GEOPHYSICS, VOL. 51. NO. II (NOVEMBER 1986); P. 2093-2107. 13 FIGS., 2 TABLES.

Effects of porosity and clay content on wave velocities

in sandstones

De-hua Han*, A. Nur*, and Dale Morgani

ABSTRACT Deviations from these equations are less than 3 percent

and 5 percent for VPand Vs, respectively.
The ultrasonic compressional (V,) and shear (VJ ve-
The velocities of clean sandstones are significantly
locities and first-arrival peak amplitude (A,) were mea-
higher than those predicted by the above linear fits
sured as functions of differential pressure to 50 MPa
(about 7 percent- fnr V, and l-I_ percent for VJ, which
and to a state of saturation on 75 different sandstone
indicates that a very small amount of clay (1 or a few
samples, with porosities 4 ranging from 2 to 30 percent
percent of volume fraction) significantly reduces the
and volume clay content C ranging from 0 to 50 per-
elastic moduli of sandstones.
cent, respectively. Both VP and V, were found to corre-
For shaly sandstones we conclude that, to first order,
late linearly with porosity and clay content in shaly
more sensitive to the porosity and clay content than is
sandstones. At confining pressure of 40 MPa and pore
VP. Consequently, velocity ratios VP/V, and their differ-
pressure of 1.0 MPa, the best least-squares fits to the
ences between fully saturated (s) and dry (d) samples
velocity data are
also show clear correlation with the clay content and
porosity.
V, (km/s) = 5.59 - 6.93$ - 2.18C For shaly sandstones we conclude that , to first order,
clay content is the next most important parameter to
and porosity in reducing velocities. with an effect which is
about 0.31 for VP to 0.38 for V, that of the effect of
K (km/s) = 3.52 - 4.91$ - 1.89C. porosity.

INTRODUCTION the velocity of the pore fluid. When both V, and V are fixed,
the only variable in the equation is porosity. To first order,
Shaly sandstones and shales comprise a major component this simple equation appears adequate for clean sandstones in
of sedimentary basins and are of foremost relevance to hydro- the middle range of porosity (10 percent < $ < 25 percent).
carbon reservoirs. The acoustic properties of shaly sandstones However, it is well-known that acoustic velocities of sand-
and shales are thus of great interest in seismic and well log stones are also related to mineralogy, pore geometry, degree of
interpretation. consolidation, cementation, confining pressure, pore fluid,
For years, the time-average equation of Wyllie et al. (1956, pore pressure, and temperature. Consequently, the short-
1958) has been used to obtain porosities from acoustic velocity comings of the time-average equation have been extensively
logs. The equation for P-wave velocity V in water-saturated discussed (Geertsma, 1961; Geertsma and Smit, 1961; Raymer
rock is et al., 1980; Kevin, 1981). A newer, empirical equation based
on well log data was obtained by Raymer et al. (1980):

where V, is the P-wave velocity of the rock matrix and V, is Equation (2) was proposed as an alternative to the time-

Manuscriptreceivedby the Editor October 18, 1985; revised manuscript received May 5, 1986.
*Stanford Rock PhysicsProject,Department of Geophysics,StanfordUniversity, Stanford,CA 94305.
$Formerly StanfordRock Physics Project, Stanford University; presently Department of Geophysics, Texas A&M University, College Station, TX.
10 1986 Society of Exploration Geophysicists. All rights reserved.

2093
2094 Han et al.

average equation for interpretation of acoustic logs. Because

the porosity is the only parameter in this equation, it is not
.eo I 1 I I 8 ~~~ very different from the time-average equation [equation (l)].
Neither equation (1) nor equation (2) can be directly applied
E BUILDING SANDSTONE
T: TIGHT GAS SANDSTONE to shaly sandstones. From the results presented here, as well
G
.!a -
G: GULF SANDSTONE as from earlier work by De Martini et al. (1976), Tosaya and
P
PI P-SANDSTONE Nur (1982a), and Kowallis et al. (1984), in shaly sandstones
cc XI CLEAN SANDSTONE
and shales the time-average equation significantly overesti-
.40- P$ mates velocities. as does Raymers model. The question then is
P
!i P t

F P how can the effect of clay best be represented in the velocity

z equation for shaly sandstones?
s .w-
Although there are many theoretical models for the effects
z c c
S Es
of porosity, pore shape, fluid, and matrix moduli on the elastic
d c
.20-
G properties of rocks (Gassmann, 1951; Biot, 1956; Geertsma,
T G
s 1961; Kuster and Toksdz, 1974; OConnell and Budiansky,
TI T P
SP P 1974; Mavko and Nur, 1978; Walsh and Grosenbaugh, 1979),
T TT T
pP c =G GE
. 10
c s
none of them includes the effect of clays on velocities in sand-
stones. Minear ji982) applied the Kuster and TokaBz (1974)

1
TS s c sg s Sk

I S Ss Ss model to simulate the effects of clays on velocities of sand-

a%cn .os .lO
I
.k
.a

- -20 .K .P .3E stones. His results suggestthat clay minerals may significantly
POROSITY reduce elastic moduli and velocities of sandstones. However,
the magnitude of these effects of clays remains far from clear.
FE. 1. The ranges of clay content and porosity for the 75 In contrast with theory, some experimental and petrographic
shaly sandstones of this study. Porosity ranges from 2 to 30 work has been published on the effects of clay minerals
percent, and clay content ranges from 0 to about 50 percent.
(De Martini et al., 1976; Tosaya and Nur, 1982a; Kowallis et
The data indicate that sandstones with high clay content tend
to have low porosities. al., 1984). Although each of the above studies was limited to a
few samples, all results have alluded to a general trend-that
increasing clay content in sandstones systematically decreases
acoustic velocities in both well-consolidated and poorly con-
solidated sandstones. Costagna et al. (1985) obtained field re-
sults which suggest a linear dependence on the porosity and
clay content for both V, and Vs, inferred from sonic log data
from the fiio formation. Thus, because of the pauciry of data,
the main goal of this study is to investigate systematically,
under laboratory conditions using a large number of samples

6.00

St BIJILOING SANDSTONE St BUILDING SANDSTONE

TI TIGHT GAS SANDSTONE Tt TIGHT GAS SANDSTONE
150- X X
CI GULF SANDSTONE G GULF SANDSTONE
PI P-SANDSTDNE PI P-SANDSTONE
XI CLEAN SANDSTONE XI CLEAN SANDSTONE
7s T I
100-
T s T@ s X
s
SS

1 I
.cl5 .lO -15 -20 .25 .30
POROSITY POROSITY

FE. 2. Measured (a) compressional and (b) shear velocities versus porosity for 75 sandstone samples at P, = 40 MPa
and P, = 1.0 MPa.
 Wave Veiocitles in Sanilstones 2095

cbvering a wide range of porosities and clay contents, the Finally, Gassmanns equation (Gasstnann, 1951) was used
effect of clay content versus porosity on the Acoustic velocities to explore the frequency dependence of rock elasticity as pro-
in sandstones. posed by Winkler (1985).
Because shear wave velocities are now available from well
!ogs and seismic refiection measurements, it is- of particular
interest to explain the relation between shear velocity and EXPERIMENTAL iROCEDURES
p&o&y. An empirical relation bet~ween shear velocity and
porosity has been proposed in a modification of the time- Sample description
average equation (e.g., Domenico, 1984). However, as we show
later, this equation cannot be used very well to interpret shear The 75 sandstone samples used in this study come from
velocity values in shaly sandstones. A systematic investigation either well cores or quarries. The porosities of the samples
of the effects of clay content and porosity on shear velocity V, range from 2 to 30 percent, and the clay contents by volume
is consequently of great interest. fraction range from 0 to 50 percent (Figure 1). Ten samples are
We measured compressional velocity V, and shear velocity tight gas sandstones (T) with very low poro$ities. Twenty-four
V, as functions of pressure in 75 sandstone samples with vary- samples (G) come from a few offshore wells in the Gulf of
ing clay conteats and porosities. All the data are tabulated in Mexico; some of these are poorly consolidated. Eleven sam-
Appendix A. We also investigated the relations among ples (P) are also borehole cores which are well-consolidated.
changes in relative attenuation of waves with varying rock Thirty well-consolidated samples (S) are from quarries, of
porosity and clay content. The relations among the velocity which five (X) are clean sandstones with less than 1 percent
ratio VP/V,, water saturation, elastic moduli, porosity, and clay content. Figure 1 shows clay content versus porosity for
elay content are also examined. all samples. It is noteworthy that samples with more than 20

Table 1. Model fitting to the experimental velocity data.

