pressure at a given height above
the free-water surface, using water and
oil
q/, = volume rate of flow at the base pressure
40
= volume rate of flow of oil
q/J
= volume rate of flow at the algebraic mean
pressure (p, +p,, )/2
qr = total flow rate
Q = rate of heat flow
r = radius or resistance
Y = mean radius
r, = radius at the external boundary
TH = hydraulic radius
r 1,
= radius of the wellbore
R,., = resistivity caused by the clay minerals
Rn = resistivity of the rock when saturated with
water having a resistivity of R,,
Ro,I, =
resistivity of a shaly sand when 100%
saturated with water of resistivity R,,
s = distance in direction of flow, always
positive
lTs = gas saturation
SL = total wetting saturation
S,, = oil saturation
S,,. = water saturation
T,f = average flowing temperature
AT = temperature drop
u.,
= volume flux across a unit area of the porous
medium in unit time along flow path s
UY
= volume flux across a unit area of the porous
medium in unit time along flow path x
v = fluid velocity
Vh = BV
V,, = oil volume
VP = PV
V, = sand-grain volume
V,,. = water volume
w = mass rate of flow
WC.I, = weight of dry core
W,.,,. = weight of wet core
W,,. = weight of water
z = vertical coordinate, considered positive
downward
0 = angle between s and the horizontal
19,. = angle of contact of the interface of two
tluids and the capillary tube
Oc.,,,p = water/gas contact angle
fI,.,,i, = water/oil contact angle
p = fluid viscosity
p = fluid density or resistivity
~1, = fluid density at base pressure
P 0
= oil density
u = IFT or conductivity
~I11
= surface tension of mercury
surface tension of water
water/gas IFT
water/oil IFT
Kozeny tortuosity
effective rock tortuosity
fractional porosity
effective porosity
potential function
Key Equations in SI Metric Units
q=14.696x IO"
k4p I -I+)
, . FLL
y=92.349x IO
MP, -P ,,.I
c1 ]n r,lr,, ,
T/,kA(p, 2 -pz )
qb=23.1454x104 _
TfW<&P/,
, .
4)) = I .4554 x 106
T/,vWp, * -P,, *)
ii.fzj~~pj, In rclr,,
(26)
(28)
where
q is in m/d,
k is in pm2,
A is in m2,
p is in kPa,
p is in Pas,
L is in m,
I7 is in m,
r is in m, and
T is in C.
References
I.
2.
3
4:
5.
6.
7.
8.
9.
IO.
Il.
I?.
13
14.
15.
Fraser, H.J. and Graton, L.C : Systematic Packing of
Sphere-With Particular Relation to Porosity and Permeability,
J. G4. (Nov.-Dec. 1935) 785-909.
Tickell. F.G., Mechem. O.E. and McCurdy, R.C.: Some
Studies on the Porosity and Permeability of Rocks, Trmv..
AIME (1933) 103. 250-60.
Core Laboratories Inc.. Dallas, TX.
Nutting, P.G.: Physical Analysis of Oil Sands. Buli., AAPG,
(1930) 14. 1337-49.
Russell, W.L.: A Quick Method for Determining Porosity,
B!i//ril.. AAPG (1926) 10, 931-38.
Stevens, A.B.: A Luhoruror~ Munuulfor Perrdrum Eqinrerirzg
308. Texas A&M U., College Station, TX (1954).
Washburn, E.W. and Bunting. E.N.: Determination of Porosity
by the Method of Gas Expansion. J. Ant. Ccram. SW, 5. 48.
Beeson. C.M.: The Kobe Porosimeter and Oilwell Research
Porosimeter. Trms., AlME (1950) 189. 313-18.
Dotson. B.J. et ni.: Pomsity Measurement Comparison by Five
Laboratories.. Trcins., AIME (1951) 192. 341-46.
Kelton. F.C.: Analysis of Fractured Limestone Cores, Truns..
AIME (1950) 189. 225-34.
Krumbein. W.C. and Sloss. L.L.: .Srrcui~rciph~ and .Scdiww~tu-
rim, Appleton-CenturyCrofts Inc., New York City (1951) 218.
Geertsma. J.: Effect of Fluid Pressure Decline on Volumetric
Changes of Porous Rocks, Truns.. AIME (19.57) 210. 33 1 and
339.
Hall. H.N.: Compressibility of Reservoir Rocks. Trmn.c.,
AIME (1953) 198. 309.
Fatt. I.: Pore Volume Compressibilities of Sandstone Reservoir
Rocks, Trun;. , AIME (1958) 213. 362-64
Hammerlindl, D.J.: Predicting Gas Reservoirs in Abnormally
Pressured Reservoirs. paper SPE 3479 presented at the 1971 SPE
Annual Meeting, New Orleans. Oct. 3-6.
16.
17.
18.
19.
20.
21.
22.
23.
24
25.
26.
21.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
26-33
Newman. G.H.: PowVolume Comprcshihllity ofConsol~datet1.
Friable. and Unconwlidatcd Reservoir Rock\ Under Hydrostutlc
Loading. J. PH. T~I. (Feb. 1973) 129-34.
