GEOPHYSICS, VOL. 34, NO. 1 (FEBRUARY 1969), P. 87-100, 8 FIGS.

, 2 TABLES
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

INDUCED POLARIZATION OF SYNTHETIC, HIGH-RESISTIVITY

ROCKS CONTAINING DISSEMINATED SULFIDES?

W. J. SCOTT* AND G. F. WEST*

The induced polarization and resistivity of a with increase in current density. A background
suite of 74 synthetic, low porosity, “rock” speci- chargeability of a few milliseconds, virtually
mens were measured by the pulse transient indistinguishable from the sulfide IP, was found
method. The specimens contained up to 8 percent in nonmineralized specimens and assumed to be
by volume of sulfide particles in three different present in all. Chargeability due to metallic
grain sizes. A measuring cycle of 5 set on and 5 minerals was found to be proportional to sulfide
set off was employed, and discharge potentials content (volume percent) and to decrease with
were recorded for 4 set by oscilloscope photog- increasing grain diameter, as predicted by the
raphy. Charging current density was varied be- theory developed by Wait. The observed decay
tween 125 w/cm2 and 0.0125 pa/cm2. forms were not exponential as this theory pre-
Chargeability-defined as the time integral, dicts, but were approximately logarithmic. The
from 15 to 415 ms after cessation of charging, of metallic mineral-electrolyte interface impedance
the instantaneous discharge potential divided by determined for the specimens by use of Wait’s
thechargingpotentiaL-was found to vary from 1 theory is cmistent with that measured directly
to 30 ms. For current density less than about by Anderson and Keller for a clean pyrite inter-
0.1 pa/cm2, it was independent of charge current face.
density, but decreased about 20 percent/decade

INTRODUCTION trolyte activity (Keevil and Ward, 1962) but,

The measurement of induced polarization (IP) although the effects of the concentration and
has been used as a prospecting method for dis- grain size of metallic minerals in a host rock have
seminated metallic minerals since about 1949. been widely discussed, only the experimental
Bleil (1953) referenced earlier IP studies and work of Collett (1959) is available in Western
discussed the state of the art at that time Re- literature. Collett’s measurements were made on
search sponsored by the Newmont Mining Com- natural mineralized and unmineralized specimens
pany was published in a monograph edited by and on samples composed of metallic mineral
Wait (i959), mcf research at MIT was summa- grains disseminated in granuiar host material
rized by Marshall and Madden (1959). Keevil and (andesite) saturated with an e!ectrolyte. From an
Ward (1962) list the references to many labora- experimentalist’s point of view, the measurements
tory studies. of the former suffer from one’s inability to con-
Laboratory experiments divide naturally into trol, or often even to measure, the basic prop-
two groups, one to study normal or background erties of the specimens (such as sulfide type, con-
IP effects (e.g. Mayper, 1959) and the other to tent, grain size) and the latter is a rather crude
study metallic mineral (electrode) polarization or model of an igneous or metamorphic rock. Thus,
overvoltage. There has been much work on the there is need for study of IP on a suite of speci-
effect on IP of varying the electrolyte and elec- mens with reasonably controlled properties, as

t Manuscript receivedby the Editor June 24, 1968; revised manuscript received October 7, 1968.
* Geophysics Laboratory, Department of Physics, University of Toronto, Canada. Mr. Scott is now at the
Geophysics Division, Geological Survey of Canada, Ottawa.

