Skip to content

Induced Polarization

Induced polarization for CSIR NET geophysics: electrode and membrane polarization, chargeability, frequency effect, metal factor, Cole-Cole, pseudosections.

  • 18 min read
  • Updated
On this page · 13 sections
  1. Prerequisites
  2. The Hook: The Voltage That Will Not Switch Off
  3. What Causes Induced Polarization
  4. Measuring IP in the Time Domain
  5. Measuring IP in the Frequency Domain
  6. Field Surveys and Pseudosections
  7. Memory Tricks
  8. Don’t Confuse
  9. Common Mistakes
  10. Practice
  11. Quick Revision Table
  12. Key Takeaways
  13. What’s Next

Prerequisites

The Hook: The Voltage That Will Not Switch Off

How CSIR NET asks it. Part C Geophysics: the two polarisation mechanisms, definitions and units of chargeability, frequency effect and metal factor (with a calculation), the link between time and frequency domains, the pseudosection plotting point, EM coupling, and which targets IP finds that resistivity misses.


What Causes Induced Polarization

Induced polarization (IP) is the capacitor-like storage of charge in the ground when current flows, caused by blocked or slowed ion movement. Two mechanisms are recognised.

Electrode (grain) polarization

Where a pore channel is blocked by an electronically conducting grain (sulphide, graphite, some oxides), current must change from ionic to electronic at one face of the grain and back at the other. That transfer needs an electrochemical reaction, which is slower than the arrival of ions. Ions pile up on the grain faces, forming a double layer and an overvoltage. When the current stops, the ions diffuse back and the overvoltage decays.

  • Strongest when metallic grains are disseminated (many grain surfaces exposed to pore water), not massive.
  • The response depends on the total grain surface area more than on the volume of metal.
  • Gives the largest IP effects and is the basis of IP for base-metal exploration.

Membrane (electrolytic) polarization

In rocks without metallic grains, clay minerals carry a negative surface charge that attracts a cloud of cations. Where pore throats are narrow and lined with clay, the cation cloud acts as an ion-selective membrane: cations pass, anions are held back. Current flow builds up ion concentration gradients across these zones, which relax when the current stops.

  • Strongest in rocks with a modest, scattered clay content; very clay-rich rocks give less, because the pathways become uniformly charged.
  • Gives a weaker effect than electrode polarization, but it is the “background” IP of almost every rock, and the response used in groundwater and environmental IP work.
Electrode polarizationMembrane polarization
Causeionic–electronic transfer at a metallic grainion-selective zones at clay-lined pore throats
Needselectronically conducting mineralsclay minerals, narrow pores
Size of effectlargesmall to moderate
Typical targetdisseminated sulphides, graphiteclay content; distinguishing clay from saline water

Measuring IP in the Time Domain

The transmitter sends a square wave with off-times (for example on +, off, on −, off), each part typically lasting one to a few seconds. During the on-time the receiver reads the primary voltage \( V_p \) (from which apparent resistivity is computed exactly as in a DC survey). After switch-off it samples the secondary (decay) voltage \( V_s(t) \).

graph LR
    A["Current on: V rises toward Vp
(charging)"] --> B["Current off at t = 0:
instant drop"] B --> C["Slow decay Vs(t)
(discharge of polarised ground)"] C --> D["Chargeability:
Vs/Vp or area under decay / Vp"]

Chargeability

The chargeability is defined in two ways.

Instantaneous ratio (at a chosen delay \( t \)):

$$M = \frac{V_s(t)}{V_p}\qquad\text{(dimensionless; quoted in mV/V or per cent)}$$

Integral chargeability (area under the decay curve over a window \( t_1 \) to \( t_2 \)):

$$\boxed{M = \frac{1}{V_p}\int_{t_1}^{t_2} V_s(t)\,dt}\qquad\text{(unit: ms, or mV·s/V)}$$

The integral form averages noise and is what most modern receivers report, often for several windows so the decay shape is recorded. Values from different instruments are comparable only if the windows and the waveform are the same; quote them with their timing.

Seigel’s definition and apparent chargeability

Seigel described IP as a change of effective resistivity: a polarisable medium of chargeability \( m \) behaves at DC (after full charging) as if its resistivity were \( \rho/(1 - m) \) instead of \( \rho \). Over non-uniform ground the measured apparent chargeability is a sensitivity-weighted sum of the true chargeabilities:

$$m_a \approx \sum_i \frac{\partial\ln\rho_a}{\partial\ln\rho_i}\,m_i, \qquad \sum_i \frac{\partial\ln\rho_a}{\partial\ln\rho_i} = 1 .$$

The weights are the same sensitivities that control apparent resistivity. A polarisable body therefore shows up in \( m_a \) diluted by how much of the current actually passes through it. With weights 0.7 (barren cover, \( m = 5 \) mV/V) and 0.3 (mineralised zone, \( m = 40 \) mV/V), \( m_a = 3.5 + 12 = 15.5 \) mV/V.


