Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Electrical Methods

Download as pdf or txt
Download as pdf or txt
You are on page 1of 25

Electrical Methods.

In this section we will look at the third potential field method, DC resistivity, and the related
methods induced polarization (IP) and self potential (SP). The DC resistivity method is a way of
estimating the electrical conductivity of geological formations, and so we will be looking at what
makes rocks more or less conductive. One of the most important factors in rock conductivity is water
content, and so DC resistivity methods are used extensively in groundwater exploration. However,
DC methods also have applications in geothermal exploration (hot water is more conductive, steam
is resistive) and mineral exploration (ore minerals are conductive).
Electromagnetic methods can also be used to study ground conductivity, but that will be the
subject of the next section.

1. Intoductory Theory.

Although in geophysical prospecting we are really interested in the effects of electric currents
flowing in the Earth, electrical methods are founded in a fundamental property of matter called
charge, which is measured in coulombs (C) and comes in two varieties, positive and negative. An
electron/proton has a charge of 1.6 × 10−19 C. Coulomb’s Law gives the force on a charge q1 due
to a charge q2 at a distance R:
q1 q2
F= r
4πo R2
where o is the permittivity of free space; o ≈ 8.85 × 10−12 C 2 N −1 m−2 (or F/m; in cgs units
1/(4πo ) = 1). r is the unit vector in the direction of q2 from q1 . Note that by this definition, like
charges repel and unlike charges attract. The electric field, E, is defined as the force per unit charge
and so the field due to a charge q is
1 q
E= r
4πo R2
and has units of N/C.

P
F
q2

Rr
q1

Fig. E.01: Geometry for Coulomb’s Law.

Potential is defined as the work required to move a point charge from an infinite distance to the
point P , and so a potential field may be related to E. The work required to move a test charge from
infinity to a point P is given by
ZP
VP = − E.dr

The minus sign indicates that we must do work against the electric field to approach a like charge.
Electric potential is measured in NmC−1 , or volts. Note that force is a vector (i.e. it has a magnitude
and a direction) but that potential is a scalar (i.e. it has magnitude only). Both are functions of

SIO182 (February 5, 2016): Electrical Methods 1


three dimensional space. Dealing with potential is often easier than dealing with the field directly.
The electric field is obtainable from the potential by differentiation:

E = −∇V

so the common unit of E is V/m.

The gradient operator, ∇ , equals

∂ ∂ ∂
x +y +z
∂x ∂y ∂z

where x, y, z are the cartesian unit vectors. Applied to a scalar, ∇ becomes gradient and
results in a vector:  
∂U ∂U ∂U
∇U = , ,
∂x ∂y ∂z

Dotted into a vector, ∇ becomes divergence and is a scalar:

∂Vx ∂Vy ∂Vz


∇·V= + + .
∂x ∂y ∂z

Crossed into a vector, ∇ becomes the curl or rotation and is another vector:

x y z
∂ ∂ ∂ ∂Vz ∂Vy  ∂Vx ∂Vz  ∂Vy ∂Vx 
∇ × V = ∂x ∂y ∂z = x − + y − + z −
∂y ∂z ∂z ∂x ∂x ∂y
Vx Vy Vz

The potential is conservative, that is, it doesn’t matter what path the charge takes to go from
point to point, the total work done is the same. Potential fields are also additive, or linear, so the
effect of several charges may be obtained by integrating the effect of all the constituent charges.
When we considered gravity and magnetics we considered distributed regions of mass or mag-
netization, but this is not our interest here because we will be concerned with electric current,
which is the flow of charge. Current in a wire is measured in amperes, 1 A being 1 C/s, but in a
three dimensional medium, charge flow has to be measured by current density, J, having units of
Cm−2 s−1 or Am−2 . Note that J has a direction as well as magnitude.
Charged particles in a material are in constant thermal motion, colliding with each other and
moving in all directions, resulting in no net current flow. However, in an electric field the charges
will have a slight preferred direction, resulting in a drift velocity. The charges will usually come
to rest in a very short time after the field has been removed, because the time between collisions is
very short, on the order of 10−14 s for free electrons in a metal (see below). The drift velocity is
proportional to E, so current density is also proportional to E;

J = σE

SIO182 (February 5, 2016): Electrical Methods 2


This expression is Ohm’s Law. The constant of proportionality, σ, is called conductivity, and
depends on the number of charge carriers per unit volume, their charge mangitude, and their
mobility. Conductivity has the unit S/m (a siemen is a reciprocal ohm) and is a measure of how
easily an electric current is generated by an electric field. It is common, especially in geophysics,
to consider resistivity, ρ, the reciprocal of conductivity. We see that resistivity has units of Ωm.

2. Resistivity of rocks and minerals.

We have seen that electrical conduction in a material is accomplished by a movement of charge,


which in practice means the movement of electrons, holes (the absence of an electron), ions or ion
vacancies. Materials may be classified on the basis of the nature of the charge carriers and the type
of movement.
Metallic conduction is due to ‘free’ electrons, that is electrons which do not take part in crys-
tal bonding, in a metal. This form of conduction is very efficient, resulting in room temperature
resistivities for metals of about 10−7 Ωm. The resistivity of metallic conductors decreases with
temperature because the thermal agitation of the conduction electrons, which impedes their move-
ment in response to an electric field and therefore lowers the drift velocity. In the extreme case
of superconductors, at sufficiently low temperatures resistivity goes to zero. Massive deposits of
native metals are extremely rare in rocks, and so metallic conduction is not of direct importance to
prospecting. (Of course, the technology used to build prospecting apparatus depends very much on
metallic conduction.) However, fine grains of metallic conductor dispersed through a porous rock
play an important role in the induced polarization effect, described later.
Semiconduction is the result of behaviour between metallic conduction and conduction in “in-
sulators". There are no free electrons as such in a semiconductor, but an applied electric field can
supply enough energy to move electrons into higher energy levels, where they are free to conduct. As
a result of these few electrons, and the ‘holes’ they leave behind in the lower energy levels, moderate
conductivities are achieved, between about 10−5 and 105 Ωm. Although rock-forming minerals are
not semiconductors at surface temperatures, semiconduction is important in geophysical prospect-
ing because many of the sulphide minerals are semiconductors with quite low resistivities, making
ore bodies detectable using electrical methods.
Solid electrolytic conduction is the result of moving ions and ion vacancies through a crystal
structure. As may be imagined, it is very difficult to move ionically bonded ions through a crystal
lattice, so the resistivities of silicate minerals are very high; quartz has resistivities between 1010
and 1014 Ωm. Halite has a resistivity of 1012 Ωm or so. It would, in fact, be close to impossible
to move ions through a perfect lattice, but defects in the lattice allow the motion of ions and holes.
Three types of defects are illustrated in the diagram below. Schottky defects are lattice vacancies
and Frenkel defects are ions in interstitial positions. Impurities also constitute defects in the lattice.
Shottky and Frenkel defects are thermally induced, so the number of defects, and hence conductivity,
increases with increasing temperature. Impurities do not depend on temperature, but rather account
for the variation in resistivity between different samples of the same mineral. Because smaller ions
such as Mg++ and Fe++,+++ are more mobile than larger ions such as Na+ and K+ , mafic rocks are
slightly more conductive than felsic rocks, but for most geophysics rock conductivity is dominated
entirely by water content.
Electrolytic conduction is the more familiar movement of dissociated ions through a liquid
or solution. Melting of a solid electrolyte vastly increases the mobility of the ions, and so the

