GEOPHYSICS, VOL. 54. NO.5 (MA Y 1989); P. 598--608. 12 FIGS.

Effect of conductive host rock on borehole transient electromagnetic responses

Gregory· A. Newman" l W·a-Iter L. Anderson] andI

Gerald W. Hohmann§

ABSTRACT

Transient electromagnetic (TFM) borehole responses response at later times if the body is well coupled to the of 3-D vertical and horizontal tabular bodies in a half­ transmitter. If the host is very resistive, the galvanic space are calculated to assess the effect of a conductive response vanishes; and the response of the conductor is host. The transmitter is a large loop at the surface of the caused only by vortex currents. earth, and the receiver measures the time derivative of The shapes of the borehole profiles change consider­the vertical magnetic field. When the host is conductive ably with changes in the host resistivity because vortex (Ioo·n·· m), the borehole response hi"due rrrairrly ur cur­ and galvanic current distributions arc- very different

rent channeled through the body. The observed When only the vortex response is observed, it is easy to magnetic-field response can be visualized as due to gal­ distinguish vertical and horizontal conductors. How­

vanic currents that pass through the conductor and ever, in a conductive host where the galvanic response is

return in the half-space. When the host resistivity is dominant, it is difficult to interpret the geometry of the

increased, the magnetic field of the conductor is influ­ body; only the approximate location of the body can be

enced more by vortex currents that flow in closed loops determined easily. For a horizontal conductor and a inside the conductor. for a moderately resistive host single transmitting loop, only the galvanic response en­

(1000 Q. m), the magnetic field of the body is caused by ables one to determine whether the conductor is be­

both vortex and galvanic currents. The galvanic re­ tween the transmitter and the borehole or beyond the

sponse is observed at early times. followed by the vortex borehole. A field example shows behavior similar to that of our theoretical results.

INTRODUCfION sulfides (Boyd and Wiles, 1984; Dyck and West, 1984; Eadie and Staltari, 1987).

The borehole transient electromagnetic (TEM) prospecting Because the borehole prospecting method was first applied method uses a large transmitting loop laid out on the surface in resistive terrain. interpretation was successfully based on and a small receiver coil lowered down a drill hole. A series of free-space models. Two free-space models that are commonly current pulses is transmitted. and a sensor, coaxial with the used arc plates and spheres. Plate models can represent con­drill hole. measures the time rate of change in the magnetic ductive veins, shears, or dipping strata, while sphere models field along the borehole during the transmitter off-time. Bore­ can represent bulk regions of sulfide mineralization. Insight hole surveys increase the prospecting depth for base metals; derived from plate and sphere models has greatly improved case histories in Australia and Canada have shown the surveys our understanding of TEM responses in resistive terrain to he very useful for exploration and development of massive (Woods. 1975; Dyck and West, 1984). However, many mining

Presented at the 57th Annual International Meeting, Society of Exploration Geophysicists. Manuscript received by the Editor March 28. 1988; revised manuscript received October 24, 1988. "Formerly Department of Genlogy and Geophysics, University of Utah; presently Institute for Geophysics and Meteorology, University of Koeln, 5000 Koeln 41, West Germany. :tV.S. Geological Survey. Box 25046, MS 964. Denver Federal Center, Denver, CO 80215. ~Departmenl of Geology and Geophysics. University of Utah, Salt Lake City. UT 841] 2-1 (83. c 1989 Society of Exploration Geophysicists. All rights reserved.

598

599 Borehole Transient EM Responses

targets are embedded in a conductive host or beneath conduc­tive overburden, in which cases interpretation of data based on free-space models may lead to erroneous results. Simply removing the host or background response from the data prior to a free-space interpretation, even if such removal is possible. does not completely eliminate the effect of a conduc­tive host. For example, the response of the target may not decay exponentially at late times (San Filipo et a1., 1985; Newman and Hohmann, 1988), as expected for a confined conductor in free-space.