87
 88 Scott and West

well as low porosity and high resistivity. in a specially designed cell, that was filled with the
Collett (1959) and Henkel and Van Nostrand specimen electrolyte, in such a manner that cur-
(1957) reported a nearly linear variation of IP rent could flow through the sample but not around
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

with current density (i.e. constant chargeability), it. The charging and measuring periods were each
for current density of the order of 100 pa/cm2. 5 set and the current polarity was reversed after
McEuen et al (19-S?)Joon- the oth~sl ha& show a each pulse. Current pulses were applied to a
less than proportional increase in IP with current sample for some time prior to measurement to
(i.e. chargeability decreases with increasing cur- ensure that dynamic equilibrium was achieved.
rent) for the same order of current density. A For each measurement the voltage drop across the
study of a simple pyrite electrode (Anderson and sample during the charging pulse, and the charging
Keller, 1964) shows that, as current density is current were recorded. The transient voltage
diminished from a value greater than 100 pa/cm2, across the specimen after termination of each
the electrode resistance increases to a maximum, charge pulse was displayed on an oscilloscope and
then decreases slightly and becomes nearly con- photographed. From these data, the resistivity
stant below a current density of about 10 pa/cm2. and the induced polarization of the specimen were
It appears, therefore, that there is some incon- calculated.
sistency in the published reports and that a study The quantity used in most of this work to de-
of the behavior of IP over a wide range of current scribe the induced polarization is the chargeability
density would also be useful. M, which is defined as:
1 Pi2
EXPERIMENTAL PROGRAM
M = ; 1 v,(t)&.
A series of induced polarization measurements
was made on a suite of mineralized, synthetic
“rock” specimens over a range at current density where
between 125 pa/cm2 and 0.0125 pa/cm2, by the V, = voltage drop across the specimen during
pulse-transient method. The samples were made charge,
of a high alumina cement that was set under high v,(t) = transient voltage drop across specimen
pressure. 74 cylindrical specimens were used-56 after interruption of charging current,
contained sulfide inclusions and 18 contained tr= 15 ms after interruption of charging
quartz grains. The sulfides were a natural mixture current,
of approximately 70 percent pyrite and 30 percent t2=415 ms after interruption of charging
pyrrhotite by weight. Three grain-sizes were used current,
for the quartz and sulfide inclusions, and, for the and
latter, there were three different abundances M is in units of ms.
(Table 1). All samples were saturated with an
electrolyte of 0.3 millinormal NaCl in distilled The integral of VP was obtained from an oscillo-
water that had a resistivity of approximately scope photograph of the transient by use of a
2 X lo4 ohm-cm. planimeter. The time interval chosen for integra-
Measurements were made on a specimen placed tion is similar to that used in field measurements.

Table 1. Distribution of specimens in the suite Apparatus

The test cell [similar to that used by Mayper
Grain size Quartz _ Sulfides
Tyler mesh 2,% hy (1959)] used for the measurements is shown in
(Diam. vol. “Fop’ 4% by g% by Figure 1. It was made of clear acrylic tubing with
in mm) VOl. vol.
an inside diameter of l$ inches, and was arranged
20-24 with a sleeve joint in the center to facilitate
(.833 to .701) 6 6 6 6 mounting of specimens of varying thickness. The
32-35 ends were closed with threaded plastic caps, and
(.495 to .417) 6 6 6 6 the cell was held together in a long clamp.
Current electrodes were made from graphite
65-100
(.208 to .147) 6 8 6 6 discs and mounted in the end caps of the cell.
The potential electrodes were mounted in acrylic
 IP of Synthetic Rocks 89

NON-POLARIZING
POTENTIAL ELECTRODES

/ \
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Cu GO4 GEL

POROUS PLUG

ELECTROLYTE C

\ GRAPHITE CURRENT ELECTRODES’

FIG. 1. Schematicdiagram of the test cell.

