Downloaded 04/11/15 to 130.194.20.173. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.

org/

Fundamentals
Electromagnetic
of the
Method
Downloaded 04/11/15 to 130.194.20.173. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Fundamentals of the Electromagnetic Method

Charles M. Swift, Jr.

Explorationgeophysicsis applied to obtain information about the subsurfaceof the earth

that is not availablefrom surfacegeologicalobservations.Becausethe electricalresistivity
of different earth materialsrangesover many orders of magnitude, electromagnetic(EM)
methodsare used to map the subsurfaceresistivity structure.Applicationscan be at any
scale--a metal detectorsearchingfor coins in beach sands,a detailedresistivitysurveyto
map the influx of conductiveseawaterinto a resistiveaquifer, a high powered time-domain
electromagneticsoundingsurveyto map the thicknessof resistivevolcanicsover conductive
possiblypetroliferoussediments,or a magnetotelluricsurveyto detectanomalouslyconduc-
tive mantle for a dissertation. However, because most base metal massive sulfide ores are
very conductiveandprovidea strongcontrastto their hostrocks, mineral explorationsurveys
havebeenthe prime applicationof the EM methodsand the mineralexplorationindustryhas
fundedmost of the development.
Electromagnetic (EM) methodsincludean initially confusingvarietyof techniques,survey
methods,applications,and interpretationprocedures,which are further complicatedby a
bewilderingarray of trade names. Each technique,however, involves the measurementof
one or more electricor magneticfield componentsby an "EM receiver," from some natural
or artificial sourceof electromagneticenergy•the "EM transmitter." A useful classification
baseduponthis commonalityof EM systemsis presentedas Figure 1. It is apparentthat the
measurement is essentiallythat of obtainingthe transferfunctionof the earth. At mostprac-
tical currentdensities,the earthis linear, so the transferfunctionis simplythe outputdivided
by the input. Propertiesof the earth are interpretedfrom these transfer functionsin ways
similar to how propertiesof an electricalcircuit are interpretedfrom its transferor system
function.
Possiblythe mostfundamentalelectromagnetic methodis the directcurrent(dc) laboratory
measurement of the outputvoltagedevelopedfrom an input currentacrossa sampleof rock.
From Ohm's law, the ratio of the voltage in volts to the current in amperesis the absolute
resistivityin ohm-metersfor a rock one cubic meter in volume. The resistivitytechnique,as
categorizedin Figure 1, is a generalizationof the abovelab measurementfor arbitrary geo-
metric configurationsof the transmitterand receiverelectrodes,yet by being dc, it is only
a subsetof the electromagnetic methodswhich arise when nonzerofrequencies[alternating
current(ac)] are utilized by the transmitter.
Based on Maxwell's laws, an ac electromagneticsourceinducessecondarycurrentsin
conductiveearthmaterial. These secondarycurrentsgeneratesecondarymagneticfields. EM
receiversmeasureboth the primary and secondaryfields. Upon normalizationby the input
sourceor by the receivedprimary field, either the secondaryor the total (primary plus sec-
 6 Swift

ondary)field responseis interpretedto yield the significantresistivityinformation.

Electromagneticresponse,usually expressedas the secondaryfield or the mutual imped-
ance betweenthe transmitterand receiver, dependsupon the frequency, the conductivity
Downloaded 04/11/15 to 130.194.20.173. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

structure,and the geometriccouplingbetweenthe transmitterand receiver. (Quantitatively,