Until now, the effect of a conductive host on borehole TEM responses has not been well understood. Eaton and Hohmann (1984) discussed the influence of a conductive host, but only for two-dimensional (2-D) structures. In Eaton and Hoh­mann's investigation, the transmitting loop was replaced with two infinite line sources parallel to the strike of the structure. Hence, their results are valid only for the 2-D inductive re­sponse. Typically, base-metal deposits are three-dimensional (3-D) in character; 2-D model studies cannot reflect the closed nature of eddy currents in the typical 3-D base-metal deposit, nor can they simulate galvanic current. If the host is moder­ately conductive, galvanic current due to charges on resistivity boundaries can make an important contribution to the re­sponse of a 3-D target. Galvanic current can be visualized as host current that is channeled through or deflected away from the target depending upon whether the target is more or less conductive than its host. McNeill et al. (1984) and West and Edwards (1985) used approximate techniques to study the relation between vortex and galvanic TEM responses for sur­face configurations.

More recently, West and Ward (1988) modeled the borehole TEM response of a 3-0, horizontal, tabular body in the earth with a view toward detection of a fracture zone. In another recent paper, Augustin et al. (1989) evaluated the effect of metal casing on surface-to-borehole, frequency-domain EM. However, their work relates primarily to borehole logging, i.e., determining the electrical resistivity in the vicinity of a drill hole. In this paper we use numerical modeling to assess the influence of a conductive host on the borehole responses of vertical and horizontal 3-D bodies that represent typical tar­gets in mineral exploration. We first discuss the general characteristics of vortex and galvanic responses of a 3-D con­ductor in a conductive host in order to provide a basic under­standing of borehole responses. We then compare borehole responses of bodies in free-space and in a conductive half­space, with a view toward interpreting the geometry and lo­cation of a conductor. We conclude the paper with a field example that exhibits similar characteristics to those of our model simulations.

MODELS AND METHOD OF COMPlJTAnON

Figures la and l b show the 3-D models considered. The bodies have a resistivity of 1 o· m and are centered at 305 m depth. In cross-section, the vertical and horizontal bodies are both 50 m by 200 m. Their strike length is 1000 m. We carried out model simulations for earth resistivities varying from 1O:l n· m (essentially free space) to 100 Q. m. A rectangular loop, 500 m in width and 1000 m in length, is used for the transmit­ter; the long axis of the loop is parallel to the strike direction. The near side of the loop is located 300 m from the origin,

VE RTICAL BODY

205m

I p=l n'm ,0 010 2,IO~ 10$ n. m

~LUa. Lengtlt -/DOOm

-140m 740m

5800 DoPj j HORIZONTAL BODY

//77l7T//7/7J7Y///l//7/! I .[ , .. , . , , , , , , , Boreholes

280m 2.10'.70'./2'/11

,0-70 I I I Length

p= 7Jl./II 1000m I i I 4

I 1150m

T200'~ I

580. Dop,J r r FIG. 1. (a) Cross-section of the vertical-conductor model used for computing TEM responses in the two boreholes shown. Borehole locations measured from center of conductor. (b) Cross-section of the horizontal-conductor model used for computing TEM responses in the three boreholes shown. Borehole locations measured from center of conductor.

which is over the center of each body. Results are computed for the vertical body in two boreholes at -140 and 140 m from the center of the body and for the horizontal body in three boreholes at -140,80, and 140 m from the center ofthe body.

Because our investigation requires borehole simulations over a wide range of background resistivities, we used the 3-D TEM numerical solution of Newman and Hohmann (1988). This solution is based on the method of integral equations and provides the TEM scattering responses of prisms in layered half-spaces. Because the solution includes a specialized set of divergence-free basis functions, it can model the response of one or more prisms in highly resistive layers and in free space.