tubes that were cemented into the cell and com- opened, the cell was short circuited. This was
municated with it through small holes in the cell done to discharge the stray capacity of the cell
wall. They were arranged to be at a constant and the charging circuit. In order to be certain
distance from the specimen faces regardless of that this process had no appreciable effect on the
sample thickness. The potential electrodes con- IP of the sample, the length of time of the short-
sisted of a clean copper wire immersed in a satu- ing was varied.
rated copper sulfate solution in a glass tube The polarization potentials were usually ob-
plugged at the bottom with the same high-alu- served directly on a Tektronix 502 scope using the
mina cement that was used in the specimen differential input and a pair of 50 megohm series
manufacture. To avoid contamination of the isolating resistances. When the potentials were
electrolyte in the sample cell, the CuS04 solution very low, a Keithly 603 Electrometer Amplifier
was mixed with sufficient gelatin to give it a with a 200 megohm differential input was em-
rubbery consistency. These electrodes proved to ployed as a preamplifier.
be stable. During the experimental work they
Preparation of Samjdes
exhibited no significant transient voltage, al-
though occasionally a very small steady voltage, Because the purpose in manufacturing samples
which was nullified by a variable bias in the mea- was to simulate natural igneous or metamorphic
suring circuit, was present. The series resistance rocks, it was necessary to produce a sample with
of the potential electrodes when immersed in the low porosity. A method was found to produce
0.3 mN NaCl electrolyte was 0.36 megohms, specimens with porosity less than 6 percent and
which was small compared to the 200 megohm resistivity (when saturated with 0.3 mN NaCl)
input impedance of the measuring circuit. greater than lo6 ohm-cm. It is a modification of
A block diagram of the circuitry used to ob- the method described by Mandel et al (1957).
serve the polarization is shown in Figure 2. De- Specimens were made from Secar 250, a pure
tails may be found in Scott (1965). high-alumina refractory cement supplied by
During the measuring interval of the cycle the Ciment Fondu Lafarge Ltd. The dry cement was
only external discharge path available to the mixed with the appropriate quantity of sulfide
specimen was through the potential circuit whose or quartz, and sufficient distilled water was added
resistance was always at least 100 megohms. to make a thick paste. About one third of the mix
Observation of the potentials began about 6 ms was placed on top of a piston in a strong-walled
after the charging current ceased. For a period of cylinder into which a second piston was inserted
about one ms before the charging circuit was from the top, and a load of about 30,000 lb was
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Scott
and
West
 IP of Synthetic Rocks 91

applied to the pistons with a hydraulic testing were again weighed to determine the weight gain
machine. After ten minutes under load the speci- and porosity.
men was pressed out of the cylinder. Three speci- In preparation for measurement, the specimens
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