WITHOUT CLAY TERM (70 samples)

1
__ = 0.194 + 0.3284 R: 0.844 rms: 6.6%
VP

1
- = 0.322 + O.ci?&$ R: 0.750 rms: 10.3%
V*

VP = 5.02 - 5.634 R: 0.840 rms: 6.8%

v, = 3.03 - 3.78f#J R: 0.754 rms: 10.8%

WITH A CLAY CONTENT TERM

1. LINEAR EQUATION

VP = 5.59 - 6.934 - 2.18C R: 0.985 rms: 2.1%

v, = 3.52 - 4.01fp - 1.89C R: 0.959 rms: 4.3%

2. TRAVELTIME AVERAGE EQUATION

1
- = 0.163 + 0.3994 + 0.119C R: 0.972 rms: 2.8%
VP

1
- = 0.242 + 0.8124 + 0.307C R: 0.945 rms: 5.1%
v,

R: CORRELATION COEFFICIENT
RMS: RELATIVE rms DEVIATION WITH 68.3% CONFIDENCE.
2096 Han et al.

percent clay content tend to have lower porosities, ranging pressure cycles. Generally, hysteresis is observed, but its mag-
from 5 to 15 percent. nitude is typically small. After the first pressure cycle, veloci-
ties measured during unloading cycles are repeatable, with
Acoustic measurement hysteresis in the well-consolidated samples of less than 1 per-
cmt and in then poorly consolidated samples of !ess than
Wave velocities VPand V, and the associated compressional 2 percent. The first peak amplitudes A, were measured using
first-arrival peak amplitude A, are measured as functions of the method of Tosaya (1982b).
pressure and the state of water saturation. Confining pressure Unless otherwise mentioned, data shown in the figures are
P, and pore pressure P, are controlled separately. With differ- for confining pressure of 40 MPa and pore pressure of 10
ential pressure P, limited to 50 MPa and pore pressure limited MPa. The detailed results are given in Appendix A.
to 1.0 MPa, the system can thus simulate pressure conditions
to depths of over 2 000 m.
The velocities are measured at ultrasonic frequencies using Density and porosity measurements
the pulse transmission technique (Birch, 1960). The central
frequencies of the transducers used for P- and S-waves are 1.0 Samples are vacuum oven-dried at less than 50C for two to
and 0.6 MHz, respectively. The wavelengths for P- and S- eight weeks, and weighed to within 0.01 g so that the density
waves are greater than 5 mm and at least five times greater of a dry sample pd can be determined to within 0.3 percent.
than a mean grain size for our samples. On each sample the The density of the wet sample p, is then calculated by the
measurements were first performed in a vacuum dry state relation
(vacuum pressure less than 0.01 Torr), then at full saturation
with water. In some samples with high clay content, brine is ku=Pd+bf. (3)
used as the pore fluid. No velocity differences are found be- where p,. is the pore fluid density and 4 is the porosity of the
tween saturation with brine and saturation with plain water. sample. The porosity of each sample at room pressure is mea-
Samples are 5.0 cm in diameter and 2.0 to 5.0 cm in length, sured with a helium porosimeter, repeatably within 1 percent
or more than 100 times the average grain size. Both sample of total bulk volume. This value of porosity is considered
dimensions are measured to within 0.05 mm. The two end equal to the porosity at differential pressurePd of 1.0 MPa.
surfacesof each sample are ground parallel to within 0.05 mm. Although the bulk volume of each sample decreases with
P-wave arrival times are picked to within 0.03 us, which increasing differential pressure, in the pressure range used the
leads to less than 1 percent absolute error in the measurement change in volume of the grains is very small, so that the
of VP. The error in the V, measurement is less than 2 percent, variation of bulk volume with pressure can be considered
except for poorly consolidated sandstones at low differential equal to the change of the pore volume only. The variations of
pressure (P,, < 10 MPa), where errors may be up to 3 percent the pore volume are monitored with a pore-pressure intensifier
due to poor signal_ while the pore pressure is kept constant. The ambient temper-
The velocities are measured during loading and unloading ature is 22C * 1C.

l/V,=.194+.328$1 1/ V, =.194+ .328@

1.a0 1. a, , I

1.10 -
cc
G 1.10 - t EC
G
c
S
c
G ITs G
s pc;p

9
.m -
St BUILDING SANDSTONE
u
e .al- sl BUILDING SAtdlSTONE
Tt TIGHT GAS SANDSTONE
F i-Kl3-T GAS SANDSTONE
GM-i= SANDsroNE
G GULF SANDSTONE
P-SANDSTONE
PI PSANDSTDNE
CLEAN SAtQSTONi
XI CLEAN SANDSTONE

FIG. 3. Compressional velocity V,,data at P, = 40 kPa and P, = 1.0 MPa fitted by the time-average equation.
(a) Relative deviations versus porosrty showing large scatter. (b) Relative deviations show a clear correlation with clay
contents.
 Wave Velocities in Sandstones 2097

Petrographic measurement THE EFFECTS OF POROSITY

AND CLAY CONTENT ON VELOCITIES
The clay content of each sample is obtained by point count-
ing on thin sections. Most results are based on 300 point Based on the combination of acoustic and petrographic
counts per section, with a few earlier results based on only 100 data, the effects of confining pressure P,, pore pressure P,,
point counts. Minerals with flaky textures, such as hematite and fluid saturation on acoustic P- and S-wave velocities and
and other iron oxides, are counted as part of the clays. Usu- amplitude are studied.
ally two thin sections are taken, from the top and bottom of Compressional velocity V, and shear velocity V, versus po-
the sample, respectively. For some samples the clay contents rosity 4 for all samples are shown in Figures 2a and 2b.
from these two sections differ by as much as 20 percent, due Despite significant scatter, clear trends indicate that both V,,
mainly to heterogeneity in the sample. and v, decrease with increasing porosity. As a first trial, the

V, = 5.02 - 5.634 v,=3.03-3.784

l.sa, G
G
BUILDING sandstone
TIGHT GAS SANDSTONE
1.10 - GULF SANDSTONE
c
s c P-SANOSTONE
c CLEAN SANDSTONE
h t r: s + l5.G G
* i.m-
1 3 7 1.10 p PP
2 (G
z
n
11 SIT s P pPp& SSs5s S@ 5
m t
kPts
T G
G