it will be developedin later chaptersthat the responsecan be plotted as a function of a
dimensionless
"response
parameter,"
{r•toa2, wherea represents
a dimension
usuallychar-
acteristicof the target but sometimesof the array.) Responseincreaseswith increasingre-
sponseparameterfrom zero, to a range of maximum rate of changein amplitudeand max-
imum responsein phase shift, to a saturationlimit of no change. This responsehas been
availablefor many yearsfor simpleearthgeometriessuchas a layeredearth, or for a sphere,
a cylinder,or a thin conductivesheettargetin an infinitely resistivehalf-space.For a complex
resistivitystructure,however, each conductiveregion will respondaccordingto its own re-
sponseparameter,and total inductionwill be complicatedbecausethe inducing fields for
someconductorswill include secondaryfields from the better conductors.Development of
the standardapproachto EM explorationfor conductivemassivesulfides 25 years ago in-
volved selectingan operatingfrequencyfor a given systemthat would producea measurable
responseonly from the massivesulfidebody. Prospectingfor thesetargetsconsistsof sys-
tematicallytraversingthe ground, and detectingan anomalousresponseover background.
The major advancesin EM methodsin the past 25 years have resultedfrom consideration
of the entireresistivitystructure.We now know that "currentchanneling"from weakly con-

Table 1. A classificationof electrical and electromagneticsystems

Receiver

Both wire Small coil Small coil

Transmitter Groundedwire and small coil (ground) (air)
Grounded wire

Galvanic Resistivity Magnetometric

IP resistivity (MMR)
Magnetic IP (MIP)
Inductive Eltran Controlled source Some TEM systems
AMT (CSAMT)
Small loop Slingram Airborne EM
Horizontal loop EM Time domain
Vertical loop EM towed-bird
Tilt angle method Helicopter
Some TEM systems rigid-boom
Coincidentloop
Borehole
configurations
Large loop Large loop systems
(long wire) Turam
Many TEM systems
Borehole
configurations
Plane wave

Vertical
antenna VLF-resistivity VLF VLF
Natural Telluric Magnetotellurics
geomagnetic currents
field

Notes:Groundedwiresmeasurepotentialdifferenceper length,thuselectricfield. Coils (or fluxgatemagnetometers

or SQUIDS) measuremagneticfield, or its time derivative.A smallloop is a 3-D source(magneticdipole).A long
wire (or the longedgeof a largeloop) is a 2-D source.NaturalEM sourcesare assumedto be 1-D sources.Receivers
canbe frequency-domain, time domain(TEM), or both. This classification
excludesthe high-frequencytechniques
(radar, etc.)
 Fundamentals of EM 7

ductinghostrock can funnel currentinto the good conductorsto producea responsealtered

from that of an identicalconductorin free space.We have developednumericalmethodsto
calculateforward solutionsand hence realistic type curves for interpretation.And we have
Downloaded 04/11/15 to 130.194.20.173. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

designeddigital instrumentationto better measurethe responseand better processthe more

complexdata.
Nevertheless,one "best"EM systemsimply doesnot exist for all applications.Depending
uponthe estimatedresistivitystructureincluding "geologicalnoise," the electrical noise en-
vironment,the availability of relevantinterpretationalaides, the familiarity of the geophys-
icist with a particular system, and nongeophysicalconstraintssuch as limited land accessi-
bility andbudget,any one of manysystemscan be the mostappropriate for a givengeophysical
problem.The diversityof the EM systemsnow in usecan be attributedto a numberof factors
as discussedin the following paragraphs:(1) different combinationsof type and orientation
of EM source and receiver, (2) the transmitter waveform and hence the receiver detection
schemes,(3) the techniquesusedto subtractout the primary field, and (4) the particularform
ofthefinalparameter
tointerpret.•

Differences in Array Configuration

The classificationin Figure 1 includesa numberof subtlebut significantdifferences.When