SIO182 (February 5, 2016): Electrical Methods 3


Dislocation Impurity

Frenkel defect

Schottky defect

Fig. E.02: Three different types of defects found in ionic crystals.

resistivity drops by an order of magnitude or more. However, unless you are engaged in the search
for geothermal resources, this will not be important, as most near-surface rocks are well below
their melting points. Of much greater significance are the low resistivities of salt solutions. Ions in
solution are very mobile so while the resistivity of pure water is very high, the addition of a soluble
ionic salt lowers the resistivity greatly. One molar NaCl solution has a resistivity of 0.22 Ωm;
the resistivity of sea water is between 0.25 (near surface) and 0.30 Ωm (below the thermocline)
depending on the temperature; to a good approximation the conductivity is seawater is

σ(t) = 3 + t/10 S/m

where t is temperature in Celsius. Increasing the concentration of salt also increases the conductivity
of salt solution. The resistivity of ground water lies between 0.1 and 1000 Ωm and is usually between
10 and 100 Ωm.
Because the resistivity of silicate minerals at surface temperatures is very high and the resistivity
of ground water is, on the other hand, quite low, electrical conduction in most surface rocks is
through water contained in pores, cracks and fissures. Archie’s law gives an empirical relationship
between water content and rock resistivity:

σr = σw φ+a or ρr = ρw φ−a

where ρw is the resistivity of the water, φ is the fractional content of water in the rock (equal to the
porosity if the rock is saturated). The exponent, a, is related to how well the pores are interconnected,
and may take values between 1 and 3, but is usually about 2. We see that as water content goes
up, rock conductivity goes up, and as water conductivity goes up, so does the rock conductivity.
This simple form of Archie’s law with an exponent of 2 can be a useful guide to porosity given
resistivity, but if reliable results are required, the exponent (and another multiplicative constant
called formation factor) must be established by making laboratory studies of a representative suite
of rocks from the area being considered.
The resistivity of wet rocks is extremely variable, varying over many orders of magnitude as
the porosity and groundwater conductivity changes. In general, igneous rocks are more resistive
than sedimentary rocks, whose resistivity goes up with consolidation and which are generally more
resistive than unconsolidated sediments. Clays are almost always relatively conductive, around
10 Ωm at least. One needs to beware of extrapolating from the porosity and resistivity of hand

SIO182 (February 5, 2016): Electrical Methods 4


samples to the resistivity of the bulk rock. For porous sediments this extrapolation may be quite
good, but for rocks of lower porosity most of the water may be contained in porosity of a larger
scale than may be sampled with a specimen, that is cracks, joints, fissures, faults and so on.
The range in the resistivity of rocks is from about 1 Ωm for clays and porous, salt water saturated
sandstones to about 106 Ωm for relatively unfractured igneous rocks.

Rock type porosity resistivity


Soil, unconsolidated rock 0.1–0.5 1–1000 Ωm
Consolidated sedimentary rock 0.01–0.4 10–10000 Ωm
Igneous rocks 0.001–0.1 100 – 106 Ωm
Metamorphic Rocks 0.001–0.1 100 – 106 Ωm
Metalliferous ores – 0.1–100 Ωm

Table 2.01: Typical resistivities and porosities. More comprehensive tables


can be found in Telford et al.; these are coarse guidelines only. Apart from ore
bodies, the main factor is porosity and groundwater salinity, and so initial porosity,
intergranular mineral growth, and weathering are all contributing and cumulative
factors.

3. Measuring rock resistivity in the laboratory.

For a simple series circuit, such as a piece of wire, Ohm’s law becomes

V = IR,

where V is voltage (actually a potential difference), I is electric current and R is resistance.


However, it is resistivity, ρ, rather than resistance, that is the intrinsic property of a material. If we
consider a regular solid, resistivity and resistance are related by

R = ρL/A

where L is the length of the prism and A is the cross sectional area of the base.

∆V

Length = L

I
Area = A

Fig. E.03: Measuring resistivity of a regular prism.

SIO182 (February 5, 2016): Electrical Methods 5


This forms the basis of measuring rock resistivity: cut a regular solid from the rock, such as a
core or a cube, apply current electrodes to the ends and see how much current may be passed for a
given voltage, that is, ρ = ∆V.A/I.L. This is called the two electrode method, and works well if the
contact resistance of the current electrodes can be kept low. If the rock is resistive, however, it is
quite likely that the contact reistance of the electrodes will be higher than the resitivity of the rock,
giving erroneous results. The solution is to devise a method which ignores the contact resistance,
and so the four electrode method was devised, whereby the potential difference is measured across
the center of the rock rather than the ends. For a high impedance voltmeter the electrode resistance
is not critical, and the resistivity can be computed from ρ = ∆V.A/(I.s) where s is the potential
electrode spacing. In Figure E.04, Rr represents the resistance of the rock specimen and Rc
represents an electrode contact resistance, either current or potential.

Two-electrode method Four-electrode method

V V
I I

Rc Rr Rc Rc+1/3Rr 1/3Rr Rc+1/3Rr

Rc Rc
V
V
I I
Fig. E.04: Two- and four-electrode methods of measuring rock resistivity, showing equivalent
circuit elements. In the two-electrode method, the potential drop across the current electrodes
is measured along with the potential drop across the sample. In the four-electrode method, the
potential drop across the current electrodes is not part of the voltage measuring circuit, and the
potential electrode impedance simply adds to the already high input impedance of the voltmeter.