I .m-
TTI S
e c km-
h
c
b
I P Pp
Ii X x ss G6;
X
P P
f
.w- G $2 %S G
P Sa BUILOING SANDSTONE
SS S
Ta TIGHT GAS SANDSTONE T
X
C GULF SANDSTONE
.70 - PI P-SANDSTONE .W- XX
X
XI CLEAN SANDSTONE X

..L, .k .io .i5 .ko .k .j, .b5 %

k .k .io .15 .a .25 .a
PDRDSITY POROSITY

V, = 5.02 - 5.634 V,= 3.03 - 3.784

l.aO* 1 1. Jo I
C C G
G
EC
t G
1.10 -

c
PP P

E PP p
TS CG

e
E
za .wt &
TI
G
BUILIYING SANDSTONE
TIGHT GAS SANOSTaJE
GULF SANDSTONE
BUILDING sandstone
TIGHT GAS SANDSTONE

I
GULF sandstone
R P-SANDSTONE PI P-SANOSTONE
.70 XI CLEAN SANDSTONE XI CLEAN SANDSTONE

I I I
*Boo .10 .20 .P .ul -50 -10 .P .J1 .uJ -50 .a
CLAY CONTENT CLAY CONTENT

FIG. 4. Compressional V,, and shear V, velocities fitted by the model V = A 0 - A,$. (a) Relative deviations of VPversus
porosity, showing large scatter. (b) Relative deviations of VPversus clay content. (c) Relative deviations of V, versus
porosity, showing large scatter. (d) Relative deviations of V, versus clay content.
2098 Han et al.

modified time-average equation l/V, = B, + B,$ is fitted to B, = l/V,;

the data. By least-squares regression, the fitted results (Table
B, = (l/V, - l/V& (6)
1) are presented in the form of relative deviations versus po-
rosities, as shown in Figure 3a. The matrix P wave velocity & = (f/T/ - l/v,),
computed from the fit is V, = l/B, = 5.15 km/s, which is
where Vmis matrix velocity, V< is fluid velocity, and I$ is clay
much lower than the value for nonporous quartz aggregates,
velocity.
!p = 605 km/s- (Robert, 1,082).The relative deviations of the
Kesuits of the fits of these equations to thevelocity data are
data from the values predicted by the equation are quite large
listed in Table 1. Both equations show similar correlation co-
(Figure 3a). However, these deviations clearly depend upon
efficients and relative rms deviations from the data. However,
the clay content (as in Figure 3b), which indicates that clay
the time-average equation is obviously meaningless for shear
content systematically affects velocity.
waves. Furthermore, the velocities computed from this equa-
We also use the empirical linear model V = A,, - A,$ to fit
tion are systematically higher than measured ones in the low
both V, and V, data. Again, the relative deviations of the data
range of porosity (+ < 7 percent) and high range of porosity
from the values predicted by the equation are quite large (Fig-
($ > 25 percent), in contrast with the porosity in the middle
ures 4a and 4c) and clearly depend upon the clay content
range (Figure 5). On the other hand, the deviations of the data
(Figures 4b and 4d).
from linearity are randomly distributed. Furthermore, the
As shown in Figures 3a, 4a, and 4c, velocities of clean sand-
stones are systematically higher than predicted from the fit. To linear equation (4) fits both V and t data slightly better.
Consequently, we choose the linear equation (4) to describe
emphasize the effects of clay content, the data used to fit the
the dependencies of velocities on the porosity and clay content
model excluded the totally clay-free sandstones.
in sandstones. The equations are similar to those proposed by
Based on the above results we conciude that any model
Tosaya and Nur (1982a), Kowalis et al. (1984), and Costagna
used to fit both VP and V, data in shaly sandstones must
et al. (1985).
account for clay content. Two simple equations that include
The following best fits to the data at the confining pressure
clay-content terms are used to describe the data by least-
squares regression; namely, 40 MPa and pore pressure 1.0 MPa are obtained by least-
squares regression. For the compressional velocity,
V=Ao-A,$-A,C, (4)
V, (km/s) = 5.59 - 6.934 - 2.1SC, (7)
and
and for the shear velocity.
l/V=&-B,@-B,C. (5)
V, (km/s) = 3.52 - 4.914 - 1.89C. (8)
Equation (5) is actually the time-average equation modified to
Velocity deviations from fits versus porosites and clay content,
include a linear term for clay content. The coefficients B,, B,,
respectively, are shown in Figures 6a and 6b for VP, in Figures
and B, may be interpreted as follows:
6c and 6d for V,. The correlation coefficient is 0.985 and the
relative rms deviation is 0.021 for VP;the corresponding values
are 0.959 and 0.043 for VT. These values show greatly im-
l/V,=.163+.399d+.119C proved fit in comparison with the results using porosity alone.
In the very clean sandstones (St. Peter, Beaver, and Fon-
tainebleau), the measured velocities are higher than those pre-

II----l E
dicted by equations (7) and (8) by about 7 percent for V, and
11 percent for V, (Figures 6a and 6b). This distinct difference
between the clean and shaly sandstones implies that even
small amounts of clay in sandstones tend to soften grain con-
tacts significantly. Such softening is most likely related to clay
particles situated between grain boundaries. Because the grain
size of the clay particles is so small and their surface area is so
large, even a small volume fraction of clays can cover the
entire pore surface area throughout the rock, including the
grain contacts. We believe that the contact clay is responsible
.m - St BUILDING SANDSTONE _ for the decrease of velocity from clean sandstones to shaly
P
TI TIGHT GAS SANDSTONE ones.
The good fit of shaly sandstone velocities [represented by
equations (7) and (S)] suggests that the velocities are nearly
independent of the type of clay or the location of clay particles
within the rock matrix. Minear (1982) has investigated models
.25 .P .35 of sandstone with (1) structural, (2) laminal, and (3) clay sus-
POROSITY pended in the pores. His model results show that the suspend-
ed clay has only a small effect on velocities, whereas both
FIG. 5. Compressional velocity data fitted by the modified
time-average equation with a clay-content term. Relative devi- structural and laminal clays have significant and similar effects
ations versus porosity indicate that the model overestimates on velocities. Both arrangements predict a nearly linear rela-
V, at porosities lower than 7 percent or higher than 25 per- tion between velocities and clay content for clay content
cent, and underestimates VPat intermediate porosities. below 50 percent. Because the exact arrangement of clays
 Wave Velocities in Sandstones 2099

within sandstones is not known, we can only suggest that Finally, the coefficients in equations (7) and (8) indicate that
something like Minears (1982) structural or laminal arrange- the influence of clay content (by volume) is about l/3.2 that of
ment is typical. the influences of porosity for V, and l/2.6 for V,. These ratios
The coefficients of the linear fits in equations (7) and (8) for are also fairly independent of pressure (Table 2).
both V, and V, are fairly constant with differential pressure In summary, it appears that the volume of clay in consoli-
over 10 MPa (Table 2). These results suggest that the effects of dated shaly sandstones is an important parameter when accu-
porosity and clay content are fairly independent of differential rate porosity evaluations from seismic or acoustic data (e.g.,
pressure, which can be extrapolated to a higher P, range. For well logs) are required. Other parameters, including pore ge-
data with P, below 10 MPa, the fit is somewhat worse. ometry, grain size, grain contacts, cementation, type of clay,