mens were made from a mix. were removed from the electrolyte and quickly
Before production of the suite of specimens to mounted in cylindrical jackets of polyester resin
be studied was begun, a series of test samples was with only the end faces exposed. They were then
made using this method. From this series, the soaked in electrolyte until they were measured.
weight of dry cement powder necessary to pro- The measurements were made using the same
duce one cc of sample was determined. Production electrolyte in the cell as that in which the speci-
of a specimen with a desired volume fraction of mens were soaked.
pyrite or quartz was then a matter of combining
Measurement of sample polarization
the correct weight of sulfides (assumed density
5.0 gm/cm3) or quartz (assumed density 2.65 The saturated sample was placed in the cell and
gm/cm3) with the required weight of dry cement the electrolyte was introduced through the poten-
powder. tial electrode mounting tubes. Care was taken to
For every combination of grain size and quan- ensure that no bubbles were left trapped in the
tity, one half of the samples were loaded to 30,000 cell.
lb and one half to 40,000 lb in an effort to vary the The clamped cell was placed in a shielding box;
porosity and, hence, the resistivity. Unfortu- the relay system was set in operation and the
nately the scatter in porosity and resistivity charge voltage increased slowly to its maximum
values was sufficiently large to mask any effect value (500 volts). Cycling was continued until
this procedure might have had. the charging current became constant and until
In order to stabilize the porosity and composi- any bias in the observed potential drop across the
tion of the specimens, they were weighed and sample had disappeared or stabilized. A state of
placed in distilled water at approximately 60°C equilibrium was usually reached within an hour.
for seven days or more to cure. They were then If a bias remained in the potential measuring
weighed and dried, first in air for one day, and circuit at equilibrium, it was cancelled with the
then in vacuum for about one week at a tempera- bias adjustment.
ture commencing at 60°C and rising to 100°C After equilibrium was reached, current was
over the first four days. Experimentation in the measured precisely by noting the voltage across a
early stages of the work indicated that after one 1000 ohm resistor temporarily inserted in the
week of drying in vacuum, no further significant charging circuit. The voltages across the cell cur-
weight loss could be produced. rent electrodes (V,) and across the potential
To ensure uniform distribution of porosity in electrodes (V,) were measured with a Hewlett-
the part of each specimen to be measured, 0.8 cm Packard Model 702A VTVM with an input
was sliced from each end to avoid end irregulari- impedance of 300 megohms. The decay voltage
ties. They were then heated for an additional 24 transient [V,(t)] was then photographed for a
hours in vacuum at lOO”C, cooled, removed and given charging current density. Either four or six
weighed. This weight was used as the dry weight exposures were photographed on the same frame
for porosity calculations. (Figure 3)-one for each polarity of decay for
Saturation was accomplished by evacuating the each of either two or three sweep rates. Two rates,
chamber containing the specimens to a pressure 50 ms/cm (500 ms horizontal extent) and 500
of less than 1 mm Hg for at least 24 hours and ms/cm (5 set horizontal extent), were normally
then introducing electrolyte and leaving the used, but, if a third was required, it was usually
specimens to soak for four hours before opening 10 ms/cm (100 ms horizontal extent). After the
the chamber to the atmosphere. The specimens last sweep was made, the charge voltage was
were. transferred to a flexible PVC plastic tube lowered until the next charging current at which
filled with the electrolyte, which was corked measurements were to be made was reached.
tightly, placed in a pressure chamber and sub- After a suitable time interval, during which the
jected to a nitrogen pressure of 1600 psi for 24 cell current stabilized, the procedure was re-
hours. At the end of this procedure the specimens peated.
 Scott and West
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

FIG. 3. A typical decay curve photograph showing the decay on two time
scales (50 ms, 500 ms/div) and both polarities of charging current. Bv photo-
graphing both polarities, the difficulty of establishing a zero poteniial base
level is eliminated. (Vertical scales are 1500 and 500 mv/div respectively.)

If initial equilibrium was attained in one hour, 4) An IP effect was found in the samples con-
measurement of the induced polarization of one taining only quartz grains. This background
sample at five current levels took about 23 hours. effect was indistinguishable in its decay
The estimated maximum error in a single mea- form from the IP effect due to the sulfides.
surement of chargeability is _t5 percent. Pro- Numerous checks showed that the effect did
vided there is no leakage of current around the originate in the sample and not in the cell or
sample, the error in the resistivity measurements measuring system.
is also estimated to be within k 5 percent. These characteristics make it possible to dis-
cuss the experimental results in two main divi-
EXPERIMENTAL RESULTS
sions. The first is the behavior of the properties at
The induced polarization of each specimen was low current, or what may be termed current in-
measured for five values of current density be- dependent levels, and the second is the variation
tween 125 pa/cm* and 0.0125 pa/cma. These of properties with current density at higher cur-
measurements constitute the main part of the rent.
experimental results. The average chargeability
Low current properties
of each group of similar specimens plotted as a
function of current density is shown in Figure 4. Five quantities were determined for each speci-
The characteristics common to all specimens are men: chargeability, resistivity, porosity, sulfide
these: content, and sulfide grain size. Only the last two
The chargeability is nearly constant be- could be controlled accurately by the specimen
tween current densities of 0.01 and 0.1 preparation method. The interrelation of the
pa/cm2. various quantities is summarized in Figure 5, in
The chargeability decreases by about 20 which the data plotted are averages over all of
percent per decade increase in current den- the specimens with a given sulfide content and
sity between 1 and 100 pa/cm.2 grain size. It is immediately obvious that sulfide
There is no indication in any of the curves content and grain size are the controlling factors
that chargeability will not continue to de- in determining the chargeability of the specimens
crease with increase in current density. (Figure Sf). No clear correlation exists between a
 IP of Synthetic Rocks 93
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

100
. . . . .._._
..- . . . . . . . . . . ..__._.... o .,,,,
” . . . . .._
“.‘ . . . .._._.