VORTEX AND GALVANIC CURRENTS

It is necessary to distinguish between galvanic and vortex currents in order to understand 3-D borehole responses in a conductive host. Let us first consider the vortex current. When the vertical conductor (Figure la) is in free-space, its magnetic-field response is caused by vortex currents flowing in dosed loops. These eddy currents are fanned because at the

600 Newman et al.

Vortex Currents

- --~---"\ Mo~n.t/( fle/ei

\~ ~ ~

I

~ I I

/

'/)/ ,/

, " ....<, '\ \ \ , ~ ~ J

.... _/_ -,,/ /1 J 1 ....... _---" / / ....... ""'-'- --."," ,,"

------'

/'" ....... ~::-:.-- .... I /"., - - - .../

I 1 / I'" v -, I I \1 , \ ~ , 1 \ " -\ -, "-

"

Galvanic Currents

\ Mo~n./I( fl.lei

\~ t ' I f

I Ii I "I J

--';,//11 --" '" /_/,.../

FIG, 2. (a) Sketch of vortex currents in a thin plate with the corresponding magnetic field shown by the dashed lines. The directions of the vortex current and the magnetic field are determined by the direction of the current in the source loop. (b) Sketch of galvanic currents (solid lines) and the magnetic field (dashed lines) of a thin plate. The directions of the gal­vanic current and the magnetic field are determined by the direction of the current in the source loop.

time the transmitter is shut off, currents begin to flow in the conductor in order to preserve the magnetic field. If the body is thin, the eddy currents (Figure 2a) flow along its depth and strike dimensions. The eddy currents decay with time, as does their magnetic field.

If the body is embedded in conductive ground, the col­lapsing magnetic field of the transmitter induces electric cur­rents in the ground. These host currents interact with the body, so that the TEM response of the body can be very different from its free-space response. An important type of interaction between the host currents and the body is galvanic current flow. Galvanic current is due to charge that appears on the surface of the body because the normal component of electric field is discontinuous there. Because of this charge, current is gathered from the host and channeled into the body

if the body is more conductive than its host. However, if the body is more resistive, then the host current is deflected away. Magnetic fields produced by galvanic and vortex. currents are very-different (compare Figure 2a-with- Figur-e 2h)._ In. general; the TEM response of a J-D body in a conductive host is due to a complex interaction between vortex and galvanic cur­rents. Either type of current can dominate the response at different times after transmitter shut-off.

Color plots of electric-field patterns in the earth provide a clearer understanding of vortex. and galvanic currents. For comparison, in Figure 3, we show the electric field of the 1000 n'm half-space without the body. The location of the conduc­tor is shown for reference, but its effect is not included in these plots,

These color contour plots are snapshots of the strike com­ponent of the electic field in cross-section; each individual color plot corresponds to a different time after the transmitter is turned off These electric-field plots can be converted to current-density plots by multiplying by the conductivity of the half-space. At early times (0.03 and 0.1 ms) the currents in the half-space are concentrated around the transmitter; at later times (greater than 0.1 ms) the currents diffuse outward and downward away from the transmitter in the manner described by Lewis and Lee (1978) and Nabighian (1979).

When the half-space electric field interacts with the body of Figure 1a, the electric-field plots of Figure 3 change consider­ably. In Figure 4, from 0.03 to 0.1 ms, the effect of the conduc­tor is to oppose the buildup in the field by retarding the outward and downward diffusion of the field maximum in the vicinity of the conductor. However, at 0.3 ms, a vortex begins to develop centered on the conductor; from 1 to 5 ms, the vortex is observable even in the presence of the half-space field. At times later than 5 ms, the vortex vanishes; and the field is depressed compared to the field in the homogeneous half-space shown in Figure 3.

To observe the galvanic part of the electric field and conse­quently its current, we must subtract the half-space electric field of Figure 3 from that or Figure 4, The resultant field in Figure 5 illustrates the time ranges when the galvanic field is dominant. At early times, from 0.03 to 0.1 ms, the electric field exhibits a bull's eye over the conductor; but this field is op­posite in sign to the half-space field of Figure 3. The bull's eye pattern in Figure 5 is due to charges induced on the surface of the body. These charges reduce the electric field inside the body in Figure 4, but the current density actually is large compared to that of the surrounding earth, because the con­ductivity of the body is a thousand times larger than that of the half-space. Figure 4 shows that when the electric-field maximum of the half-space is moving toward the body, the response is dominantly galvanic.