1/,=5.59-6.93$-2.18C IS=3.52-4.91g-1.89C

SI BUILDING SANDSTONE SI BUILDING SANDSTONE

TI TIGHT GAS SANDSTONE T: TIGHT GAS SANDSTONE

I;L .bs , , ;FfEEff 6

Pa
GULF SANDSTONE
P-SANDSTCNE
)G CLEAN SANDSTONE

.10 .I5 .20 .a5 .P .15 .a0 .25 -30

POROSITY POROSITY

l,=5.59-6.93 b-2.18C l,=3.52-4.91$-1.89C

T
1. LO -
sT
# f_!fiT CC G G

. t.m I- ..c
d P- G
5 pPp
0)
T P
I .su P
.
B
Sn BUILDING SANDSTONE E & BUILOING SANDSTONE
.m -
T: TIGHT GAS SANDSTONE 2 TI TIGHT GAS SANDSTONE
Ga GULF SANDSTONE C GULF SANDSTONE
PI P-SANDSTONE R P-SANUSTDM
x8 CLEAN SANDSTONE 1 .70 - XI CLEAN sandstone

.io .;O .30 .;O .;O .1, -a& .I0 .p .;o .z .;o
CLAY CONTENT CLAY CONTENT

FIG. 6. Linear fit between velocity and porosity and clay content: (a) VP relative to porosity, (b) V, relative to clay
content; (c) V, relative to porosity; (d) V, relative to clay content, all at P, = 40 MPa and P, = 1.0 MPa.
 Han et al.

I
distribution of clays, and mineralogy, have much smaller influ-
St BUILDING SANDSTONE
ences on velocities at high differential pressure for shaly,
Cc GULF SANDSTONE water-saturated sandstones.
PI P-SANDSTONE
XI CLEAN SANDSTONE
c
EFFECTS ON WAVE AMPLITUDES
Fp s G X
P We have also studied the dependence of wave amplitude on
s G
porosity and clay content, using a comparative method as
EP c X
described by Tosaya (1982b). All the amplitude data were cor-
P S rected and normalized by that of an aluminum sample with
c P
S S the same length. For P-waves, we find that the amplitude -4,
c s increases with decreasing porosity (Figure 7), although the
G c s P SS G scatter is quite substantial. For 46 water-saturated samples,
G
FCG B the best fit by linear regression of the amplitudes as functions
5
c G GGG of the porosity and clay content is
s GG

A, = 0.58 - 1.514 + 0.23C, (9)

I I
I .Gs -10 .15 .a .x with a correlation coefficient of 0.68. The results show that the
POROSITY
amplitude depends strongly on the porosity and only weakly
FIG. 7. Normalized amplitude A, versus porosity for 46 sand- on clay content. We conclude that parameters other than clay
stone samples at P, = 40 MPa and P = 1.0 MPa. The results content are probably important. No clear relations are ob-
indicate that A,, tends to decreasewitg porosity. served between shear amplitude A, and porosity, probably

Table 2. The pressuredependenceof the coefficientsin the linear velocity-porosityclay model.

VELOCITY EQUATION

(bars) (km/s) (km/s) (km/s) (km/s) PO)

400 5.59 0.93 2.18 0.985 0.09 2.1

300 5.55 6.86 2.18 0.985 0.09 2.1

200 5.49 6.94 2.17 0.981 0.10 2.4

100 5.39 7.08 2.13 0.978 0.11 2.8

50 5.26 7.08 2.02 0.969 0.14 3.4

300 3.47 4.84 1.87 0.957 0.11 4.5

200 3.39 4.73 1.81 0.951 0.12 4.9
100 3.29 4.73 1.74 0.937 0.14 5.8
51

R:

Sd: STANDARD DEVIATION

RMS: RELATIVE rms DEVIATION WlTK 6KWi CUTUFI-

DENCE.
 Wave Velocities in Sandstones 2101

because the shear signal is often distorted in a sample with values (Picket& 1963; Gregory, 1977; Benzing, 1978; Johnson,
high clay content, so that A, cannot be measured accurately. 1978; Tatham, 1982; Eastwood, 1983; Domenico, 1984; Rafa-
vich et al., 1984; Costagna et al., 1985). Our data show that
THE EFFECT OF CLAY CONTENT the velocity ratio for water-saturated shaly sandstones de-
ON THE ELASTIC MODULI pends upon both porosity and clay content. By least-squares
regression, this dependence is found to be
The effective elastic moduli for our samples are computed
V,/v, = 1.55 + 0.564 + 0.43c, (11)
from the velocities by the expressions

K = p($ - 4Vi/3) with a correlation coefficient of 0.70. The results show that
increasing porosity or clay content (Figure 9) increases VP/V,
and (IO) and that the velocity ratio is niore sensitive to porosity
G = pVf, changes, in agreement with the results of Costagna et al.
(1985).
where p is the density of the sample. Clearly, elastic moduli Sandstones with high clay content have velocity ratios and
relate to porosity and clay content in a manner similar to the Poissons ratios similar to carbonate rocks. The resulting am-
velocities, as shown for bulk and shear moduli versus porosity biguity in the interpretation of velocity data may be resolved
in Figures 8a and Sb. For shaly sandstones, clay content af- by the combined use of the velocity and the velocity ratio,
fects the shear modulus more than the bulk modulus. More- providing a useful tool for reliable lithology discrimination.
over, as seen in Figure 8b, clean sandstones have much higher Costagna et al. (1985) found that shear velocity is nearly
shear moduli than do shaly sandstones with the same poros- linearly related to compressional velocity for water-saturated
ities, suggesting that grain boundary clays significantly reduce elastic silicate sedimentary rocks; the relation is
the shear modulus but have only little influence on the bulk
modulus. Consequently, the effects of clay content on VP and V,(km/s) = 1.161/,+ 1.36. (12)
V, [equations (7) and (S)] are more due to the fact that clays Our data also show that V, is nearly linearly related to V,,
reduce the shear modulus of sandstones. with somewhat different coefficients than in equation (12). For
The effects of porosity on the shear modulus are also larger 75 samples, the best linear least-squares fit yields
than effects on bulk modulus, as indicated by equations (7)
and (8). V,, (km/s) = 1.26% + 1.07. (13)

with a correlation coefficient of 0.97.

THE EFFECTS OF CLAY CONTENT
ON VELOCITY RATIO
THE EFFECTS OF WATER SATURATION
Because more shear velocity data are becoming available in
seismic exploration and well logging, the velocity ratio V,/V, is The data shown so far are all for water-saturated samples.
becoming a useful parameter in the determination of rock In contrast, data for dry samples show much more scatter in
properties. Previous laboratory and well-logging studies have the relations among velocities, porosity, and clay content, with
suggested correlations between lithology, porosity, and V,/V, large deviations appearing when fitting linear equations to the

400.0, 350.0 I 1 I

SI BUILDING SANDSTONE I x Sr BUILDING SANDSTONE

T: TIGHT GAS SANDSTONE T; TIGHT GAS sandstone
EW.0 GULF SANDSTONE
G: GULF SANDSTONE ti
PI P-SANDSTONE i 7 I: P-SANDSTONE
XI CLEAN SANDSTONE T Xr CLEAN SANDSTONE
ij 300.0

e t
; x0.0- ;2ouo-
Y c3

5 5
is 200.0- Y
v, 150.0 -

150.0 100.0 -

t 4;
I
loaaon .05 . 10 .15 .20 .25 .3O Yloll .05 . 10 .15 .2n
1 .25 .30
POROSITY POROSITY

FIG. 8. (a) Bulk and (b) shear moduli versus porosity at P, = 40 MPa and P, = 1 MPa. Relative decrease of the shear
modulus is more than the bulk modulus with increasing porosity. Scatter of the shear modulus, caused by the clays, is
larger than corresponding scatter of bulk modulus data.
2102 Han et al.