-““-.~.-a
,‘....(
.‘.,
.‘.,
.‘.,
. ...
.‘.,
..,
. .....
.‘.,
2x9 2=%5 4x.3 8Y& .‘..
.‘.,
“...,
FINE I 4 7 lo D‘
.‘..
EO- MEciuM
_--___ 2 5 8 II .... - 20

awfSE 3 6 9 12 .‘..
i‘,
\,
.‘.

....

k
‘.
j‘,lO _

0 I I1 ,,,,,I I I ,,I 1 111,111 1 I,,,/,,

0.1 1.0 IO 100 IX0
TOTAL CURRENT (jut)
I I I I I
0125 .I25 125 12.5 125

CURRENT DENSITY ( pohn*)

FIG. 4. Chargeability versus charge-current density. Each curve rePments the average
chargeability of all specimens of the same grain size and sulfide or quartz content.
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
94
Scott
and West
 IP of Synthetic Rocks 95

specimen’s chargeability and its resistivity or cmz) whose thickness is small compared to
porosity, except in the case of the background the radius of the particle.
chargeability found in the quartz-loaded speci- -1. The dielectric film behaves as a simple lossy
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

mens (Figures 5d & e). The data (Figure SC) sug- capacitor having a decay constant a! and a dc
gest a relation between resistivity and porosity specific resistance 70, i.e. q(w) =v&(a+iw)
consistent with Archie’s law [p (Y (porosity)“]. 5. The volume fraction o of the metallic par-
Sulfide content and grain size does not seem to ticles is very small and the separation be-
have a clear effect on the resistivity of the speci- tween particles is much larger than the
mens, although there is a suggestion that small particle radius. Thus, the particles have no
concentrations actually increase resistivity and interaction.
that increasing concentrations decrease it (Figure
5a & b). Such behavior has been noted by Mandel, It then follows from electromagnetic theory that,
et al (1957) and McEuen, et al (1959). if a charging potential V, is applied uniformly
The main induced-polarization effect observed across a part of the medium for a long time the
depends on the sulfide content and grain size of decay of potential after interruption of the charg-
ing current is
the sample, with the fine grains being more polar-
izable than the large grains. If we assume that
the background polarization is additive with the v,(t) = 5 (exp [-(K + 2)at/k]), (2)
sulfide polarization, the chargeability is approxi-
mately proportional to sulfide content (Figure
where k =ap/vo and is called the particle size
5f and also Figure 7 in the next section). Such a
parameter. Wait normalized (2) with respect to
situation is in general accord with Wait’s (1959)
V, and v to define P(t) the normalized decay
theory of particle polarization. A quantitative
response as
comparison of the experimental data with the
theory was made. 2v,(t)
P(t) = ___
9rv,
Low current decay curves

It is possible to consider the dependence of the = & {exp [-(k + 2)cutjk]J. (3)
decay curves, as well as chargeability, on the vari-
ous parameters. Figure 6 shows some representa-
tive decay curves plotted with time linear and The theoretical chargeability M is found from (1)
potential logarithmic, a display that permits and (3) to be
visual comparison with theory. Over the time
interval used in these experiments the decay
curves were most nearly linear when plotted with
M=
&
9kv
+ 2j2
{erp [1-(y) ali]

a logarithmic time axis and a linear potential axis.