At later times, a vortex develops in Figure 5. The vortex is centered on the conductor at 0.3 ms and is most strongly developed from I to 5 ms. This vortex current creates a mag­netic field that opposes the collapse of the primary magnetic field in the half-space. However, at late times, the vortex van­ishes and the galvanic component of the electric field again dominates,

After 10 ms, the electric field inside the body (Figure 4) is smaller than that of the surrounding earth and has a decay rate of t - ~/2, which is that of a half-space. The magnetic field of the conductor also decays at this rate. Kaufman and Keller

T , 03 ms CI ZSE -3 (Vim 1 ..e- 2 T :: 31M CI :: .23E - 7 IVim ! .. £ - 6 . ; ~

~,.. . ·i . .. . ..,·.~-f • • • • ·"..... ·

.. , ... I • • • • .. . :'H

·.­ • , I .: ;::f ' 1. ... · .....: :1::" "'UI

. ­ • 1 - •............­ .. . · . ­

f •·I •.·.. ..­..,. • ..w ·., ...,.....: n;.., ,.­ : ~

. ~:

1 ms ci • .36£ - 4 IV/ m i ..E - ) T " 5 InS CI " 64E-8 {Vim t ..£ - 7 : ::: ··............

I e . ·, .:..­... • ' ,I •·.. ..: :::

• 11_..­.. . t •.­* .... ·: ~ .(..­;ID.­·i .-;$ J .... .

· fi U _ · .".• I ~ '·.­ ::5 i::.!E ' II' . ... i' 001;;

T 3m!. CI:: loS E - 5 I Vim J .E-4 T:llms CI a . n E ~ 8 {Vim ) toE - 7 i • ' ·;" .

0" .i :i ~.-::;:·­,..: Ii: : E!!..­I• ::I I . • r ..: :::: ·.­ : ~ . :

··: :::,­' I , ' 111

·.~­.~.

..• , I' , ~ ... . ·. ·..

.. ."' .

f .

" " "

f m C :If••'8

!='

ms CI 3L E- 5 IV/m ! ..E - 5 T: :J) ms CI " 68 E ~l) [ V/m j .f - 9

: :~ : · .... • ·1.': .'::: · " ..., ...: ::::f • • 1,

.. . .t .·· .....­-. · "·..,-... .. ....: :l:: : ..,::·, ...... " .. • ,' , I.. ....:::;:

FIG. 3. Contour plots of the primary electr ic field (V1m ) in a 1000 n ·m homogeneous half-space. The cross-section is J500 m in the x direction and 580 m in the z direction. Measurement times range from 0.03 to 30 ms, and the contour interval and levels are given for each plot. For reference, the location of the conductor of Figures 4 and 5 is indicated by the rectangle; the position of the source loop is denoted by the + and I symbols. At early times, the currents in the half-space are concentrated around the transmitter; as time progresses, the current maxima move outward and downward. g...

8 N

T 03 ms Cl e 21. E - 3 I Vim ' .E· .2 T,. 3 ~ cr =82E-7 { V/ m l -£-6 .. .....­., . ...· -. _A

.~

...., .... ·.,:.

. .~

~ , '.4 ' 1 """·· ..-....

ft . ; . ·,* .... .... ... ' ~..­ .': .

T 1 m!> CI • 26 E • 4 VIm .E - 3 T:5 ~ CI : 16E -1 I Vim I .E - 6

·.. ._.­ .. ..... ·.. . .. '.•; ~iii • _• "" M

. ~- ..,.....'" -.. , ~ ... .." .......• .. " . ' ,.

it . ...