Although the D.C. resistance is usually required, the current source must be low frequency (1 Hz
to 1 kHz) A.C. to avoid polarization (note that when we consider field experiments, this frequency
will be lower because the length scale is larger). Polarization results from a build-up of charge
carriers on the ends of the sample or within the sample itself, inhibiting further flow of current. This
property is actually exploited as an exploration method (induced polarization) described later.
Anisotropy. It is possible that the resistivity of a mineral, rock, or rock formation depends on the
direction of measurement. Thus the resistivity measured along the c crystal axis may be different
to that measured along the a crystal axis. The coefficient of anisotropy, λ, is the ratio of maximum

SIO182 (February 5, 2016): Electrical Methods 6


to minimum resitivity, thus
ρmax
λ= .
ρmin
Mineral (micro) anisotropy is not likely to be of any importance in prospecting, but (macro)
anisotropy due to bedding structure or metamorphic foliation is common, where the alignment of
non-spherical grains or microcracks produces a preferred direction for current flow. This alignment
is most often controlled by gravity, and therefore produces the situation in which ρz > ρx , ρy (i.e.
horizontal resistivity in either direction is less than vertical resistivity). This is horizontal anisotropy.
Interbedding between otherwise isotropic sedimentary units of different resistivity will also produce
electrical anisotropy on a much larger scale. The terms macroanisotropy and microanisotropy are
often used to indicate the scale on which the anisotropy is caused.

4. DC resistivity method

The measurement of the Earth’s resistivity is very similar in concept to the laboratory resistivity
measurement of rock samples. A DC electric current is passed through the ground via a pair of
current electrodes and a resulting potential difference is measured between a second pair of potental
electrodes. It is common to designate the current electrodes A and B and the potential electrodes
M and N. Although the two-electrode method is occasionally practiced in the laboratory, it is never
used in the field because the resistance of the current circuit is always dominated by the electrode
contact and the material immediately surrounding the electrodes.
Theory. As in all potential methods, we consider a point source first and later combine point
sources to model more realistic situations.

I
Air: conductivity = 0

current density

Earth: constant,
isotropic resistivity equipotentials

Fig. E.05: Current I being injected into homogeneous ground from a single electrode. Current
density is directed radially out from the electrode and equipotentials are hemispherical.

Consider a single current electrode, considered to be a point source of current, on the surface of
a homogeneous, isotropic halfspace, injecting a current I into the Earth (Figure E.05). The flow of
electric current will be radially symmetric in the halfspace. This should be intuitively obvious, but
is also a consequence of the two boundary conditions which must be satisfied at a contact between
materials of differing conductivity, namely that the tangential electric field must be continuous and

SIO182 (February 5, 2016): Electrical Methods 7


the normal electric current density must be continuous. Since there is no current flowing in the
atmosphere, there must be no normal component of J at the Earth’s surface, i.e. J is tangential at
the surface.

(A planar surface may be described by a vector, S, whose magnitude is the area of the
surface and whose direction is the (outward) normal to the surface. Thus in the above case
ds describes the surface element.

To compute the E-field inside the earth we will balance the current flowing into the earth at the
electrode with the total current flow out of a hemispherical surface. The total current across the
hemisphere must be equal to I because there are no sources or sinks of current other than I (i.e.
charge is conserved). Because of the radial symmetry, the current density will be constant at a
distance r from the current electrode, so the total current flow across the hemispherical surface will
be Z
I= J · ds = 2πr2 J
hemisphere

(ds is defined in the box.) The integration is simply the surface area of the sphere times the (constant)
current density which is always normal to the hemisphere. So, we have that at any distance r, the
current density is
I
J= r
2πr2
where r is the outward normal to the hemisphere defined by r. Now substituting the above expression
into Ohm’s law gives
I
r = σE
2πr2
or using the definition of resistivity and rearranging

ρI
r=E .
2πr2
To obtain the potential at a distance R we integrate the electric field from infinity to R:

ZR ZR
ρI ρI R
VR = − E.dr = − dr =
2πr2 2πr ∞
∞ ∞

SIO182 (February 5, 2016): Electrical Methods 8


General 4-electrode array
N

A (+)
B (-)
M

Fig. E.06: The general 4-electrode array. M and N are potential electrodes, A and B are current
injection electrodes.
ρI
= .
2πR
Now for the general four-electrode array: we have that the potential at electrode M is simply
the sum of the effects of the two current electrodes:
 
Iρ 1 1
VM = −
2π AM BM
and similarly the potential at N is
 
Iρ 1 1
VN = −
2π AN BN
so the potential difference measured across M N is
 
Iρ 1 1 1 1
∆V = VM − VN = − − +
2π AM BM AN BN
Iρ 1
=
2π k
where k is called the geometric factor (we will see why 1/k is used soon):
 
1 1 1 1
k = 1/ − − +
AM BM AN BN

Apparent resistivity: What we have derived above is often called a forward problem in geo-
physics; that is, given a possible structure for the Earth (in this case an isotropic, homogeneous
halfspace of resistivity ρ), we can compute what measurement would be made over that Earth (in
this case ∆V for a given combination of I and electrode geometry). The forward problem tells us
how to go from an Earth structure to a theoretical measurement. However, in real life we will have
the measurement and want to know what the structure of the Earth is (after all, that is the point of
making the measurement in the first place). The mechanism for going from a measurement to a
theoretical structure is called the inverse problem. In the above case the inverse problem is very
easy and unique:
∆V
ρ = 2π k .
I
SIO182 (February 5, 2016): Electrical Methods 9
That is, given a measurement of ∆V the above expression would correctly yield the resistivity of
a homogeneous, isotropic halfspace. If the Earth were not a homogeneous, isotropic halfspace the
above expression would not yield the resistivity of the Earth. However, to aid interpretation, an
apparent resistivity is defined for any measurement over any Earth structure as
∆V
ρa = 2π k
I
so although this resistivity is not likely to represent the truth, it is a sort of ‘average’ resistivity. The
apparent resistivity is the first step in the analysis of electrical sounding data; ρa is computed for
every measurement and is then further interpreted for more complicated structures than halfspaces.
Sounding and Profiling. There are two possible objects to using electrical methods. Electrical
sounding is an attempt to determine the resistivity of the Earth as a function of depth. Sounding
makes the assumption that the structure is locally one dimensional (1D), that is resistivity varies
only with depth and does not change significantly in lateral directions on the scale of the sounding
array (i.e. ‘layer cake geology’). It is accomplished by making the elecrode array successively larger
between measurements, making the depth of penetration successively deeper. Profiling, on the other
hand, has the explicit object of determining lateral changes in resistivity and hence structure. It is
accomplished by moving a fixed size array across the ground.
It is clear that ρa can easily be computed for any arrangement of electrodes, but the further
interpretation is complicated and depends heavily on previous workers’ efforts at solving difficult
mathematical problems, so in practice only a few specific geometries are used:
The Schlumberger array is the most commony used arrangement, and was developed by Conrad
Schlumberger in the 1930’s. Like most of the standard arrays it is colinear and symmetrical. The
particular feature of the Schlumberger array is that the potential electrode spacing is very much
smaller than the current electrode spacing.