On a log-log plot there is a suggestion of two linear - exp[-(~~)&--J . (1)
segments, the second segment beginning at about
0.5 set and having a more rapid falloff than the
first. Equation (4) predicts that chargeability M
should be proportional to sulfide content v, and,
In his theory of particle polarization, Wait
for a particular film decay constant CX,it should
(1959) assumes a homogeneous medium contain-
exhibit a maximum at some value of particle size
ing a distribution of very low resistivity spherical
parameter k. For values of k greater than this,
particles coated with a relatively insulating film.
the chargeability decreases with increasing size
The basic assumptions of the theory are:
parameter.
1. The matrix has a constant resistivity p. Wait’s (1959) theory predicts an exponential
2. Spherical particles of radius a and resistivity decay for the induced polarization voltage, but a
pr<<p are uniformly distributed in the logarithmic decay was observed. It is possible to
matrix. explain this difference by assuming one or more of
3. Each particle is uniformly coated with a the properties of the particles to be statistically
dielectric film of specific impedance ~7(ohm- distributed over a considerable range (for ex-
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

CURVE

I
20
2b STR
2~ CUR
2d SPE
3

4b CU
4a
5

FIG. 6. Normalized decay voltage (log scale) versus time (linear scale) for a
The line 2b shows how exponentials were “fitted” to decay c
 IP of Synthetic Rocks 97

ample, Keller, 1959). On the other hand, Ander- weak T\jaCl solution at a current density below 10
son and Keller (1964) have shown the specific pa/cm2.
surface resistance of a clean pyrite-electrolyte
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Closer examination of the results in Table 2

interface to be nonlinear with current density shows that the values of ~0 vary in a manner
for values of current density greater than about similar to that of specimen resistivity. When
10 pa/cm2. It seems most likely that the decay of specimens of similar resistivity (but differing
the charged particle film is itself a nonlinear pro- particle diameter) are selected, the variation is
cess, even at very low current densities. Prelimi- much less. Since only five decay curves were
nary experiments done in our laboratory onasingle fitted, the evidence is not strong, but it seems that
interface indicate a decay which is more hyper- variations in specimen resistivity do not have the
bolic or logarithmic than exponential. effect predicted for variations in matrix resistivity
To test the other aspects of Wait’s (1959) by the theory. Possibly the observed specimen
theory, one must attempt to fit an exponential resistivity variations reflect the presence or
decay to the observed decay curves. If this can be absence of a few cracks or dry spots in the speci-
accomplished, one hopes to find that the values mens, rather than variations in the average resis-
of Q:and ~0 obtained are independent of the other tivity to be found in the vicinity of a sulfide
specimen properties p, ZIand a. A fitting proce- grain.
dure was carried out on the decay curves mea- As a check to see how well the whole suite of
sured at 0.125 pa/cm2, of five specimens, with specimens conforms to Wait’s model, the theoret-
8 percent sulfide content. The highest sulfide ical chargeability was calculated for specimens of
content was chosen so that the decay curves each grain size and sulfide content under the
would not be significantly contaminated by back- assumptions that cu=O.22 mm (the average of
ground effect. The values obtained are shown in Table 2), VO= 2.5X lo3 ohm-cm2 (Anderson and
Table 2. Since observed chargeability has already Keller, 1964), and p = 3.5 X lo6 ohm-cm (produces
been shown to have a v dependence in accord with best fit of data to theory, and is consistent with
equation (4), it is not necessary to investigate the observed specimen resistivities). The observed
effect of sulfide content. and theoretical data are shown in Figure 7.
The curve-fitting procedure used is subjective. When we attribute the upward displacement of the
An example “fit” is shown in Figure 6 for curve
2a. The method used apparently produced satis-
factory fits because the individual fits to a group
of decay curves for a given sample agreed reason-
ably well with one another. However, the decay
constant (slope) of the fitted line certainly de-
pends on the time range over which the fit is
determined.
Table 2 shows that there is a reasonable con-
sistency in the values of q. and LYof the five speci-
mens. Most notable is that the qo’s are all near
the value of 2.5 X lo3 ohm-cm2 found by Anderson
and Keller (1964) for a clean pyrite interface in