".. ,­ - • MM0 ...... .. ..--~

l! • • •. •.... .I ~" "" ...... - .., .,_-,".. "... ­!ll' 1 . ·.. . ."~ .. •

Z ... ' .. •: t:':·.,--­

:I•::t T .3 ms CI : ,3'1 E• 5 I VIm I -E- 4 TzlOms CI z 11 E • 8 ( Vi m I -E -7 !t

" .... !!­f , \ ' . I . • • • ~ I I I . · ­

·f f· ... · f , . _ . · ·II . 1'. ~ ,.­ , ....·f .·"......~, ..­f • • • • .. • , 1.4l .-....'" ,1 ft··.,,.oil II,. ... "

...:­III • • ' ": :::: .-I

~

.iO .. I ..• ...• ­::: ~

T . 1 ms CI : 47E- 5 ! '11m l .E·S T .. Xlms CI = liSE - () t VIm I .E- 8 •........• _. 'oJ, ....

f ' { f .. .'" ",;,;',alI,;/.II".:t ............r ._­

III •• ••"'y ·.....r: , , ~ ", " .,..~~ <~~~ .. ," " .

r:-'" ». ",':"",,' r, .,.. , ", _ I I • •a i. ....", ' .• : 1\:: , ~ p, ' . .. ' ''''1[ : ~~

_~~~

.,.­.... , -,'

....

. ~ .

II ," . .. e ,o' , ' 0. ' - '......_.

• . c"

FIG. 4. Contour plots of the total electric field in a 1000 Q . m half-space containing a 1 n ·m vertical conductor. Compare with Figure 3. Galvanic currents flow in the body at early and late times; vortex currents. at intermediate times .

T om ms CI • .fiDE - L ( VI m I

T = 1 ms CI .eE -L {V/m !

T : .3 ms CI • 21 E- 5 I VIm I

T ; 1 ms CI =.t.7 E -6 [ VI m)

.e - 4..­,,-­J:.::·­-­:=· ~ .... '­~ = .....­ ­...... ..... i i ::::

.e -L ..: L....,­= .­,., ,-­

.;\ :c(.....­·j ...j '­,=r,_.1_ ­

' ­

-£-4 .. ..... : ~ ..­• 'I".:·i .:-\: t.: : l~··• ',-."I'...''

' f ··. :

..u .... ,a

I N/ a ..• .t'".... • ' 1"­

.e -s ··........ .._­..::::­.­...... ..;

.4'1

' , ­..!i.'.... ....~• 1 • •

··...l. .

T,,3ms Cl " .83E - 7 (VI M )

T . 5IM CI : .I6E - 7 [ Vi m J

T.llms CI " .55E- 9 [ Vi m J

T.301M Cl • 32E - 10 : V/ m )

. E - 6 ·..... ­.. ...... ..~...1'·..-"

" " : :::: · . 0­• I f ... ·....." 0 • ··. ­, .

.~

., ~

. E - 6 · ....• ­

.'00-.1 ·,'-

"

-e-B .'.

~ .... .. .. "

· ' . 'I "

- E- 9 ... . .... , " ;,.

" ..... ', I •

lD

! ~

•2.

I ....

l! m ~

j:D

I•

F IG, 5. Contour plots of the scattered electr ic field due to the conductor only. These plots were obtained by subtracting the electric field of Figure 3 from that of Figure 4. Galvanic currents are apparent at early and late times ; vortex currents, at intermediate times. II)

e

604 Newman et al.

(1985) showed a similar result for a sphere in a conductive half-space. The decay rate of t - 5:2 shows that the late-time response is driven by the half-space electric field.

BOREHOLE PROFILES

In our model simulations, the borehole profiles are based on a voltage measurement after the current is terminated in the transmitter. Because the boreholes are orientated vertically in Figure 1. this voltage measurement is related to the vertical component of the magnetic field by -Ilo Dhz !()t Aa, where ~o

is the magnetic permeability of free-space and Aa is the area­turns product of the receiving coil. All borehole responses are normalized by Aa and are based on a step shutoff of the transmitter.

Vertical conductor

We begin with a study of the responses of the vertical con­ductor in free-space, for direct comparison with the response of the same conductor in a conductive half-space. The general characteristics of such responses are well known from previous scale and numerical model studies (Woods, 1975; Dyck and West, 1984). Figure 6 shows borehole profiles at seven measurement times ranging from 0.5 to 25 ms for the vertical conductor of Figure 1 in a 105 Q. m host, which simulates free-space. The geometry of the eddy currents, as illustrated in Figure 2a, causes the magnetic field, and therefore the voltage, to change sign as the coil moves past the center of the body. The shape of a profile for the magnetic field (not shown) is similar to that of its time derivative, except that the sign is reversed. Figure 2a also shows why the signs of the borehole profiles on the left side of the body in Figure 6 are opposite those of the profiles on the right. This sign difference is impor­tant because it can be used to indicate the direction to the target.