Schlumberger array: AM = BN >> MN

A (+) M N B (-)

L = AB/2

Fig. E.07: The Schlumberger array. M N is small compared with AB and the array is symmet-
rical and colinear.

To compute the expression for apparent resistivity for this array we firstly note that for a
symmetrical array AM = BN and BM = AN so that
1 2 2
=( − )
k AM BM
Now for the Schlumberger array we write
BN = AM = AB/2 − M N/2 = (AB − M N )/2

SIO182 (February 5, 2016): Electrical Methods 10


and
AN = BM = (AB + M N )/2
so
1 4 4
= −
k AB − M N AB + M N
(AB + M N ) − (AB − M N ) 8M N
=4 =
(AB − M N )(AB + M N ) (AB − M N )(AB + M N )
or
(AB − M N )(AB + M N )
k=
4M N
yielding an expression for apparent resistivity:

π∆V
ρa = (AB − M N )(AB + M N )
4.I.M N
Because AB >> M N we may set AB ± M N ≈ AB and also write E ≈ ∆V /M N so

AB 2
 
πE πE
ρa ≈ (AB)2 =
4.I I 2

The Wenner array was developed by Frank Wenner in the U.S. at about the same time that
Schlumberger developed his techniques in France. Wenner worked in the national standards labo-
ratory on material properties, and realised that he could apply the same four-point method used in
the laboratory in the field to measure bulk Earth resistivity. Consequently, the array that bears his
name has equal spacings between all electrodes.

Wenner array: AM = MN = NB = a

A (+) M N B (-)

a a a

Fig. E.08: The Wenner array. The array is symmetrical and colinear but now the electrodes are
equally spaced.

From our expression for a symmetric array we have very simply that 1/k = 2/a − 2/2a = 1/a
so that k = a and
∆V
ρa = 2π a
I
Finally we have the dipole-dipole array, which comes in two varieties, equatorial and polar. We
have for the polar array
πr3 ∆V
ρa = ·
I.AB M N
SIO182 (February 5, 2016): Electrical Methods 11
and for the equatorial array
πr3 ∆V
ρa = 2 ·
I.AB M N
Observe that as in magnetism, the dipole field falls of as r3 and the polar field is twice as strong as
the equatorial field. ∆V /M N is approximately E, the electric field, and I.AB is the dipole moment
of the transmitter, and is a meausure of power supplied to the Earth.

Equatorial Dipole-Dipole array


A (+) M
r N
B (-)

Polar Dipole-Dipole array


B (-) A (+) M N
r

Fig. E.09: The dipole-dipole arrays. Both the current electrode pair and potential electrode pair
form dipoles which are separated by a distance which is large compared to the dipole length.

5. Electrical sounding.

In an electrical sounding the electrode array is systematically made larger while the center of the
array remains fixed over the area of interest. As the array gets larger, the electric currents flowing
deeper and deeper in the Earth are sensed, and so the resistivity of deeper and deeper structure is
measured.
An apparent resistivity is computed for each electrode spacing and plotted as a function of the
electrode spacing, both the resistivity and spacing scales being logarithmic. This plot is the sounding
curve. Logarithmic scales are used because a) the range in resistvity of Earth materials is more
than 5 orders of magnitude and b) the resistivity method is only sensitive to structure which is of
comparable size to its depth of burial.
Roughly speaking, the apparent resistivity curve represents the true resistivity as a function of
depth, smoothed by the response of the experiment. Thus, if the Earth were composed of a 10 Ωm
layer 30 m thick overlying a 1000 Ωm layer, when the electrode spacing is much less than 30 m
the Earth looks like a 10 Ωm halfspace to the experiment and the apparent resistivity would truely
represent the resistivity of the top layer. When the electrode spacing is much greater than 30 m the
apparent resistivity is the resistivity of the lower layer. At electrode spacings of about 30 m both
layers influence the apparent resistivity and the sounding curve makes a smooth transition between
the two layer resistivities.

SIO182 (February 5, 2016): Electrical Methods 12


A MN B
105

10 Ωm 30 m
Apparent Resistivity, Ωm

104
10000 Ωm

103

102

101

1 10 100 103 AB/2 (m)

Fig. E.10: Electrical sounding. The electrode spacing is made progressively large, sensing
progressively deeper structure. Apparent resistivity is plotted against electrode spacing to give a
sounding curve.

For the Schlumberger array, ρa is plotted as a function of AB/2, for the Wenner array, a, and
the dipole-dipole array, r.

6. Electrical profiling

Electrical profiling consists of moving an array of fixed geometry across the surface of the Earth
in an attempt to resolve lateral changes in resitivity. As in sounding, the depth of penetration of the
electric current is comparable to the electrode spacing, but now the electrode spacing is chosen to
be sensitive to the sort of structure being investigated. Thus, if the resistivity of an aquifer at a depth
of 20 m were to be mapped, an AB/2 or a of 30 - 50 m would be used. For profiling the apparent
resistivity is plotted as a function of the lateral position of the center of the array. By its nature,
resistivity profiling is less quantitative than sounding.
A combination of sounding and profiling can be conducted to produce a pseudo-section. A
dipole-dipole array is moved laterally, but the source-reciever spacing is also changed. The apparent
resistivities are plotted on a cross section at a depth of half the electrode spacing and beneath the
center of the array. A sort of two-dimensional picture is built up which qualtitativly indicates
resistivity structure as a function of both depth and lateral position. Great care must be exercised
when interpreting data presented as pseudo-sections, because there is no guarantee that the strucure
of the real Earth will be reflected when the data are plotted this way.
Electrode effects in electrical sounding. The object of electrical sounding is to estimate resistivity
as a function of depth. For the results to be realistic the true structure must be, for the most part,
only a function of depth, or one-dimensional (1D). We know, of course, that this will never be
perfectly true, but in many circumstances the 1D approximation is good enough, for example when