Van der Knaap. W.: Nonlinear Behavior of Elastic Porou\
Media. Trcrns.. AIME (1959) 216. 179-87
Krug. J.A.: The Effect of Stress on the Petrophyhical Propcrtics
of Some Sandstones. PhD dissertation (T- 1964). Colorado
School of Mines. Golden. CO (1977).
Graves, R.M.: Biaxial Acoustic and Static Measurement ol
Rock Elastic Properties. PhD dissertation (T-2596). Colorado
School of Mines, Golden. CO (1982).
Lachance. D.P. and Anderson. M.A.: Comparison of Uniaxial
Strain and Hydrostatic Stress Pore-Volume Compre%\ibilitic\ in
the Nugget Sandstone. paper SPE 11971 prcscntcd at the 1083
SPE Annual Technical Conference and Exhibition. Sdn Francirco.
Oct. S-8.
Hubben. M.K.: Entrapment of Petroleum Under Hydrodynamic
Conditions. Bull., AAPG (Aug. 1953) 1954-2026
Croft, H. 0. : Thrnnr)fl~~~crt~li~.~, Fluid NON (UK/ Hotrr Trrr!wri.c-
\io/l, McGraw-Hill Book Company Inc.. New York City (1938)
129.
Klinkenberg. L-J.: The Permeability of Porou\ Media to Liquids
and Gates. Drill. crrtd Prod. Pnrc.. API. Dallas ( lY4 1 ) 200- 13.
Johnston. N. and Beeson. C.M.: Water Pemrcability of Rcxr-
voir Sands, Trw~s.. AIME (1945) 160. 43-S?
Fatt. I. and Davis. D.H.: Reduction in Permeability waith Ovcr-
burden Pressure, Trtrm.. AIME (19.52) 195. 329.
Wyllie. M.R.J. and Spangler. M.B.: Application of Electrical
Resistivity Measurements to Problem of Fluid Flow in Porow
Media, Bull., AAPG (Feb. 19.52) 359-403.
Kennedy. H.T., VanMeter, O.E., and Jones. R.G.: Saturation
Determination of Rotary Cores. P<,f. Efr~r. (Jan. 1954)
B.52-B.64
Gates, G.L., Morris, F.C., and Caraway. W.H.: f$cr oj Oil-
Bose Drilhg Nuid Fihrurc on Aucr/~sis of Corr v ,frm Sour/~ C&s
Lewe, Crdijimtiu urld Rcmgc/~. C&r&> /G/d, technical report,
Contract No. RI 4716. USBM (Aug. 1950).
Welge. H.J. and Bruce, W.A : The Restored-state Method for
Determination of Oil m Place and Connate Water. Drill. crrrrl
Prod Prm.. API, Dallas (1947) 166-74
Slobod. R.L., Chambers. A.. and Prehn. W.L. Jr.: USC of Ccn-
trifuge for Determining Connate Water. Residual Oil, and
Capillary Pressure Curves of Small Core Samples. Tram..
AIME (1951) 192. 127-34.
Brown. H.W.: Capillary Pressure Investigations. Trms.,
AIME (1951) 192, 67-74.
Owen, J.D.: Well Logging Study-Quinduno Field. Roberls
County, Texas, paper 593-G presented at the 195.5 AIME For-
mation Evaluation Symposium, Houston. Oct. 27-28.
Leverett, M.C.: Capillary Behawor in Porous Solids, Trmv.,
AIME (1941) 142, 152-68.
Wright. H.T. Jr. and Wooddy. L.D. Jr.: Formation Evaluation
of the Borregas and Seeligson Fields, Brooks and Jim Wells Cow-
ties, Texas. paper 591-G presented at the 1955 AIME FormatIon
Evaluation Symposium. Houston. Oct. 27-28.
Cornell, D. and Katz. D.L.: Flow of Gases Through Con-
solidated Porous Media. hi. cd Etzgr. Chew (Oct. 1953) 45.
Wyllie, M.R.J. and Gardner, G.H.F.: The Generalized Kozeny
Carman Equation. World Oil (March and April 1958).
Winsauer. W.O. PI ol.: Rewtiwty of Brme-Saturated Sands in
Relation to Pore Geometry. Bull.. AAPG (Feb. 1952) 253-77.
Rust. C.F.: Electrical Resistivity Measurements on Reservoir
Rock Samples by the Two-Electrode and Four-Electrode
Methods, Trms., AIME (1952) 195, 2 17-24.
Tlxier. M.P.: Porosity Index in Limestone from Electrical
Logs-Part I , Odrmci Cm .I. (Nov. 15. 1951) 140.
Wyllie, M.R.J. and Gregory. A.R.: Formation Factors of Un-
consolidated Porous Media: Influence of Particle Shape and Effect
of Cementation, Truns.. AIME (1953) 198, 103-09.
dewitte. A.J.: Saturation and Porosity from Electrical Logs in
Shaly Sands-Part I, Oil and GNS J. (March 4. 1957) 89.
Hill, H.J. and Mdhum. J.D.: Effect of Clay and Water Salinity
on Electrochemical Behavior of Reservoir Rocks. Tmm., AIME
(1956) 207. 65-72.

26 Properties of Reservoir Rock

Uploaded by

Copyright:

Available Formats

26 Properties of Reservoir Rock

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

26 Properties of Reservoir Rock

Uploaded by

Copyright:

Available Formats

Chapter 26

Properties of Reservoir Rocks

You might also like