data for dry samples. One reason for the scatter is that veloci-
ties for several Gulf sandstones are systematically lower than
1.90 1 I other samples with the same porosity and clay content. Some
G t

Gulf sandstones with high clay content are poorly consoli-

Es G dated. It thus appears that water saturation affects the veloci-
GG G

1.80 -
S G ties in shaly sandstones as a function of differing degrees of
S CT
G t
G consolidation, via interactions between the pore fluid (water)
c S
P and the clays.
P The shear and bulk moduli for our dry samples are com-
+ 1.70 - P
P P puted from velocities and densities by equations (10). It is
.t? P
commonly expected that the shear modulus should remain
(
3
constant or increase only slightly as rock is saturated from its
1.80 5 T dry state (Gassmann, 19.51; Biot, 1956; Kuster and Toksiiz,
P s 51 BUILDING SANDSTONE 1974; OConnell and Budiansky, 1974). This assumption
Ta TIGHT GAS SANDSTONE
agrees with our data for clean sandstones (Figure 10). The
t GULF SANDSTONE
1.50 PI P-SANDSTONE data for shaly sandstones, however, display a more complex
i Xr CLEAN SANDSTONE situation. The ratios of shear moduli G(sat) to G(dry) deviate
significantly from the expected value of 1 (Figure lOa). King
(1966) and Toksiiz et al. (1976) have alluded to possible inter-
actions between pore fluid and clay minerals as processesto
CLAY CONTENT soften rock matrix and decrease the shear modulus. We pres-
ent data in Figure 10a to show how such interaction affects
FIG. 9. Velocity ratios V,/V, at P, = 40 MPa and P, = 1 MPa shear moduli for shaly sandstones. The ratio of G(sat)/G(dry)
versus clay content. shows (excluding the poorly consolidated G sandstones) a de-
creasing trend with increasing clay content, suggesting that
shear moduli of clays can be significantly reduced by water
saturation. Many data, from Gulf sandstones in particular,
deviate from this trend with their ratios of G (sat)/G (dry) being
greater than 1 even though they have high clay contents,
The effects of water saturation on the bulk modulus of rock
were studied by several researchers (Gassmann, 1951; Biot,
1956; Nur and Simmons, 1969; Kuster and Toksoz, 1974;
OConnell and Budiansky, 1974) who showed that water can
significantly increase bulk moduli of rocks. This agrees with
data for clean sandstones (Figure lob). However, effects of
clay content on the bulk modulus upon water saturation have

250 1 I r

Sa BUILDING SANDSTONE
G: GULF SANDSTONE 2iz _ SI BUILDING SANDSTONE
GI GULF SANDSTONE
PI P-SANDSTONE G
Xt CLEAN SANDSTONE PI P-SANDSTONE
XI CLEAN SANDSTONE G
2
L zco-
-u G G
G
G
GG s G E
P P

G sGG
E
,P
GP S G
P P
s P G s
P
SP G
P
p P
P

I
.i -20 .so .40 .So .%I .lO .a0 .-a -.50.40 .6 0
CLAY CONTENT CLAY CONTENT

FIG. 10. The shear modulus (a) and bulk modulus (b) ratios of saturated (s) over dry (d) state versus clay content.
 Wave Velocities in Sandstones 2103

not been mentioned in the literature. Our data show that the
ratio of K(sat)/K(dry) increases from 1.25 for clean sand-
stones to over 2.0 for some samples with a high clay content
(Figure lob). This large effect of increasing K(sat) suggests
that bulk moduli of saturated clays are much greater than
those for dry clays, in contrast to the effect on shear moduli.
This effect on K may be the consequence of water-saturated
micropores being more structurally rigid than vacuum-dry
ones. Though the data are scattered, it is clear that most Gulf
sandstones have larger K (sat)/K(dry) ratios than do other
samples, especially those with high clay content.
Moreover, the P-wave amplitude ratio of A,(sat)jA,(dry) x zoo- G G
reveals that for Gulf sandstones this ratio is greater than 1, as z G
opposed to the effect for all other sandstones (Figure 11). This b G G
result again suggests that Gulf sandstones are systematically G
g 1-w_ GG G G
G
different from the other sandstones in our study.
Indeed, most Gulf sandstones are poorly consolidated and
t. G G G
GGG
k S G P
show larger porosity reductions than do other sandstones with
increasing P,. Moreover, Gulf sandstone clays are mainly
( 1.00
t c:s p
pPP P

1t sS4 4
P
pP
composed of illite and montmorillonite, which differ from SI BUILLJINC SANDSTONE
.a- S
other sandstones in that their clays are mainly composed of G GULF SANDSTONE
R P-SANDSTONE
kaolinite. It appears that in these rocks one of the effects of
.%a CLEAN SANDSTONE
water saturation is to stiffen the contacts between the quartz
grains. This effect is particularly pronounced in poorly con- a%oo .10 .a3 .30 :a .50 .w
solidated sandstones, possibly because grain contacts are al- CLAY CONTENT
ready loose.
FIG. 11. The amplitude ratio A,(sat)/A (dry) versus porosity
The combined effects of water saturation and matrix hard-
at P, = 40 MPa and P, = f.0 MPa.
ening on the moduli of shaly sandstones may explain much of
the scatter in Gulf sandstone data presented in Figures 10 and
11. More accurate petrographic description of these rocks may
be required in the future.
Finally, owing to differing responsesof bulk modulus K and
shear modulus G to the water saturation versus the clay con-
tent for shaly sandstones, the differences between velocity .35 I , I I 7

ratios V,/V, of fully saturated (s) and dry (d) samples can be G
more clearly related to the clay content and porosity as fol- G
.3G- G G
lows :
G
D = (VP/r;,),- (VP/v,), = 0.018 + 0.36+ + 0.47C, (14) -u .25- G G
P
a
with a correlation coefficient of 0.89. The effect of the clay < 6 GGS Pp P
content on D is shown in Figure 12. This effect is greater than P .20- S
P
that of the porosity in equation (14). This relation suggests
G
that the velocity ratio may be used as an index of saturated !jG G p2 GG G
S
P
states in shaly sandstones. S+
SPP

EXAMINATION OF GASSMANNS EQUATION

Our velocity data for dry and saturated sandstones are used
;jm:, G, g5 1
to test Gassmanns (1951) relations:

K Kd
K,=K,+K,-_K+-K~)
Kf
(15) .
10 .zo .P .4G .50 .w
CLAY CONTENT
and
FIG. 12. Differences of velocity ratios V,/V, between saturated
G, = G,, (16) (s) and dry (d) samples versus clay content at P, = 40 MPa
and P, = 1 MPa.
where K,, G, and K,, G, are the bulk and shear moduli of
saturated and dry sandstones, respectively, and K, and K, are
matrix and pore-fluid bulk moduli. These relations coincide
with the low-frequency limit of the Biot equations (Biot, 1956).
Because our data were obtained at high frequencies (lo-lo6
Hz), the dependence of moduli on frequency may be explored.
2104 Han et al.