Measuring IP in the Frequency Domain

Because the ground “charges up” over time, its apparent resistivity measured with a slowly alternating current is higher than with a rapidly alternating one: at high frequency the polarisation does not have time to build up. Resistivity falls with increasing frequency.

Frequency effect (resistivities at a low frequency \( \rho_{dc} \) and a higher one \( \rho_{ac} \), commonly about a decade apart):

$$\boxed{FE = \frac{\rho_{dc} - \rho_{ac}}{\rho_{ac}}}, \qquad PFE = 100 \times FE\ \ (\text{per cent frequency effect}).$$

Metal factor (normalises the frequency effect by resistivity):

$$\boxed{MF = A\,\frac{\rho_{dc} - \rho_{ac}}{\rho_{dc}\,\rho_{ac}} = \frac{A\cdot FE}{\rho_{dc}}}, \qquad A = 2\pi\times10^5 .$$

The idea behind MF: a sulphide body tends to have both high IP and lower resistivity, so dividing by \( \rho \) enhances it. But dividing by resistivity also enhances anything conductive, so MF must be read with both FE and \( \rho_a \).

Phase (complex resistivity) and spectral IP

Measured as a complex quantity, resistivity has a small negative phase: the voltage lags the current. Phase IP reports this phase angle (usually in milliradians) at one or more frequencies. Spectral IP (SIP) measures amplitude and phase over several decades of frequency, which carries information about grain size and texture, and can help discriminate sulphide types and remove coupling.

The standard empirical description is the Cole–Cole model (as adapted to IP data by Pelton and co-workers):

$$\boxed{\rho(\omega) = \rho_0\left[1 - m\left(1 - \frac{1}{1 + (i\omega\tau)^c}\right)\right]}$$
ParameterMeaning
\( \rho_0 \)DC resistivity
\( m \)chargeability (0 to 1)
\( \tau \)time constant: controls where in frequency the dispersion happens; tends to increase with polarisable grain size
\( c \)frequency exponent (0 to 1): controls how spread out the dispersion is

Limits. At \( \omega \to 0 \), \( \rho \to \rho_0 \). At \( \omega \to \infty \), \( \rho \to \rho_0(1 - m) \). So the total fall in resistivity is a fraction \( m \): this is Seigel’s chargeability again, and it links the two domains:

$$FE_{\text{full}} = \frac{\rho_0 - \rho_\infty}{\rho_\infty} = \frac{m}{1 - m}\quad\Longrightarrow\quad m = \frac{FE}{1 + FE}.$$

For small \( m \), \( FE \approx m \). (A real survey measures between two finite frequencies, so its FE is smaller than this full-range value.)

Frequency (Hz)0.0111015.91001,00010,000
\( \lvert\rho\rvert/\rho_0 \)0.9960.9660.9140.9010.8520.8180.806
Phase (mrad)−3.4−25.9−44.6−46.0−38.4−18.3−6.6

(Cole–Cole model with \( m = 0.2 \), \( \tau = 0.01 \) s, \( c = 0.5 \).) Amplitude falls from \( \rho_0 \) toward \( 0.8\rho_0 \); the phase peaks at a few tens of Hz, in the neighbourhood of \( f = 1/(2\pi\tau) \approx 15.9 \) Hz (for these parameters the exact maximum, about −46 mrad, falls near 20 Hz).


Field Surveys and Pseudosections

Arrays

ArrayUse in IP
Dipole-dipolestandard for 2-D profiling and pseudosections; symmetric, good lateral resolution; lower signal
Pole-dipolestronger signal and deeper reach for a given cable; asymmetric response
Gradientlarge fixed current electrodes, potential dipole moved over a central area; fast reconnaissance mapping of a wide block
Schlumberger / WennerIP soundings (for example, clay layers in groundwater work)

Plotting a pseudosection

For dipole-dipole, each reading is plotted below the midpoint between the two dipoles, at the intersection of two lines drawn at 45° from the dipole centres. Dipole centres \( (n + 1)a \) apart give a pseudo-depth of

$$z_{\text{plot}} = \frac{(n + 1)a}{2}.$$

So \( a = 25 \) m, \( n = 4 \) plots at 62.5 m. Contoured readings of \( \rho_a \), chargeability or PFE then form a pseudosection. Over a compact polarisable body, a dipole-dipole pseudosection shows a characteristic inverted-V (“pant’s leg”) pattern: the anomaly appears along both 45° lines from the body, because a reading is high whenever either the current or the potential dipole is over the body. Pseudosections must be inverted to obtain a true chargeability section.