currentsin a conductiveearth are generatedwithout the transmitterdirectly contactingthe
earth,the EM responseis termedEM induction,andthe resultingtechniquesare "inductive."
For currenttransmitteddirectly into the earth in dc resistivity, the techniqueis "galvanic."
Galvanictechniquesdependupongoodelectrodecontactand thusare not appropriatein areas
of surfacehigh resistivity such as dry sand or glaciers.
EM responsecan be inducedby naturalearth currents(passivetechniques)or by artificial
or controlledsources(active techniques).Most EM systemsemploy an active transmitterso
that the sourcegeometryand frequencycan be controlledby the geophysicist.However, the
ambientelectromagneticradiation from ionosphericoscillations(below 10 Hertz) and from
lightningdischarges(above8 Hertz) is utilized as a sourcefor the telluric, magnetotelluric,
and audio-frequency magnetic(AFMAG) techniques.Advantagesare the avoidanceof the
financialand logisticalproblemsof a transmitter,and the availability of low frequencyen-
ergy, which is expensiveto generateartificially. However, sufficientsignalstrengthis not
alwaysavailable,particularlyin the 0.1-10 Hz frequencyrange.
In addition,the naturalfield magnetotelluricand AFMAG fields are "plane wave" sources
for mostall applications,which simplifiesthe analysisbecausethe locationof the transmitter
can be ignored. In VLF (for "very low frequency"in radio jargon, although20 kHz is a
high frequencyfor explorationgeophysics)and controlledsourceaudiomagnetotellurics,an
active sourceis usedto simulatea "far field" plane-wavesource.This assumptionusually
holds for VLF. In CSAMT, as the receiver-transmitterseparationbecomesless than three
skindepths(8 = X/2/trlxto)at the low frequencies,
thefield becomes
"near-field"
rather
than "far-field," and the transmitteractsmore as a local EM sourcethan as a magnetotelluric
(plane-wave) source.
Array configurations
are selectedin advancedependinguponthe application,whetherfor
"sounding,"
to measurethe variationof resistivitywith depth,and/or for "profiling,"to
measurethe variationof resistivityacrossthe surface.Whereasin galvanictechniquesdepth
penetration for a given resistivitystructureis controlledonly by the array geometry,in in-
ductivetechniques deeperpenetrationcan alsobe obtainedby usinglower frequencies.How-
ever, lower frequenciesin practicerequirehigherpower. In addition,the EM skin depth
physicsthat allowsdeeperpenetrationwill also diminishresolution.In the earth, electro-
magneticwave propagationat geophysical frequenciesis controlledby a diffusionequation,
not a wave equation.EM wavesattenuatesignificantlyand dispersively,so resolutionanal-
ogousto seismicsis impossible.
In profiling, choiceof a receiver-transmitter combinationis usuallydeterminedby maxi-
mizingthe responsefrom the target. Bestresponseoccurswhen the primary magneticfield
 8 Swift

is perpendicular to the planeof the target, and whenthe long axis of the targetis parallel to
the primaryelectricfield. Plane wave sourceshave greathorizontalextent, and thus couple
well with very long targets such as clay-rich shear zones. For steeply dipping targets, a
Downloaded 04/11/15 to 130.194.20.173. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

horizontalloop sourcelocated on the ground away from the target generatesa horizontal
magneticfield in the earthfor maximumcoupling.A commonconfigurationfor profiling is
the so-calledSlingramarray, a systemusing small coplanarloopsfor transmitterand receiver
at a fixed separation.Another configurationfor profiling is Turam, a systemusing a large
loop transmitterand two receivercoils at a fixed separationmeasuringamplituderatios and
phaseshiftsof the two receivedsignals.The popularmodernversionusesone receivercoil
measuringthe responsewith referenceto the transmittedwaveform. For sucha large loop
source,conductivegroundand target conductorscloseto the loop can rotate and attenuate
the field so that coupling away from the loop can be reducedfrom that anticipatedin the
absence of other conductors.