If we assume that velocities in dry rocks at high frequencies and shear velocity V, are linearly related to porosities of from
are equal to the corresponding low-frequency values (Winkler, 2 to 30 percent and to clay contents of from 1 to 50 percent.
1985) Gassmanns equation can be used to estimate the satu- The effect of clay content in reducing velocity is about l/3.2 as
rated moduh at low frequencies. great as the effect of porosity for VP and l/2.6 times as great
The bulk modulus of the matrix depends upon clay content. for V,.
For shaly sandstones, we simply assume that the bulk modu- Generally, the effects of porosity and clay content on the
lus can be computed by equation (10) while substituting the shear velocity V, are larger than on the compressional velocity
velocities of equations (7) and (8) (assuming zero porosity). V,. Thus, a sample with high porosity and clay content tends
The bulk modulus for water is K, = 2.2 GPa, calculated from to have a high VJV, ratio.
the compressibility data for water by Robert (1975). With P- and S-wave velocities V, and V, of clean sandstones are
these values and the K, values obtained from the velocity and significantly higher than for shaly sandstones with the same
density data, we compute K,. The computed bulk modulus porosity. The matrix velocities 5.59 km/s for V, and 3.52 km/s
K, is systematically less than the measured value (Figure 13). for V, obtained from equations (7) and (8) with porosity and
As bulk modulus increases (the porosity decreases),deviations clay content set to zero are significantly lower than the corre-
increase, up to around 10 percent. Good agreement between sponding velocities for quartz aggregates, i.e., 6.05 km/s for V
data and computed values is found at porosities greater than and 4.09 km/s for V, (Robert, 1982). This difference implies
0.25. The systematic deviations of computed values from mea- that a small amount of clays (1 or 2 percent volume fraction)
sured values suggest that the dispersion of bulk modulus with can significantly soften the sandstone matrix, leading to re-
frequency depends somewhat on the porosity. Nevertheless, it duced velocities.
appears that laboratory data in the megahertz range can be The effects of all clays on velocities, other than the first 1
used in the seismic frequency range through the above ap- percent or so, are described by equations (7) and (8). The effect
proach. depends upon the volume clay fraction. A simple model by
Figure 13 shows quite a bit of scatter which might be Minear (1982) shows that to first order these clays are ar-
caused by the effects of clay content. Further study of the ranged as lamina in the rocks or as grains between the sand
effects of the porosity and clay content on velocities using grains. Furthermore, the effects of clays on velocities VPand V,
Gassmanns equation, adjusted to the seismic frequency range are more from the reduction of the shear modulus than from
by the above approach, might prove valuable. the buik moduius.
The interactions between clay minerals and water also influ-
CONCLUSlON ence elastic moduli and velocities. The pore fluid (water) ap-
pears to decrease shear modulus while increasing bulk modu-
On the basis of our experimental results, we conclude that lus of the clays in sandstones. In poorly consolidated sand-
in water-saturated shaly sandstones, compressional velocity V, stones, however, water-saturated clays tend actually to stiffen
grain contacts. Consequently, bulk and shear moduli increase
as a result of better grain contacts in water-saturated clays.
Differences in consolidation among~dry sandstcnes are nearig
overwhelmed by the effects of water saturation.
The differing responses of shear and bulk moduli G and K
BUtLlJING SANDSTONE
GULF SANDSTONE in relation of water saturation versus clay content indicate
P-SANDSTONE that the differences between V,/V, ratios for saturated and dry
CLEAN SANDSTONE / s j
samples increase with increasing clay content,

5s
ACKNOWLEDGMENT
G
The authors thank C. Tosaya and D. Yale for helpful
suggestions to start this study. Thanks are also due to
B. Turner, A. Christina, L. Yale, and K. Velasco for collecting
data from thin-section analysis5SEMI and petrographic micro-
photography and analysis, and data-base management.
Thanks are also due to B. Quinn for extensive reviewing of the
final draft.
1!33.0 ZO.0 73.0 300.0 350.0 This research was supported by a grant from the Division of
MEA. K MOD. (. 1 GPo)
Engineering, Mathematics, and Geosciences, Office of Energy
Research, U.S. Department of Energy.

FIG. 13. Computed bulk moduli for fully saturated samples

from velocity data of dry samples (assumed to be independent
REFERENCES
of frequency; Winkler, 1985) using Gassmanns relation versus
measured data on 63 samples. The bulk modulus for rock
matrix is calculated from equations, and that for water is 2.2 Benzing, W. M., 1978, V,/V, relationshipsin carbonatesand sand-
GPa. Deviations of data from the solid line can be considered stones:laboratorydata: Presentedat the 48th Ann. Internat. Mtg.,
Sot. Explor. Geophys.,San Francisco.
estimates of bulk modulus dispersion with frequency, from Biot, M. A., 1956, Theory of propagationof elastic waves in fluid
zero to the megahertz range. saturatedporoussolid:J. Accoust.Sot. Am., 28, 168-178.
 Wave Velocities in Sandstones 2105

Birch, F., 1960, The velocity of compressional waves in rocks to 10 Mavko, G. M., and Nur, A. M., 1978, The effect of nonelliptical
kilobars, 1: J. Geophys. Res., 65, 1083-t 102. cracks on the compresstbility of rocks: J. Geophys. Res., 83, 4459-
Costagna, J. P., Bat&, M. L., and Eastwood, R. L., 1985,. Relation- 44hX
ship between compressional wave and shear wave veloctttes in clas- Minear. M. J., 1982, Clay models and acoustic velocities: Presented at
tic silicate rocks: Geophysics, 50, 551-570. the 57th Ann. Mtg.. Am. Inst. Min. Metall. Eng., New Orleans.
De Martini, D. C., Beard, D. C., Danburg, J. S., and Robinson, J. H., OConnell. R. J.. and Budiansky, B., 1974, Seismic velocities in dry
1976, Variation of seismic velocities in sandstones and limestones and saturated cracked solids: J. Geophys. Res., 79, 5412-5426.
with lithology and pore fluid at simulated in situ conditions: Proc. Nur, A., and Simmons, G., 1969. The effect of saturation on velocity
EGPC exploration seminar. in low porosity rocks: Earth Plan. Sci. Lett.. 7, 1833193.
Domenico. S. N., 1984, Rock lithology and porosity determination Pickett. C. R., 1963, Acoustic character logs and their application in
from shear and compressional wave velocity: Geophysics, 49, 11% formation evaluation: J. Petr. Tech., 15, 6599667.
1145 Quirein, J. A.. Baldwin, J. L.. Terry, L.. and Hendricks. M., 1981,
Eastwood, L. R., 1983, Basis for interpretation of V,,V, ratios in com- Fstimation of clay types and volumes from well log data-an exten-
plex lithologies: Presented at the Sot. Prof. Well Log Anal. 24th sion of the global method: Presented at the Sot. Prof. Well-Log
Ann. Mte.. naner G. Anal. 22nd Ann. Mtg., paper Q,
Gassmann,-F:, 1951, Elastic waves through a packing of spheres: Rafavich. F.. Kendall. C.H.St.C.. and Todd, T. P., 1984, The relation-
Geophysics, 16,6733685. ship between acoustic properties and the petrographic character of
Geertsma, J., 1961, Velocity log interpretation: the effect of rock bulk carbonate rocks: Geophysics, 49, 162221636.
compressibility: Sot. Petr. Eng. J., I, 235248. Raymer. D. S., Hunt. E. R.. and Gardner, J. S.. 1980, An improved
Geertsma, J., and Smit, D. C., 1961, Some aspects of acoustic wave sonic transit time-lo-porosity transform: Presented at the Sot. Prof.
propagation in fluid-saturated porous solids: Geophysics, 26, 169- Well Log Anal. 2 1st Ann. Mtg., paper P.
181. Robert. C. W.. 1975, CRC handbook of chemistry and physics: F-18.
Gregory, A. R., 1977, Aspects of rock physics from laboratory and log Robert, S. C., 1982, 1982 CRC handbook of physical properties of
data that are important to seismic interpretation: Am. Assn. Petr. rocks: 2, 213.
Geol. Memoir 26, 15-46. Talham, R. H., 1982, FO/r( and lithology: Geophysics, 47, 336344.
Johnson, W. E., 1978, Relationship between shear-wave velocity and Toksoz, M. N., Cheng, C. H.. and Timur. A., 1976, Velocities of
georechnical parameters: Presented at the 48th Ann. Internat. Mtg., seismic W~YCS in porous rocks: Geophysics. 41,621-645.
Sot. Explor. Geophys., San Francisco. Tosaya. C., 1982, Acoustical properties of clay-bearing rocks: Ph.D.
Kevin, B. H., 1981, Factors affecting acoustic compressional velocities dissertation. Stanford univ
and an examination of empirical correlations between velocities fosaya, C., and Nur. A., 1982, Effects of diagenesis and clays on
and porosities: Presented at the Sot. Prof. Well Log Anal., 22nd compressional velocities in rocks: Geophys. Res. Lett., 9. 58.
Ann. Mtg., paper PP. Walsh, J. B., and Grosenbaugh, M. A., 1979, A new model for ana-
King, M S., 1966, Wave velocities in rocks as a function of changes in lyzing the effect of fractures on compressibility: J. Geophys. Res.,
overburden pressure and pore fluid saturations: Geophysics, 31, 84 1537-1536
S&-73. Winkler. K. W., 1985, Dispersion analysis of velocity and attenuation
Koaallis, B., Jones, L. E. A., and Wang. H. F., 1984, Velocity- in Berea sandstone: J. Geophys. Res., 90,679336800.
porosity-clay content: systematics of poorly consolidated sand- Wyllie. M. R. J.. Gregory, A. R., and Gardner, L. W., 1956, Elastic
stones: J. Geophys. Res., 89, 10355510364. wave velocities in heterogeneous and porous media: Geophysics,
Kuster, G. T.. and Toksoz, M. N., 1974, Velocity and attenuation of 21,41~ 70.
seismic waves in two-phase media, I : Theoretical formulation: Geo- -__ 1958, An experimental investigation of factors affecting elastic
physics. 39. 587 606. wave veloctties in porous media: Geophysics, 23,4599493.
2106 Han et al.