The profiles in Figure 6 show the expected exponential

0)·140. b)14011

, \ I \~\ v I \ \ \ \ II

\ \ \ \ \ \.~ .. 100 I r~ 9 5, ~ 'llJ\\ \\~..

I

z~ I \ I I J/ l I I , I II

200 '/ ',1//~ ""lJJ ) .u. 'ii. 300 -, " " \ \ \~\& \ I \ S j II i: ')) -~ -~~)

400 2~ I t I J /.5 I 15 I I I " I' I I I II

Ii I! / '! II 500 1/ I / I# 600

700

tOI 102 30°'0'6 10'3 ,0" 10' 1 '0 .,1 10" 100 -d 10'

(nW.. )

F1G. 6. Borehole profiles (voltage induced in a vertical coil) for the 1 n' m vertical conductor of Figure 1a in a 105 n· m host that simulates free-space: (a) borehole between transmitting loop and conductor, (b) borehole beyond conductor. Dashed lines indicate negative voltages; solid lines, positive voltages. The geometry of the eddy currents causes the sign change as the coil moves past the center of the conductor.

0)·140 .. b)140 ..

5

<, :::-;::-',

0'1 .... -, '\ '\ \,~:: : \\ \\" 'OO~ , \ 's•• \ \ \ \ I I \ \

~ 200[-25~ 200 I I J 5 j 11 /} / \159 ~

Q. 30:>t15 15 it } J Q. 300 Cl~ II ~/~ 5M a ''1~/)400

400 I'" (I 500 1i /q/"'f-! ,</

I

600 600

700 700

800'0'3 10.2 10') 100 10' 102 'oJ 800 , 0_3 10- 2 10- 1 100 10' 102

(nV/",2) (nV/.,2)

FIG. 7. Borehole profiles for the vertical conductor of Figure Ia in a 1000 Q. m half-space: (a) borehole between transmit­ting loop and conductor, (b) borehole beyond conductor. Note the galvanic response at early times and the vortex response at intermediate times; the half-space response dominates at late times.

r I b)14011','40. - I ,'"

1-- \, ,', s''-, '00 --- 9'",,",,\," )Of ~Y./'/1\

: ,

\- _

\'(100 '_ zoo '.~)

e \ 1/1 ~ tOOt !.l \ \") J'OO __:'1) L, '~j;.,1.Sln

] 300 000 • /I I / I~~/o' <,'00 ."00 " ,

I I 5 3

::tl-­ 'I:: 700 I 3 800 , > 10"800 • 210 10 2 10 10

FIG, 8. Borehole profiles for the vertical conductor of Figure la in a too n· m half-space: (a) borehole between transmitting loop and conductor, (b) borehole beyond conductor. Due to the low host resistivity, only the galvanic response is observed.

decay, with an estimated time constant of 1.28 ms. This time constant agrees with the empirically determined time constant of a thin plate in free-space given by Lamontagne (1975) as

l' = Ilo at L!lO,

where L is the intermediate dimension of the plate and at is its conductivity-thickness product.

Figure 7 shows borehole profiles for the vertical conductor in an earth of resistivity t000 n· m. In the first two time channels, 0.5 and I ms, a significant component of host cur­rent is channeled through the body along strike. The behavior of the magnetic field produced by this current is illustrated in Figure 2b. To the left of the body, over the entire length of the borehole, there is a peak in the profiles, as shown in Figure 7a. Here the response of the conductor adds to that of the half­space. However, on the right side of the body, as shown in

605 Borehole Transient EM Responses

Figure 7b, the response is opposite in sign and greater than the response of the half-space over most of the profile. Hence, the profile is opposite in sign to the profile on the left, but it still peaks near the center of the conductor.