Differences in Transmitter Waveform

Early EM systemsusing analog electronicsutilized a sine wave sourcewaveform, or a

square-waveof equivalentfundamentalfrequency,with different frequencies,if employed,
beingmeasuredas separatemeasurements.Square-waveformsare more commonin the newer
digital instrumentation, much of which utilizes synchronous detection.Most of electromag-
netictheoryfor geophysicalapplicationswas developedin the frequencydomain(seeWard,
1967, Grant and West, 1965). Measurementof some electromagneticfield componentis
expressed as the amplitudeandphaseshift (relativeto the primary), or as the in-phase(real)
and out-of-phase(quadrature)components.One importantairborneEM system(INPUT) and
somegroundsystems(see Keller and Frischknecht,1965), however, measuredthe transient
decayof the secondaryfield, after turn-off of the primary field.
Much of the developmentof the EM methodsin the past twenty years has resultedfrom
the recognitionthat this "time domain"measurement,measuredin the absenceof the primary
field, offers somepractical advantagesto a "frequencydomain" measurement.Time and
frequencydomain alternativeshave existedfor years for the inducedpolarizationmeasure-
ment in the 0.1-10 Hz frequencyrange. Theoretically, one transientwaveform (or a "stack"
of the sum of many measuredwaveforms) possessesthe same information content as the
frequencyresponseobtained in a more time consumingfashion over many separatemea-
surements.The time domainmeasurementis wide band, however, and thus is more suscep-
tible to noise(mostlytellurics)which can be filtered out in a frequencydomain measurement.
Currently, the better approachfor IP is debatable,and the final survey choice frequently
dependsupon the systemwith which the geophysicistis most familiar.
For EM, the better approachappearsto be time domain EM [transientelectromagnetic
method(TEM)], particularlyin areasof high surfaceconductivitysuchas Australia. A con-
ductive environmentforces an EM systemto operate at lower frequenciesor longer times,
where the signal-to-noiseratio deteriorates.In fact, modeling has shown that the optimum
time for detectionof a finite targetmay be over a finite time window when the targetresponse
is greaterthan the backgroundresponseof conductivehost and overburden. Nevertheless,
the direct measurementof the secondaryfield in TEM avoids the contaminationof noise
causedby errorsin removing the primary field in FEM methods.

Differences in Techniques to Remove the Primary Field

An accurateestimateof the secondaryfield requiresan accuratetechniqueto remove the

primaryfield from the total observedfield. Thesetechniquesdiffer accordingto the type of
measurement, andthe type of final interpretiveparameter.For VLF and simple"dip-angle"
configurationsof the horizontal-loopEM methods,the primary field is not removed; a sec-
 Fundamentals of EM 9

ondaryfield tilts the total primary field (from the backgroundorientationof horizontal or
vertical, respectively).Similarly for magnetotellurics,where the measuredelectric field can
be thoughtof as indicatingthe inducedEM responsefrom the measuredinducingmagnetic
Downloaded 04/11/15 to 130.194.20.173. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

field, no secondaryfield is normally considered.

For most active EM systemsfrom simple sources,the primary field is calculatedusing
carefulgeometricinformationfrom surveyingin the field, thenthe secondaryfield is obtained
from subtractingthis theoreticalestimatefrom the observedtotal field. Resultsare then ex-
pressedas a percentof the primary field. Deviationsfrom zero thusreflect either anomalous
secondaryfields, or erroneousremoval. For real primary fields, measurementof the quad-
raturecomponent(or phaseof the total field with reflect to the primary) revealsdirectly the
presenceof a secondaryfield. Simple quadraturesystemscan thereforeavoid the necessity
of carefulprimary field removal. Thus, someearly airbornesystemswere quadraturesystems
to avoid the complicatedcompensationsystemsused to monitor variations in transmitter-
receiverseparationdue to wind and vibration.Anothertechniqueto avoid careful surveying
utilizesthe low-frequencyasymptoticresponse.Secondaryresponseis thus obtainedby sub-
tractinga low-frequencymeasurementfrom a higher-frequencymeasurementwhich includes
responsefrom someconductorwhoseresponseparameteris sufficientlyhigh.
As mentioned,one very effective techniquefor avoiding this problem is to measurethe
transientresponsein the absenceof the primary. All the time domain systems,includingthe
airborneINPUT system, utilize this technique. Normalization by the transmitter moment
resultsin unitsof volts/(ampstimesmeterssquared).

Differences in Interpretive Parameter

The variousEM systemsyield final parametersthat vary in their ease of understanding.