SIO182 (February 5, 2016): Electrical Methods 13


Apparent Resistivity, Ωm

Wenner Array
Schlumberger Array

1 Ωm 10 m

10 Ωm/0.1Ωm

AB/2, meters

Fig. E.10b: Electrical sounding. Sounding curves of the Schlumberger and Wenner arrays are
different over the same structure.

mapping sedimentary strata with little deformation and relatively small dips, or mapping the depth
to basement, where again the structure of the sediment/basement interface is mostly horizontal.
Remember that the 1D approximation only has to hold for scales on the order of the largest
electrode spacings. Thus variations in the depth, say, to a horizon may be mapped by multiple
soundings spaced well apart.
While electrical sounding is quite tolerant to deviations from the 1D approximation that occur at
depth (it is possible to map beds with slopes of up to 15o ), variations in resistivity near the surface,
where, of course, the electrodes are, are capable of producing large effects in the sounding curve.
When an electrode is moved and placed over a region of near-surface heterogeneity, a distorted
measurement will be made. The distortion depends on whether a current electrode or potential
electrode is placed on an inhomogeneity. For obvious reasons distortions are called current (or AB)
electrode effects and potential (or MN) electrode effects. Current electrode effects are not as severe
as potential electrode effects, because small inhomogeneities only serve to alter the effective shape
of the current electrode, and as long as the distorting structure is small compared with the electrode
spacing the point source approximation will hold. On the other hand, a distortion of the electric field
near the potential electrodes will have a large effect, no matter how far apart the current electrodes
are.

SIO182 (February 5, 2016): Electrical Methods 14


ρ1

resistivity

ρ2
distance

A M N B A M N B

ρ2 ρ1 > ρ2

Fig. E.11: Electrical profiling. An resistivity array is moved along the surface while holding the
dimensions fixed. The apparent resistivity is plotted as a function of distance to give a resistivity
profile or map.

Resistivity Pseudo-section
M N A B

5.0 6.0 6.5 7.1 8.0 9.0 9.5 8.6 7.0

5.1 6.2 6.8 7.5 8.4 9.1 8.8 7.5


6.5 7.1 7.6 6.2 6.1 8.1 7.7
7.3 5.0 4.0 4.5 7.1 8.0

8.5 7.0 6.5 7.5 9.0

Fig. E.11A: Electrical pseudo-section. The equally spaced stations across the surface of the
earth are all occupied by dipole transmitter and receiver arrays. For each transmitter/receiver pair,
the apparent resistivity is plotted mid-way between the stations and at a depth proportional to the
station spacing. The apparent resistivities can be contoured and interpreted graphically, or modeled
quantitatively.

How may electrode effects be recognized? A sounding curve for a truely 1D earth must be
smooth; just as in gravity and magnetics, variations due to distant structure are broad on nature.
Thus scatter between adjacent points on a sounding curve, is not due to sloppy measurement
procedure, is due to electrode effects. This type of ’noise’ is easy to handle because it is random in
nature and we can draw a smooth curve through the scattered data for interpretation (which is not,
in fact, the best way to handle such noise but gives an idea of its manageability nevertheless). For a
Wenner or dipole-dipole sounding, where the potential electrodes are moved at every measurement,
scatter due to MN electrode effects is common and there is little that may be done to eliminate it. For
the Schlumberger array, however, the MN electrodes do not have to be moved, and so MN electrode
effects are not a problem. Well, nearly not a problem. The trouble is, one may start a sounding with

SIO182 (February 5, 2016): Electrical Methods 15


A1 M2 M1 N1 N2 B1 B2
A2

Current electrode effect Potential electrode effect


Fig. E.12: Electrode effects during a Wenner sounding. While the deeper structure is 1D,
variations in the surface layer (say, in the thickness of the weathered layer) cause effects when both
the current and potential electrodes are moved. The effect on the data is smaller for the current
electrodes, because they are spaced further apart and the distortion is a small part of the total current
path.

a MN spacing of about 0.5 m and an AB spacing of a few metres, but by the time the AB spacing is
up to a few hundred metres the signal on the 0.5 m dipole is very small. Now, the potential gradient
in the central 1/3 of an array is very nearly linear (see diagram on p. 637 [525] of Telford et al.), so
all that needs to be done to increase the signal strength is to increase the potential electrode spacing,
because even with MN equal to a few tens of metres it will still be small compared to the larger
AB, as required by the Schlumberger array. But once the potential electrodes are moved there can
be distortion due to MN effects. The solution, always carried out for Schlumberger sounding, is to
make repeat measurements at 2-4 current electrode spacings every time the MN dipole needs to be
enlarged. The result will then be a sounding curve which looks like that in figure E.13.
Sections of sounding curve where the MN electrodes did not move are free of MN electrode
effects, but between these sections any distortion due to moving the MN electrodes is clearly visible.
The procedure now is to simply slide the sections of the curves vertically so that they all line up
to produce one sounding curve. The section of curve with the largest MN spacing is usually taken
to be the most reliable, since for the larger spacing any small variations in surface resistivity have
proportionately least effect. Also, what is the mathematical effect of sliding the data vertically on a
log-log plot? The ability to remove MN electrode effects in this way accounts for the popularity of
the Schlumberger array.
More severe trouble arises when the Earth departs from the 1D approximation in a more sys-
tematic, or large-scale fashion, because then the source of the distortion may affect the current
electrodes at several adjacent spacings. The danger then is that the electrode effect, which has
now graduated into what may be better called a 3D effect, will be incorrectly interpreted as 1D
structure. One danger sign is that for a truely 1D earth, an ascending part of the sounding curve
may not exceed 45◦ in slope (on our log-log plots). To see why this must be so consider the most
extreme transition; a perfectly conductive layer underlain by a perfect resistor. The current will be
trapped in the conductive layer, and so instead of the electric field falling of as the inverse square
of distance, as it does in a 3D medium, we essentially have a 2D medium and the electric field falls
off as the inverse of linear distance, that is ∆V ∝ 1/AB. When this is substituted into our formula
for apparent resistivity for the Schlumberger array, for example, which says ρa ∝ ∆V.AB 2 , we get
that ρa ∝ AB. That is, double AB and ρ doubles too, and so there is the origin of the 45◦ slope.