Electromagnetic coupling and other noise

Other noise: electrode polarisation at the potential electrodes (use non-polarisable electrodes), telluric currents, and capacitive coupling between cables and ground.

Interpretation and targets

TargetResistivityIPComment
Disseminated sulphides (porphyry copper)modest lowhighclassic IP target
Massive sulphidesvery lowmoderate to highalso a strong EM target
Graphite, graphitic schistvery lowhighthe main false target; geology needed
Pyrite halovariablehighbarren but polarisable: high IP does not mean ore
Claylowmoderatemembrane polarisation
Saline watervery lowlowdistinguishes it from clay
Fresh, clean sand or massive rockmoderate to highlowbackground

Applications beyond minerals: mapping clay layers and distinguishing clay from saline water in aquifers, detecting some hydrocarbon-contaminated and landfill sites, and characterising disseminated mineralisation for engineering.


Memory Tricks

Don’t Confuse

PairFirstSecond
Electrode vs membrane polarizationmetallic grain, large effectclay-lined pores, smaller effect
Chargeability vs frequency effecttime domain: \( V_s/V_p \) or integral (ms)frequency domain: \( (\rho_{dc} - \rho_{ac})/\rho_{ac} \)
Frequency effect vs metal factordimensionless, independent of \( \rho \) scale\( FE/\rho_{dc} \) scaled by \( 2\pi\times10^5 \): favours conductive polarisable ground
IP vs self-potentialinduced by an applied currentnatural potential, no transmitter
IP signal vs EM couplingground polarisation, low-frequency or late-time effectinductive wire–ground coupling, grows with frequency, separation, conductivity
\( \tau \) vs \( c \) in Cole–Coleposition of the dispersion in frequencywidth (spread) of the dispersion
Pseudosection vs inverted sectionreadings plotted at nominal depth, pant’s-leg artefactsmodel of true chargeability

Common Mistakes


Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper.

Level 1: Foundation

  1. Name the polarisation mechanism caused by clay minerals.
  2. What is the unit of integral chargeability?
  3. Is resistivity higher at the lower or the higher frequency in an IP survey?
Answers
  1. Membrane (electrolytic) polarization.
  2. Milliseconds (mV·s per V).
  3. At the lower frequency: the ground has more time to polarise, so it presents a higher effective resistance.

Level 2: Part B Style

Practice Q1. A time-domain survey records \( V_p = 250 \) mV and, at a chosen delay after switch-off, \( V_s = 5 \) mV. The chargeability at that delay is:

(a) 2 mV/V (b) 5 mV/V (c) 20 mV/V (d) 50 mV/V

Answer: (c) 20 mV/V

\( M = V_s/V_p = 5/250 = 0.02 = 20 \) mV/V (2%). (a) is the percentage written as mV/V. (b) quotes \( V_s \) itself. (d) inverts and rescales.

Practice Q2. Resistivities of 120 \( \Omega\,\text{m} \) at 0.1 Hz and 112 \( \Omega\,\text{m} \) at 1 Hz are measured. The per cent frequency effect and metal factor are closest to:

(a) PFE 6.7, MF 349 (b) PFE 7.1, MF 374 (c) PFE 7.1, MF 401 (d) PFE 0.071, MF 3.7

Answer: (b)

\( FE = (120 - 112)/112 = 0.0714 \), PFE \( = 7.14 \). \( MF = 2\pi\times10^5\times0.0714/120 \approx 374 \). (a) divides by \( \rho_{dc} \) in FE (\( 8/120 \)). (c) divides by 112 in the metal factor. (d) gives FE as a fraction and drops the \( 10^5 \) scale.

Practice Q3. In the Cole–Cole model, a rock has DC resistivity 200 \( \Omega\,\text{m} \) and chargeability \( m = 0.1 \). Its resistivity at very high frequency is:

(a) 180 \( \Omega\,\text{m} \) (b) 200 \( \Omega\,\text{m} \) (c) 220 \( \Omega\,\text{m} \) (d) 222 \( \Omega\,\text{m} \)

Answer: (a) 180 Ω m

\( \rho_\infty = \rho_0(1 - m) = 200 \times 0.9 = 180\ \Omega\,\text{m} \). Resistivity falls with frequency. (c) and (d) have it rising (multiplying by 1.1 or dividing by 0.9).

Practice Q4. A dipole-dipole IP survey uses 25 m dipoles. The reading for \( n = 4 \) is plotted on the pseudosection at a pseudo-depth of:

(a) 25 m (b) 50 m (c) 62.5 m (d) 125 m

Answer: (c) 62.5 m

Dipole centres are \( (n + 1)a = 125 \) m apart; the 45° lines meet at half that depth, 62.5 m. (b) uses \( na/2 \). (d) uses the full centre separation.