Someyield geometricdescriptionsof the total field suchas the tilt angle or ellipticity. Some
yield a measureof the outputtotal field normalizedby the input source,the impedance.The
magnetotelluricimpedancecan be easily transformedinto the conceptuallymore convenient
apparentresistivity.The mutual impedancebetweentwo loopscan be normalizedand trans-
formed with a little more difficulty into an apparentresistivity, although some physicists
prefer to considerthe observedfields per se. Most EM systemsyield a measureof the sec-
ondaryfield normalizedby the local primary field to produceunits of "percentof primary,"
or "partsper million." The diversity in data presentationcan be confusingto those used to
the standardpresentations of gravity, magnetics,and refractionand reflection seismics.
However the EM responseis presented,the geophysicalinterpretationinvolves two inter-
relatedquantitativesteps.First, the geometryof any anomalousconductoris interpretedfrom
the spatialcharacteristicsof the coverage--length from the anomalousextent, depth from
the horizontalshape,dip from the asymmetryof response,etc. Second,the resistivity as a
functionof depthor the conductivity-thickness product(the conductance)of a confined con-
ductoris interpretedfrom the frequencyor time variationof the response.Awarenessof the
resultsof EM modeling, whether by scale models or analytic formulas or numerical com-
putations,is critical. Forward modelingprovidesnot only type curvesfor interpretiveuse,
but also insightfor increasedunderstandingof the physicsof EM induction and information
for the intelligentdesignof field surveys.Automatic interpretationof the data to producean
appropriateand consistentresistivitystructure--the inverseproblem•is availablefor some
simplegeometries,and is an active researchtopic for the more complex earth geometries.
Nevertheless,for the limited irregularly spacedmeasurementsover a geologicallyheterog-
enousearthin the presenceof instrumentand ambientnoise, the ultimate act of geophysical
interpretation remainsan artful balanceof physicalunderstanding,awarenessof the geolog-
ical constraints,and pure experience.
 10 Swift

The Future
Downloaded 04/11/15 to 130.194.20.173. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Electromagnetic methodsare necessaryand acceptedin mineralexploration,in geothermal

exploration,andin engineeringgeophysics. In petroleumexploration,electricaland electro-
magneticmethodshavebeenlumpedwith gravity andmagneticsinto a nonseismiccategory,
but recentsuccessfulapplicationsin mappingpermafrostand in determiningthe thicknessof
high resistivitybasaltand overthrustcarbonatesectionsin areasof poor seismicdata have
broughtthe EM methodsinto more seriousconsideration.(Of course,inductionloggingtools
have been used with confidencefor years by the petroleumindustry, but are not usually
mentionedwith the surfaceexplorationmethods.)Further acceptancein petroleumexplora-
tion by explorationists
not experiencedin EM methodsrequiresthat any EM surveyor system
be explainedfully and quantitatively,with an understanding of the relative capabilitiesand
limitationsof the EM technique,particularlythe decreasingresolutionwith increasingdepth
of penetration.
For the near-termfuture, the emphasisin the applicationof electromagneticmethodswill
be on precision.Engineeringapplicationsin particularrequireaccuracy.Moreover, the con-
cern over cost-effectivenessnow permeatingall mineral and petroleumexplorationprograms
will forcegeophysi,cists
to design
theirEM surveys
morecarefully
andto interpret
theirEM
responses
moreprecisely.The contentsof this book providea solidfoundationfor this work.

References

Grant, F.S., and West, G.F., 1965, Interpretationtheory in applied geophysics:McGraw-Hill Book
Co.
Keller, G.V., and Frischknecht,F.C., 1966, Electrical methodsin geophysicalprospecting:Pergamon
Press Inc.
Ward, S.H., 1967, Electromagnetictheory for geophysicalapplications:Mining geophysics,Vol. II.,
Societyof ExplorationGeophysicists.
Downloaded 04/11/15 to 130.194.20.173. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/