SIO182 (February 5, 2016): Electrical Methods 16


Raw data:
105

Apparent Resistivity, Ωm MN=30m


104

103
MN=3m
102
MN=0.3m
101

1 10 100 103 AB/2 (m)

Corrected data:
105
Apparent Resistivity, Ωm

104

103

102

101

1 10 100 103 AB/2 (m)

Fig. E.13: Correction for electrode effects in a Schlumberger sounding; three different MN
spacings have been used (0.3, 3 and 30 m), with an overlap between each (i.e. both MN spacings
measured with the same AB). The offsets between the different segments of the curves are removed
by sliding the curves vertically. Usually, the largest MN is taken to be the most accurate.

(What would the slope be on a linear-linear plot?) Although not very common, an ascending slope
of > 45◦ on a sounding curve is a sure indication of 3D structure. A descending part of the curve
may be very steep, however.
A very good practice is to test for 1D structure during the actual sounding. This may be done in
one of two ways. A second electrical sounding may be completed at 90◦ to the first, and of course
the two soundings should give the same result for a 1D earth. However, apart from requiring a
lot of work, the 1D approximation holds better when a sounding is carried out with the electrode
array along the strike of any 2D structure than across it. Thus the crossed sounding, if it crosses the
structural strike also, may show the effects of 2D structure while the first sounding is interpretable

SIO182 (February 5, 2016): Electrical Methods 17


using the 1D approximation. The second method to test for inhomogeneity is to move only one
current electrode at a time as the array expands. Thus, say that the current electrodes are at an AB/2
spacing of 100 m, and that the next measurement is to be made at a spacing of 150 m. Instead of
moving both current electrodes to 150 m as would normally be done, move only one to 150 m and
make a measurement using this asymmetrical array. Then move the second electrode and make a
normal measurement. The apparent resistivity of the 100/150 m measurement should fall between
ρa for the 100/100 m and 150/150 m spacings; in fact it is possible to combine the 100/100 m
and 150/150 m measurements mathematically to predict the 100/150 measurement. Any structure
which is affecting only one end of the sounding array (i.e., non 1D structure) will show up as an
inconsistency in the measurements (see Zohdy, 1978, Geophysics 43, 748-766).

7. Equipment.

The equipment for the electrical sounding or profiling may in principle be very simple; a set
of dry cell or lead-acid batteries provides a current source, insulated wire connects the electrodes,
which may be merely metal pegs driven into the ground, an ammeter measures the current and a
voltmeter measures the potential differences. Very useful work may be done with such equipment,
but there are refinements which improve the precision and accuracy of the measurements.
Current circuit. The bulk and cost of the current source for the AB circuit will be measured by the
power of the transmitting unit (V.I), yet the parameter that determines the size of the signal across
the potential electrodes is only current (I). Thus we have an easy optimization problem: maximize
I for a given power (V.I). If the resistance of the AB circuit is R, then the power will be simply
P = I 2 R so I = (P/R)1/2 . We see that the resistance of the current circuit must be minimized.
Unless the cable connecting the AB electrodes is very long, most of the AB resistance will actually
be at the grounding electrodes. Because of the inverse square law we have already examined, the
grounding resistance will be determined by the resistivity of the material in immediate contact with
the electrode and with what amounts to the surface area of the electrode. AB electrodes may be
(a) metal (aluminum, steel, brass) pegs driven into the earth, (b) metal foil buried in a shollow pit,
or (c) an extensive antenna of buried copper braid. The latter is only used for a relatively fixed
installation such as a very large dipole-dipole spread. The choice between (a) and (b) depends on
several factors. The foil is potentially a better ground, but digging pits is harder work, and a peg may
be more effective if the nearsurface material is dry and resistive because a 1 m peg is more likely
to reach down to a zone of capillary water. 10 − 100 Ω represents the typical range of grounding
resistance. It is perfectly normal to water the electrode with a salt solution to lower the resistance
of the immediate area, but then the grounding resistance is likely to change with time as the water
seeps into the soil. The one restriction to be observed with current electrodes is that they do not
violate the point source approximation used in the computation of apparent resistivity. That is, the
size of the electrode should be small compared with the current electrode separation AB. Thus, a
1 m long stake would not be driven fully into the ground until AB was about 10 m or more.
Because the transmitter circuit is likely to have a resistance of 100 Ω or more, a fairly high
transmitter voltage is required to drive a significant current through the ground. 100 V would be
typical and up to 1000 V possible. Voltage higher than this would too hazardous to use and produce
problems with leakage. The power source may be dry cell or lead-acid batteries or a portable
generator connected to a DC power supply. Although the DC behaviour of the ground is required,
in practice the current is low frequency AC or switched to reverse polarity every few seconds. This
prevents polarization of the current electrodes and removes any DC offset between the potential

SIO182 (February 5, 2016): Electrical Methods 18


electordes (see below).
Potential circuit. Stainless steel pegs are commonly used as potential electrodes for simple
electrical soundings, but they are quite noisy. This is because metal in contact with damp ground
corrodes, and corrosion is an oxidation rection which has associated electrical potentials which are
fairly large. The potential difference across MN may be quite small, and mostly this corrosion noise
makes accurate measurement difficult. The solution is to use non-polarizable electrodes; a solution
of metal salt held in a porous container to which contact is made via a metal rod:
Most commonly copper-copper sulphate is used, but lead-lead chloride are also employed. The
‘porous pot’ electrode is emplaced into the soil so that the porous ceramic is in good electrical
contact. Watering potential electrodes to lower grounding resistance is a risky business because
excessive drift (see below) may develop.
It is important that the device used to measure the potential difference (PD) across MN does not
draw too much current (if it does the potential will be less than the correct value). That is, its input
impedance must be much bigger than the resistance across MN. In bygone days this was ensured
by using a potentiometer, but today voltmeters, with input resitance of 10 MΩ or greater are readily
available.

Contact point

Saturated electrolyte Insulating stopper

Metal rod Porous material (ceramic)

Fig. E.14: Non-polarizable electrode, or ’porous pot’. Electrical contact between the metallic
parts of the AB circuit and electrolytes in the ground is made via a metal-metal salt system, for
example copper and copper sulphate. The compatible metal in the salt, and a saturated salt solution,
ensure that electrochemical reactions on the metal surface are kept to a minimum. Small amounts
of salt solution leaking through the ceramic case provides an electrical contact with the ground.