At intermediate times (3 and 5 ms), vortex currents domi­nate the response. Profiles at these times show sign reversals near the center of the body. These sign reversals and profile shapes are similar to those observed in free-space (Figure 6).

At later times, the response of the body is galvanic, but the half-space response dominates.

Figure 8 shows borehole profiles for the vertical conductor when the host resistivity is decreased to 100 Q. m. Only the galvanic response is observed. The formation of eddy currents is not indicated in the profiles at any time, and the profiles bear no resemblance to the corresponding ones in free-space shown in Figure 6. In Figure 8, the profiles show a peak over the entire depth extent of the conductor. As expected, the response on the right is opposite in sign to the response on the Jeft. At both early and late times the profiles are dominated by the response of the hair-space.

Information on the location of the conductor with respect to the borehole and the transmitter also is available when the response is galvanic. This information is in the 0.5 and 1 ms time channels in Figure 7 and the 0.5 through 15 ms time channels in Figure 8. A positive peak in the profiles indicates that the target is to the right of the borehole, while a negative peak or depressed response with respect to the host response indicates that the target is between the borehole and the trans­mitter.

Horizontal conductor

When the conductor is horizontal and in free-space (l05

0-· m host), the borehole profiles (Figure 9) are considerably different in shape from those of the vertical conductor (Figure 6). However, the shapes of the profiles are again explained by

0) b)

-80-140

0>- \'Or I I I \ I II \ \ 25 /5 9 5 3 1.5ms 25 /5 'I I I I II

100>- \ .1001- I I I I I J I \ \ \ \

2001- I, / /

\ I I \\ / J I I'

":"7-9~ 7 J.l J I300>­ - \'_, , \, .... Ii ::t)))J) ....1 j I 1 \ (I

4001­" r I I II .... ' " ',', , \ " I I I ,'1\ \ \ \' \ \ 25 15 9 5 J ts5001­

25 15 9 5 j 1.5 I I J I I II

6001­

500f- I I I 1 I 11

6001­

7001-­700~

800 1 I8001 ! I I ! I I I 10.7 10-b ,o-~ 10-4 '0- 3 '0. 2 10-' 100 io' 10- 7 10. 6

(nV/m 2 )

their vortex-current sources. In this case, the eddy currents flow horizontally. Their magnetic field can be visualized by rotating the vertical body and its field in Figure 2a by 90 degrees. Thus, in boreholes on either side of the body ( - 140 and 140 m) there is a strong positive anomaly at the level of the body. For all three boreholes illustrated in Figure 9, the anomalies above and below the body are negative as expected. However, when a borehole intersects the body as shown in Figure 9b. t he profile is different from those of Figures 9a and 9c, which are for boreholes that do not pass through the body. Inside the body, the profile reverses in sign at 1 ms because eddy currents diffuse into the conductor past the receiving coil. Finally, all the profiles in Figure 9 decay exponentially with the same time constant as that of the vertical conductor in free-space: 1.28 ms.

Note particularly that the free-space response of the hori­zontal conductor does not provide information on whether the target is between the borehole and the transmitter or beyond the borehole if only one transmitter loop is used. However. as noted by West and Ward (1988), the vortex re­sponse changes sign if the transmitting loop is positioned over the horizontal body, because of the change in direction of the primary field. If the conductor were dipping, part of its re­sponse would be asymmetric; and therefore there would be information on the location of the conductor with respect to the borehole and transmitter (Dyck and West, 1984). Direc­tional information is important because it is used to locate the next borehole in a drilling program.