Table z

Speci- Curve No. ’

men (Figure 6) mm
meg:hm-
cm
ozk- seF_1
cm2

:E 32a 0.22
0.09 3.57
1.88 3.500
1800
0.21
0.23
SULFIDES (vol.%)

157B 4a 0.38 2.26 2700 0.26 FIG. 7. Theoreticaland observedchargeability(aver-

153B not shown 0.09 1.86 2000 0.23 ages)as a function of sulfidecontent and grain size.The
158B not shown 0.22 1.88 2200 0.25 only sourceof appreciablediscrepancyis the presence
155B not shown 0.38 1.86 2100 0.31 of a backgroundpolarizationwhich seemsadditive with
the metallic polarization.
 98 Scott and West

curves to background polarization, the corre- compatible with those of Anderson and Keller
spondence is excellent. for large and small values of current density.
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

S@ecimelt current density cannot be compared

Curre’elttdependewe of churgeabilily
directly with inlerJuce current density, but the
The observations of fall-off in chargeability two cannot differ by more than about an order
with increasing current density were summarized of magnitude. The fact that the ranges of linearity
in Figure 4. For current density less than 0.1 differ only by a factor of ten in the two experi-
pa/cm2 the chargeability is not a function of cur- ments is, therefore, remarkable. The decrease of
rent density, while for values greater than q. at the higher values of current, found by
1 pa/cm2 it decreasesas current density increases, Anderson and Keller, implies a decrease in charge-
regardless of specimen parameters. Individual ability of as much or more than that observed.
decay curves show no obvious change in shape or However, an increase in chargeability which
rate of decay with increasing current density. The corresponds to the increase in qo that they found
effect is apparent only as a change in amplitude. at intermediate values of current density, was not
Anderson and Keller (1964) found 70 to be observed. Perhaps the reason for the difference
independent of interface current density at values lies in the fact that their interface was scrupu-
less than 10 pa/cm2. Increasing the current den- lously cleaned, where no special precautions were
sity produced a rapid increase in 70 to about three taken with the sulfides used in this work.
times the initial value, followed by a steady Besides the variation in the chargeability, a
decline towards low values at a rate of about a slight decrease of resistivity with increasing cur-
factor of five per decade increase in current. rent density was observed. Two examples are
The results of our experiments are entirely shown in Figure 8.

I I I ~~ -..
,125 125 125 !5
CURRENT DENSITY (pafcm2)

FIG. 8. Two examplesof specimenresistivity versuschargecurrent density. The points at the

lowestcurrent density are not very reliable becauseof uncertainty in the current value.
 IP of Synthetic Rocks 99

Hnckground polarization 2. Chargeability increases about 20 percent

The investigation of the background polariza- /decade as charge current decreases, until a
tion effect was not a major part of this work. low-current limit is reached at approximately
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

However, a few points are worth mentioning. 0.1 pa/cmz.