It will be found that a potential difference across MN will be several mV to several V without
any current in AB. This is because of self potentials (see next section) in the ground, low frequency
currents induced by variations in the Earth’s magnetic field, and DC offsets between electrode pairs.
Using an AC or reversed AB current has the effect of averaging away this DC potential.
Figure E.15 shows a switched or square wave current, but AC has much the same effect. The
+∆V and −∆V values may be measured manually and subtracted to obtain 2∆V , but many modern
instruments will do this manipulation automatically and moreover average (‘stack’) several cycles
transmitted waveform to improve signal to noise ratio.

SIO182 (February 5, 2016): Electrical Methods 19


Voltage

+I

∆V

2∆V
Self Potential

0V

-I Time

Fig. E.15: There will be some voltage on the MN circuit even with no current flowing through
AB (‘a self potential’). To remove this, and double the signal size, the current is reversed in the AB
circuit, to measure (−∆V ) to (+∆V ) rather than 0 to (+∆V ).

8. Interpretation

There are three approaches to the quantitative interpretation of both profiling and sounding
resistivity results; matching pre-computed curves to the data (curve matching), comparing the
computed response of a guessed model to the field data and then adjusting the model to obtain
a fit to the data (interactive forward modelling), and mathematical/numerical inversion schemes
to automatically obtain a subsurface resistivity structure from the data. With profiling data it is
often not necessary to make a quantitative interpretation; the detection of an anomalous region or
plotting a psuedo-section is usually sufficient. With sounding data, however, it is always necessary
to make some kind of quantitative estimation. The theory allowing the computation of synthetic
apparent resistivity curves for a given model consisting of discrete layers in the earth is complicated
but has been developed for a long time. However, the equations relating ρa to a layered structure
(the forward problem) are complicated and it is not easy to produce an expression relating a layer
structure to ρa (the inverse problem). Thus many workers rely on curve matching and iterative
forward modelling.
Curve matching. Pre-computed sounding curves are plotted on transparent paper of the same
log-log scale as the sounding data (or vice-versa). Two-layer curves plot ρa /ρ1 (apparent resistivity
divided by the resistivity of the top layer) as a function of A/h (electrode spacing divided by the
thickness of the top layer) for many different ratios of ρ1 /ρ2 . Because of the logarithmic axes,
the type curves may be slid horizontally and vertically to obtain a match between one of the type
curves and the field data. The position of the point (A/h = 1, ρa /ρ1 = 1) on the field plot then
gives the resistivity and thickness of the first layer. The ratio ρ1 /ρ2 for the matching curve yields
the resistivity of the second layer.
For soundings reflecting 3 layers, the same exercise may be performed with a suite of 3 layer
curves, or repeated application of the 2 layer curves to both the top and bottom parts of the sounding
data can produce satisfactory results.

SIO182 (February 5, 2016): Electrical Methods 20


‘Stacking’. A signal to noise ratio (SNR) is simply how many times bigger the signal is
than the noise obscuring the measurement process. That is, if the signal (2∆V ) is 1 mV and
the natural background and electrode noise is 0.2 mV the SNR is 5. For one measurement the
error will therefore be 20%. If better than 20% accuracy is required several measurements
may be made and averaged. If e is the error in one measurement then the error in a sum of
N independent such measurements is
sX
esum = e2
N

= Ne
so the error in a mean of N measurements is
1 e
emean = esum = √
N N
and we see that the noise decreases as the square root of the number of measurements.
Another way to look at the problem is to realise that as measurements are summed, the
signal grows linearly with N (that is, if 10 measurements of 1 mV are summed then the result
is 10 mV). The noise, on the other hand, is as often negative as positive (i.e. in the above
example the noise will fluctuate around ±0.2 mV) √ and so will tend to cancel.
√ In actual
√ fact,
for ‘well behaved’ noise the noise will grow as N , so the new SNR is N/ N = N .
Stacking is of great importance in many aspects of quantitative science. By its use a signal
well below the noise level can be recovered with as much precision as the experimenter has
time for. The use of stacking is prevalent in all aspects of exploration geophysics, particularly
as modern microcomputers can do the necessary signal processing in real time.

Since the advent of fast computers and fast, simple algorithms for generating 1D sounding
curves, curve matching has become obsolete.
Forward modelling. Modern computers and the development of efficient algorithms for predict-
ing the resistivity response of layered structures makes trial and error or automatic curve matching
using computers quite easy. A guess is made for a layered structure to fit the data and the resistivity
response of the structure computed and plotted on the same graph as the data. The regions where
the computed response fails to fit the data are noted, and with experience or an efficient computer
algorithm the model can be corrected to bring the fits closer. This process is repeated (i.e. one iter-
ates) until a satisfactory fit is obtained, leaving the final model as the estimate of the true structure of
the Earth. If the Earth really is 1D with the same number of layers as was used in the interpretation,
this method produces very accurate results. If the assumptions are wrong, of course, the final model
may be meaningless.
Inversion. There are schemes where the computer can automatically generate new models to test
against the data, and make estimates of how well the model response fits the data. These inversion
schemes are moderately complicated, and vary both in their efficiency and the type of models they

SIO182 (February 5, 2016): Electrical Methods 21


So how do you compute the resistivity over a layered earth? This isn’t examinable, but
just so you know...
For a Schlumberger array over an earth composed of layers with resistivities ρ1 , ρ2 ...ρN
and thicknesses t1 , t2 , ...tN −1 (counting from the top down), the apparent resistivity is given
by a thing called a Hankel transform:

AB 2 ∞
  Z  
AB
ρa (AB) = T1 (λ)J1 λ λ dλ
2 0 2

where J1 is the first-order Bessel function of the first kind and T1 (λ) is the resistivity transform,
given for the top of each layer by

Ti+1 + ρi tanh(λti )
Ti = .
1 + Ti+1 tanh(λti )/ρi

Notice how the transform at the top of each layer depends only on the resistivity and thickness
of that layer and the transform of the layer below. By starting at the bottom of the heap with
TN = ρN , we can compute TN −1 , ...T2 , T1 in succession. This is called a recurrence relation
and is how one gets T1 at the surface of the earth.
This used to be a very difficult computation to make, because the Bessel function oscillates
as it decay for large λ, making the integral have chunks of positive and negative contribution,
but in 1971 ’fast Hankel transforms’ were discovered.
If we make the change of variables x = ln(AB/2) and y = ln(1/λ), then
Z ∞
ρa (x) = T1 (y)J1 (ex−y )e2(x−y) dy
−∞

and if we define F (x − y) = J1 (ex−y )e2(x−y) then


Z ∞
ρa (x) = T1 (y)F (x − y)dy
−∞

which is a convolution integral with T1 (y) as the input function, ρa (x) as the output function
and F (x − y) at the filter function. Now, F (x − y) is independent of the layer parameters ti
and ρi which appear only in T1 (y).
The convolution may be approximated by a linear filter:

kX
max

ρa = T1 (λk )fk
k=kmin

where fk are the coefficients of a moving average filter. The values for this filter are computed
by other people using P2C2E, but are published and available in lengths from 7 points to
hundreds of points. So what was previously a fiendishly complicated problem has been
reduced to computing the resistivity transform (a few hyperbolic tangents or exponentials)
and a simple sum.