APPENDIX

Values of bulk density D,, clay content (volume percent), porosity $ for the 75 sandstones of this study, at differential
compressional velocity V(km/s), shear velocity V,(km/s), and pressures of 5, 10, 20, 30, and 40 MPa.

- -
AMP T- 5 HPa 10 MPa 20 HPa 30 HP0 40 MPa I
DW :LhY
vs

I;
9 Q
- P
I s vp
i- 233 3. 4.2E 2.53 2.84 .I831 4.64 2.91
2 2.31 3.00 4.08 2.35 2.66 .I996 4.40 2. 72
3 2.53 3.00 5. 15 3. 17 3.49 .o652 5.47 3.60
4 2.39 3.00 4.61 2.91 3.06 .1544 4.78 3. 10
5 2.32 3. 00 4. 18 2.59 2.~91 .~1990 4.43 2_05
6 2.25 . 10 3. 43 2.02 .2302 3.49 2.10 .2377 3.5@ 2.16 .2368 3.64 2.20 .2361 3.68 2.22 I2355
7 2.24 . 16 3.02 1.72 .2635 3.15 1.81 .2625 3.22 1.91 ,261s 3.29 1.97 .2605 3.36 1.99 I2597
e 2.24 . 10 3.35 1.92 .2432 3.47 2.03 .2425 3.58 2.10 .2416 3.64 2.15 .2409 3.69 2. 17 ,24G3
9 2.38 .2B 3.51 I.88 . 1640 3.59 1.96 . 1629 3.71 2.02 .I611 3.77 2.05 .I599 3.82 2.07 I1589
10 2.45 .06 4.51 2. a5 . 1091 4.62 2.89 .loe2 4.66 2.93 .I072 4.70 2.96 .I063 4.73 3.00 I1056
11 2.23 .04 3. SE 2.01 .2334 3.74 2.17 .2324 3.84 2.29 .2314 3.89 2.33 .2305 3.92 2.35 I2297
12 2.38 .03 4.40 2.62 . 1580 4.45 2.69 .I572 4.51 2.74 . 1562 4.55 2.78 .1553 4.60 2. 81 1546
13 2. 47 .05 4.35 2. 56 . 1090 4.50 2.68 . IOel 4.61 2.90 .1071 4.68 2.85 ml062 4.73 2. 89 1056
14 2.18 .06 3.56 1.98 .2575 3.63 2.01 .2568 3.69 2.05 .2556 3.72 2.07 ,2546 3.74 2.08 82536
15 2.53 . 07 4.90 2.94 . 0447 4.99 3.02 .0439 5.09 3.09 .0427 5.16 3.14 .0419 5.23 3.17 .0412.
16 2.41 .27 3.67 1.94 . 1289 3.79 2.03 .1280 3.93 2.15 . 1270 4.01 2.20 . 1262 4.06 2.24 51256
i? 2. 36 .06 4.02 2.33 . 1829 4.13 2.44 . 1823 4.22 2.52 .I817 4.26 2.54 .18:1 4.30 2.57 I1007
18 2.25 . 16 3.24 1.81 .2590 3.34 1.91 .2583 3.43 1.98 .2573 3.49 2.02 .2564 3.54 2.05 ,2557
19 2.50 .06 4.81 3. 00 .0582 4.83 3.04 .0579 4.89 3.08 .0575 4.91 3.10 .0572 4.94 3. 12 ,0569
20 2. 47 . 14 3.85 2. 14 . 1339 3.96 2.21 .1332 4.07 2.31 . 1322 4.17 2.37 . 1315 4.23 2.41 >1309
21 2.35 .06 4.03 2.35 .I785 4.14 2.46 .I777 4.23 2.55 . 1771 4.28 2.59 .I765 4.32 2.62 ,I761
22 2.28 .04 3. se 2.06 .2100 3.73 2.18 .20921 3.91 2.30 .2084 3.98 2.36 .2070 4.03 2.40 ,2072
23 2.34 .05 3.79 2. 16 . 1908 3.94 2.28 .I930 4.08 2.41 .I891 4.14 2. 47 I* lees 4.18 2. 50 ,le83[
24 2.57 . 09 4.41 2.69 .0935 4.51 2. T7 ,. 0928 4.60 2.86 a0920 4.65 2.90 .0916
4.69 2. 94 IO912)
25 2.57 .08 4. 65 2. 82 .0940 4.72 2.91 .0934 4.83 1. se .0928 4.85 3.02 .0924 4.8~ 3. 05 IO920
26 2.27 .03 3.58 2.15 .2397 3.69 2.24 .2389 3.79 2.30 .2380 3.94 2.34 .2374 3.89 2.37 -2369
27 2.34 .06 3. 71 2.15 .I928 3.92 2.33 .I923 4.04 2.43 .I913 4.11 2.48 .1907 4. i5 2.51 -1903
28 2.30 .03 3.66 2.13 .2194 3.77 2.23 .2184 3.86 2.32 .2176 3.91 2.37 .2170 3.95 2.39 -2165
29 2.28 .06 3.70 2.11 .2246 3.85 2.26 .2236 3;Q.t 2.33 .2226 4.00 2.37 .2219 4.03 2.40 -2213
30 2.31 .09 3.73 2.23 .I927 3.86 2.35 . 1917 3.99 2.45 .I905 4.04 2.50 . 1895 4.08 2.54 .1887
31 2.51 . 13 4. 18 2.40 -0902 4.34 2.53 .OB7R 4.48 2.65 .0057 4.58 2.75 .0848 4.62 2. BO .0835
32 2.57 . 13 4.30 2.42 .0682 4.52 2.52 .066G 4.57 2.66 .0636 4.76 2.75 .0624 4.77 2.80 no612
33 2.55 . 12 4.32 2.38 .0755 4.57 2.87 .0735 4.72 3.03 .0712 4.77 3.19 .0700 4.78 3.23 .0693
34 2.54 . 13 4. 45 2.29 .06e9 4.54 2.41 I.0671 4.67 2.49 .0653 4.72 2.60 .0638 4.79 2.67 SO624
35 2.56 . 12 4.63 2.59 .0370 4.80 2.77 .0340 4.95 2.90 .0330 4.99 3.02 .0320 5.00 3. 13 .0313
36 2.61 . 15 4.92 3.06 .0314 5.11 3.14 .0284 5.18 3.18 .0273 5.22 3.23 .0268 5.23 3.26 a0264
37 2.57 .07 4.73 2.61 .0365 4.09 2.75 .0345 4.97 2.97 .0324 5.09 3.07 .0315 5.23 3.09 -0312
30 2.54 . 18 9. 66 2.73 .0450 4.82 2.84 .0425 4.99 2.99 .0404 5.06 3.09 .0395 5.13 3. 13 ,039o
39 2.62 . 15 4. a7 2.90 IO260 5.00 2.95 .0245 5.Oe 3.06 .0235 5.10 3.09 .0229 5.11 3.10 no225
40 2.61 . 15 4. 44 2.51 ,0672 4.53 2.60 .0642 4.61 2.70 .0627 4.67 2.72 .0617 4.68 2.73 IO612
41 2.55 .38 4. 11 2.41 no683 4.23 2.49 .0665 4.30 2.56 .0651 4.33 2.59 .0641 4.37 2.62 -0634
42 2.56 .40 4. 04 2.30 -0762 4.10 2.36 .0747 4.16 2.41 .0735 4.21 2.46 .0725 4.24 2.49 *0719
43 2.49 .3? 3. 81 2. 13 .I170 3.91 2.22 .1153 3.99 2.28 .I138 4.04 2.30 .I127 4.08 2.34 I IIIB
44 z 53 , SD k 97 2; 29 .0932 4. a6 2; 3% . i3XB~R i6 2; 45 .0963 4.21 2.50 .0692~4:24 2; 52
45 2.55 .35 3. B9 2.20 .0983 3.97 2.28 .0964 4.05 2.35 .0947 4.12 2.40 .0935 4.17 2.43
46 2.57 .4s 4.0: 2.3c .0739 4.1; 2.39 .0713 4.22 2.48 .0697 4.2E 2.53 .0685 4.3; 2.57
47 2.41 . 13 3.92 2.23 .I437 4. IC 2.39 .1425 4.31 2.54 .I413 4.4C 2.60 . 1407 4.4; 2.64
48 2.42 . 14 3. 9E 2.28 . 1668 4.1c 2.39 .I656 4.21 2.47 . 1646 4.28 2.53 . 1637 4.3: 2. SE
49 2.3E . 10 3.81 2.13 . 1593 3.9E 2.26 .I582 4. IC 2.39 .I571 4.18 2.45 . 1565 4.2~ 2.51