Level 3: Part C Style

Practice Q5. Three anomalies are found in a frequency-domain IP survey (resistivities at the low and the high frequency):

Anomaly\( \rho_{dc} \) (\( \Omega\,\text{m} \))\( \rho_{ac} \) (\( \Omega\,\text{m} \))
P8076
Q87.6
R400388

Which statements are correct?

  1. P and Q have the same per cent frequency effect, about 5.3%.
  2. Q has a metal factor about ten times that of P.
  3. R has the lowest metal factor of the three.
  4. The metal factor alone proves Q is the best sulphide target.

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 2, 3 and 4 only (d) all four

Answer: (b) 1, 2 and 3 only

FE: P \( = 4/76 = 0.0526 \); Q \( = 0.4/7.6 = 0.0526 \); R \( = 12/388 = 0.0309 \).

MF \( = 2\pi\times10^5\,FE/\rho_{dc} \): P \( \approx 413 \); Q \( \approx 4{,}134 \); R \( \approx 49 \).

  1. True. 5.26% each.
  2. True. Same FE, one-tenth the resistivity.
  3. True.
  4. False. Q’s high MF comes entirely from its low resistivity. Its polarisability is the same as P’s. An 8 \( \Omega\,\text{m} \) zone with modest FE could be graphite, clay or saline water with some polarisable material; MF alone cannot rank them.

Practice Q6. A time-domain decay is modelled as \( V_s(t) = 10e^{-t/0.5} \) mV after switch-off (t in s), with \( V_p = 500 \) mV. A second receiver integrates from 0.45 s to 1.45 s instead of 0.1 s to 1.1 s. Which is correct?

(a) Both receivers report about 7.1 ms. (b) The first reports about 7.1 ms; the second reports less, about 3.5 ms. (c) The second reports more, because its window ends later. (d) Both report 10 ms, since \( \tau V_0/V_p = 10 \) ms.

Answer: (b)

First window: \( 10\times0.5(e^{-0.2} - e^{-2.2}) = 3.54 \) mV·s; \( M = 3.54/500 \) s \( = 7.1 \) ms.

Second window: \( 5(e^{-0.9} - e^{-2.9}) = 5(0.4066 - 0.0550) = 1.76 \) mV·s; \( M \approx 3.5 \) ms.

The same ground gives half the chargeability with a later window, because the decay is fastest early. (d) is the integral from 0 to infinity, which no receiver measures. The lesson: quote windows with integral chargeability.

Practice Q7. Which statements about IP are correct?

  1. Disseminated sulphides can give a strong IP anomaly with only a small resistivity anomaly.
  2. Saline groundwater gives low resistivity and strong IP.
  3. EM coupling increases with frequency, dipole separation and ground conductivity.
  4. For small chargeability, the full-range frequency effect is approximately equal to the Cole–Cole chargeability.

(a) 1, 3 and 4 only (b) 1 and 2 only (c) 2, 3 and 4 only (d) 1, 2 and 3 only

Answer: (a) 1, 3 and 4 only
  1. True. Unconnected grains polarise strongly but carry little current.
  2. False. Saline water lowers resistivity but polarises weakly; low IP is how it is told apart from clay.
  3. True.
  4. True. \( FE = m/(1 - m) \approx m \) for small \( m \).

Quick Revision Table

IdeaFact or formula
Electrode polarizationionic–electronic transfer at metallic grains; large; disseminated sulphides
Membrane polarizationion-selective clay-lined pores; smaller; moderate clay content
Chargeability\( M = V_s/V_p \) (mV/V) or \( (1/V_p)\int_{t_1}^{t_2}V_s\,dt \) (ms)
Seigelpolarisable medium acts as \( \rho/(1 - m) \); \( m_a = \sum(\partial\ln\rho_a/\partial\ln\rho_i)m_i \)
Frequency effect\( FE = (\rho_{dc} - \rho_{ac})/\rho_{ac} \); PFE \( = 100\,FE \)
Metal factor\( MF = 2\pi\times10^5\,FE/\rho_{dc} \)
Cole–Cole\( \rho_0[1 - m(1 - 1/(1 + (i\omega\tau)^c))] \); limits \( \rho_0 \), \( \rho_0(1 - m) \)
Domain link\( m = FE/(1 + FE) \approx FE \)
Phase peaknear \( f \approx 1/(2\pi\tau) \)
Pseudo-depth (dipole-dipole)\( (n + 1)a/2 \); pant’s-leg pattern over compact bodies
EM couplingrises with frequency, separation, dipole length, conductivity
False targetsgraphite, barren pyrite, clay

Key Takeaways

What’s Next