SIO182 (February 5, 2016): Electrical Methods 22


generate. As we proceed through the geophysics course we will discover that there is rarely a unique
solution available to describe a geophysical measurement. Particular trouble in resistivity occurs
if a thin resistive (or conductive) layer is sandwiched between thicker conductive (or resistive)
layers (equivalence), or if there exists several layers of steadily increasing (or decreasing) resistivity
(supression).
In the case of equivalence, it is very difficult to tell apart models with beds of similar transverse
resistance, T = ρt where ρ and t are the resitivity and thickness of a thin resistive layer within a
conductive host, and models with layers of similar longitudinal conductance, S = t/ρ = tσ where
now the thin conductive bed is within a resistive host. These are called T and S equivalence
respectively. Supression occurs in sections of steadily increasing or decreasing resistivity. The
intermediate layers are not resolvable and are usually lost in the interpretation.
If a layer is anisotropic, with transverse resistvity ρt greater than longitudinal resistivity ρl , as is
usually the case, we have the coefficient of anisotropy
p
λ = ρt /ρl

and mean resistivity



ρm = ρt ρl
Anisotropic layers appear as layers of resistivity ρm and thickness λt, that is, depths are overesti-
mated and resistivities are underestimated.

9. Self potential method

The natural potentials in the ground which produce noise for resistivity sounding are themselves
capable of telling us something about subsurface strucure. In particular, there are often large
potentials associated with conductive ore bodies. The nature of the production of the electric
currents responsible for these potentials is not thoroughly understood, but this has not diminished
the value of mineral resources discoved using the method. The field procedure for the self potential
(SP) method is simply to take a sensitive voltmeter and a pair of non-polarising electrodes and,
using one electrode as a reference, map the potential on the ground surface. Negative anomalies of
potential characteristically occur over mineralized zones.

10. Induced polarization method

We mentioned polarization in our discussion of laboratory measurements of electrical conductiv-


ity. It was pointed out that if charge builds up on the surface of a specimen the apparent conductivity
will decrease. If a the voltage driving the DC current is then turned off, a residual voltage will be
seen across the sample for a short time as the charge re-equilibrates. This same phenomenon is seen
in rocks in situ; if a current is passed through the ground while the potential difference is monitored
across a second pair of electrodes, after the current is interrupted there is an initial drop in voltage
followed by a slow decay.
The cause of the polarization is twofold. The charged surfaces of minerals attract ions from the
electrolyte in pores and thereby produce concentrations of charge, which can impede the applied
current flow if the pore spaces are small enough. This is called membrane polarization and is found
everywhere, especially in clayey materials. Electrode polarization, on the other hand, is caused by
conductive mineral grains blocking the pores of the host rock. Charge builds up on the surfaces of

SIO182 (February 5, 2016): Electrical Methods 23


these grains, which also inhibit the movement of ions in the electrolyte filling the pores. Electrode
polarization is of greater magnitude than membrane polarization, and is obviously the reason IP is
used for exploration, as oxides, sulphides and native metals all exhibit this sort of IP effect (as does,
unfortunately, graphite).

Voltage

Vc

Voff

0
0 t1 t2 Time

Fig. E.16: Induced polarization. After the current is suddenly switched off in the AB circuit, the
voltage in the MN circuit drops, but not to zero. The residual voltage decays to zero over a period
of time. The chargeability is measured by integrating the residual voltage between two time t1 and
t2 .

The IP meausurement must somehow quantify the extent of the polarization. This is done either
in the time domain by sharply interrupting a DC current, as illustrated above, or in the frequency
domain by making resistivity measurements at two frequencies, one low enough to make a DC
meausurement and the other high enough to be influenced by polarization effects.
Chargeability, a time domain measure, is the integral of the decay voltage as a proportion of the
continuous voltage before cutoff:
Zt2
1
M= V (t) dt
Vc
t1
M has units of seconds (usually ms), and clearly depends on t1 and t2 , which must be fixed to obtain
comparable results.
Percent frequency effect is a frequency domain measurement, obtained by estimating the apparent
resistivity at low and high frequencies:

P F E = 100(ρdc − ρac )/ρac

Again, the result will depend on the actual frequencies used, which will usually be in the range of
0.1 - 10 Hz.
Metal factor, another frequency domain measure, attempts to normalise the IP measurement by
the DC resistivity:
M F = 2π × 105 (ρdc − ρac )/ρac ρdc

SIO182 (February 5, 2016): Electrical Methods 24


Charged ions

Clay grain

Matrix

Conductive grain

Applied electric field

Fig. E.17: The storage of charge occurs in the ground for two reasons. Firstly, surface charges
on clay minerals attract dissolved ions and impede their flow in response to the applied electric field.
Secondly, conductive grains (metal, sulphides or graphite) blocking pores conduct electricity but do
not allow the flow of electrolytic charge, causing charge buildup near their surfaces, or polarization.

= 103 P F E/(ρdc /2π)

Equipment for IP work is very similar to that for resistivity. Indeed, resistivity measurements
are almost always made at the same time as IP measurements. However, the decay voltages are
smaller than the Vc voltage which gives the DC value, so transmitter currents tend to be larger and
stacking times longer. A motor-driven generator is usually required, whereas resistivity work may
often be accomplished with batteries. The dipole-dipole array is preferred for IP work, for two
main reasons. One is that larger currents are needed, so more time is spend on reducing the AB
grounding impedance and then one does not wish to move the AB electrodes. The other is that
psuedo-sections are a common way of presenting the results.
Results are presented as profiles, but psuedo-sections are favoured in IP studies. Quantitative
modelling is rare, but can be accomplished in conjuntion with 2D resistivity modelling.

SIO182 (February 5, 2016): Electrical Methods 25

You might also like