- . 11 . 1772 2.18 .I760 4.11 .I741

2.36 2.4:
-50 -3.7E 2.04 3.9 2.31 .1747 4.17 2.38 4.2:
 Wave Velocities in Sandstones 2107

DW XhY
I-
# P vs Q P vs @ P VS Q P vs
51 2.38 .16 3. 2.06 . 1739 3.91 2.32 .1716 4.13 2.38 -1701 4.19 242
52 240 . 4~4 3; 4 5. 71 . 1360~ 3.52 i. 96 .i3fL? 266 :.94 3. 7: 1.37
53 2.38 .46 3. 37 1. a1 . 1380 3.44 1.93 ml338 3.59 1.97 , 1323 3.64 1.99
54 2.35 .51 3. 33 1.75 . 1246 3.43 1.94 .1178 3.63 1.98 .ll@J 3.66 201
55 2.09 . 11 2. 86 1.51 .3061 3.01 1.69 -3019 3.16 1. 73 ,300s 3.20 1.75
56 2.12 . I2 2 94 I.57 .3004 2.Q9 1.72 -2970 3.13 1.75 .2Q56 3.17 1.77
57 2.35 .27 3.44 1.72 . 1575 3.60 2.01 .1525 3.80 2.06 .1512 3.99 2.13
58 2.35 .27 3.55 1.76 , 1535 3.70 1.94 .1505 3.03 2.03 .1479 3.93 2.11 , 1467 4.00 2 16 .1454
59 2.20 .22 z93 1.47 .2492 3.08 1.64 .2475 3.24 1.79 .2458 3.31 1.06 .2446 3.36 1.09 .2435
60 2.19 . 12 3.05 1.53 .25QO 3.22 1.71 .2574 3.41 1.84 I2554 3.49 1.90 .2541 3.55 1.94 .a31
61 2.41 .37 3.41 1.79 . 1504 3.54 1.90 .1480 3.65 2.00 , 1458 3.73 2.08 .I440 3.76 2 11 ,I430
62 2.48 . 44 3.59 1.92 . 1146 3.64 2.00 .1129 3.74 2.08 .1113 3.80 2. 13 .1095 3.04 2. 15 .I089
63 2 47 .bl 3.63 1.91 . 1035 3.76 2.03 -0991 3.85 2.12 I0965 3.92 2.16 .0947 3.97 2 19 ,0937
64 2.37 .27 3.65 I. 88 .1506 3.74 2.00 .I493 3.88 2.09 -1461 3.QS 2. 15 *lb45 3.08 2. 19 ,143b
66 2.17 .08 3.27 I. 05 .2675 3.42 2.00 .2656 3.57 2 13 ,264l 3.62 2.17 02632 3.67 220 .2625
66 2.25 .06 3. 15 1.73 .2724 3.33 1.04 .2710 3.50 2.00 a2696 3.56 2.06 .2686 3.61 2.09 m2679
67 2.12 .ll 3. 12 1.66 ,263s 3.28 1.84 .2810 3.46 l.Q8 ,2803 3.52 2.03 .2793 3.58 207 .2785
60 2.17 .07 2.98 1.50 .2705 3.13 1.75 .26Q2 3.33 1.89 -2674 3.43 1.95 .2664 850 l.Q9 .2655
69 2. lb .07 3.04 1.60 .278Q 3.23 1.81 .2774 3.43 1.98 .2760 3.53 2.05 -2750 3.58 2.09 m2742
?a 229 ,l! 3zl2 L-76 1.~91 ,2052 3.69 208 -2037 3.81 2. 17 ,202a 3.88 2.23 .2021
71 2 47 .21 3.71 2.06 2 32 2.42 2.48 ,lOe9
72 2.39 .06 3.96 2. 16 253 2.66 2.73 I 1508
73 2.47 .23 3.91 2. lb 2.46 2.55 2.61 I1021
74 2.64 .24 4. 10 2.28 2.64 2.72 2.77
2.38 3.67 2.32 2.37
-75 - . 18 2.02 2.27

* Dw: saturated bulk density in gmlcc; CLAY: clay volume fraction; vp, vs: velocity in km/s.