The background chargeability observed in the 3. Low-current chargeability increases with
quartz-loaded specimens had the same depen- decreasing sulfide grain diameter, within
dence on current density as the sulfide particle the range of sample parameters employed.
polarization (Figure 4). The background charge- 4. Low-current chargeability increased linearly
ability apparently increases with increasing with sulfide content.
Consideration was given to the theory devel-
quartz grain size (Figure Sf). This is opposite to
the relationship for metallic particles. However, oped by Wait (1959) for identical metallic spheres
there is also a strong relationship, in the quartz- coated with a thin insulating film, and its predic-
loaded specimens, between grain size and porosity tions were compared with the experimental ob-
(positive correlation) and grain size and resistivity servations with the following results:
(negative correlation). Thus, it is impossible to The decay curve is not exponential as pre-
dicted.
decide which factor is causative. The results are,
however, consistent with Marshall and Madden’s Theoretical decay curves force-fitted to the
(1959) model of membrane polarization. If experimental decay curves gave reasonable
values for the parameters v0 and (Y.
decreasing porosity implies decreasing capillary
diameter, then the restricting effect of a cation- The variation of low-current chargeability
selective zone would be increased, and so would with sulfide content and grain size was in
the concentration gradients established by a fixed good agreement with the theoretical predic-
tions.
current density over a fixed interval. Thus,
It was noted that decreasing the time inter-
Figure Se, chargeability versus porosity, may
show the important relationship. val over which the decay voltage is inte-
The decay curves of the quartz-loaded speci- grated reduces the value of the particle size
mens have the same general form as those of the parameter ap/qo at which the low-current
sulfide-loaded specimens, although there may be a chargeability is a maximum.
tendency for a slightly more rapid fall-off, as the As a general conclusion, it seems that labora-
example in Figure 6 shows. tory measurements of the induced polarization of
Considerable precaution was taken during the specimens can be made at sufficiently low current
manufacture of specimens to ensure no contami- levels to make the results applicable to the field
nationof thematrixcementwith extraneous metal- situation.
lic minerals. Polished sections of non-mineralized ACKNOWLEDGMENTS
specimens showed no trace of such material.
Support of the experimental program from
Likewise, the apparatus was carefully tested to
grant Al 187 of the National Research Council of
make sure that effects such as polarization of the
Canada is gratefully acknowledged.
current electrodes made no appreciable contribu-
tion to the observed voltages. REFERENCES

Anderson, L. A., and Keller, G. V., 1964, A study in in-

SUMMARY AND CONCLUSIONS duced polarization: Geophysics,. v. 29, p. 848-864.
Bleil, D. P., 1953, Induced polartzation-a method of
Measurements of resistivity and chargeability geophysical prospecting: Geophysics, v. 17, p. 636-
were made on a suite of high resistivity synthetic 661.
Collett, L. S., 1959, Laboratory investigation of over-
specimens of varying sulfide content and grain- voltage, in Overvoltage research and geophysical ap-
size over a range of charge-current density from plications, ed. J. R. Wait: New York, Pergamon
125 pa/cm2 to 0.0125 pa/cm2. The following Press, Chapter 5.
Henkel, J. H., and Van Nostrand, R. G., 1957, Experi-
observations were made. ments in induced polarization: Trans. AIME (Min-
ing Engineering), v. 9, p. 355-359.
1. For the time interval of recording (0.01 to 5 Keevil, N. B., Jr., and Ward, S. H., 1962, Electrolyte
set) the decay voltage after interruption of activity and its effect on induced polarization: Geo-
physics, v. 27, p. 677-690.
the charge current is approximately a linear
Keller, G. V., 1959, Analysis of some electrical transient
function of log t. measurements on igneous, sedimentary and meta-
 loo Scott and West

morphic rocks, in Overvoltage research and geo- York, Chapters 10A and 10B.
physical applications, ed. J. R. Wait: New York, McEuen! R. B., Berg, J. W., Jr., and Cook, K. L., 1959,
Pergamon Press, Chapter 7. Electrical properties of synthetic metalliferous ore:
Mandel, P., Jr., Berg, J. W., and Cook, K. L., 1957, Geophysics, v. 24, p. 510-530.
Downloaded 02/24/16 to 128.111.121.42. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Electric properties of synthetic metalliferous ores: Scott, W. J., 1963, Experimental measurement of in-
Geophysics, v. 22, p. 398-411. duced polarizatron of some synthetic metalliferous
Marshall, D. J., and Madden, T. R., 1959, Induced samples at low current densities: M. A. Thesis, Dept.
polarization-a study of its causes: Geophysics, v. of Physics, University of Toronto.
24, p. 79&816. Wait, J. R., 1959, A phenomenological theory of over-
Mayper, V:, Jr., 1959, The normal effect, Part I and voltage for metallic particles, in Overvoltage research
Part II, zn Overvoltage research and geophysical ap- and geophysical applications, ed. J. R. Wait: Perga-
plications, ed. J. R. Wait: Pergamon Press, New mon Press, New York, Chapter 3.