Results for the ]000 n· m host are shown in Figure 10. Because the body is not well coupled inductively to the trans­mi tting loop, we decreased its resistivity to 0.1 Q. m to ob­serve a vortex response. The early-time anomaly, which is positive to the! left of the horizontal body and negative to the right, is due to galvanic currents trowing through the long direction of the body toward the viewer in Figure lb. The galvanic response of the horizontal conductor is important in

\ \ I II \ \ I II 9 5 J , .5/u \ \ \ \ \ \ \ \\ \ \ \ \ \

I / , I I /{ I I I / I 'I

! I I ! I I 10- 5 10. 4 '0-3 10-2 10-1 10° 10'

(nV/m 2 )

c)

140

01- I' \ I I 1\ ,\ I I I "

25 /5 9 5 3 1.5115 100r I I I I I II

J / / I I t t / / / / I J/

200~~ ............................~

25 IS )9 )5)J}J_) ~5300~ )) ~

4001- //

", "\ " ''\ " \ \ \ \ \ \ I 1 II500~ 21,'5 9

\ 5 3 /.5

I I I , I I I (/

6001­

'001­

800! I I I I I I I 10. 7 ' 0 ' 6 10-S 10.4 10-3 '0-2 10·\ 100 10'

(nV/m 2 )

FIG. 9. Borehole profiles for the horizontal conductor of Figure 1b in a lOS n· m host that simulates free-space: (a) borehole between transmitting loop and conductor, (b) borehole through conductor, (c) borehole beyond conductor. The sign of the; peak response is the same on both sides of the body; thus, for a single transmitting loop the vortex response provides no information about the direction to the conductor.

606 Newman at al.

a) b) c)

-140 -80I

140

1~"

I I -s..~, I I\,'

u

\ L!I \ \ \ \ , 5100~ 100.5/11S . I 1002~ 3

5 3 I

, I ...... 200 \ \

I ) I

I I

I J I

200'iS~\ I.s J_~).

""'" <25~ )3~0Q. 300 Ql ~ '\ ~1

400 400115/ I ... t\ 2J!\500f­ 500 500/;1 5 I2S1 II __ 9-'1,VI \

600 600''1 ./

700 700 700

80~o·, 10'3 10. 2 10'\ 100 ic' 102 103

(nVlm 2) ("Vim 2) ("Vim 2)

FIG. 10. Borehole profiles for the horizontal conductor of Figure 1b in a 1000 n· m host: (a) borehole between transmitting loop and conductor, (b) borehole through conductor, (c) borehole beyond conductor. The resistivity of the body is decreased to 0.1 n· m to emphasize the vortex response, which can be seen at late times. The signs of the early-time galvanic responses are different on opposite sides of the body, thus providing information about the direction from the borehole to the body.

a practical sense: it enables the interpreter to determine in which direction the conductor lies with respect to the drill hole. In free-space, where the response is purely inductive, this information is not available. Because of the high conductivity of the conductor, the vortex response dominates at later times with a behavior similar to that of the free-space response in Figure 9. However, when the resistivity of the target is in­creased to I Q. m (not shown), this vortex response is not large enough to be seen in the presence of the half-space re­sponse.

DISCUSSION

In free-space, the shapes of the borehole profiles for vertical and horizontal bodies do not vary with changes in the re­sistivities of the bodies. However, the shapes of the profiles in Figures 6 and 9, which are for free-space, change drastically when the bodies are placed in a conductive host (Figures 7 and 10). Our model simulations show that the galvanic re­sponse is important for host resistivities as high as 1000 n .m, especially at early times. At later times, the vortex response is observed if the body is a good conductor and is well coupled to the transmitter.

It is obvious that simple free-space interpretations of the profiles in Figures 7 and 10 are not possible. Subtracting the host response from the profiles does not eliminate the galvanic response. OnJy after the galvanic response has vanished could free-space models (Dyck and West, 1984) or models based on equivalent current filaments in free-space (Barnett, 1984) be used with confidence. For example, after 9 ms, the response of the horizontal conductor in a tOOO n· m half-space (Figure 10) is similar to its response in free-space (not shown).

In practice, attempts often are made to estimate the host response and remove it prior to applying free-space interpreta­

tion techniques. However, estimating the host response accu­rately is not easy with field data; and small errors in removing the host response could lead to an erroneous interpretation. As an example, Figure 7 shows that, because the host response is large, small errors in removing it could result in large errors in the estimated secondary field at 3 and 5 ms. The response for the more conductive host of Figure 8 is dominantly gal­vanic at late times. Hence, a free-space interpretation of the estimated secondary field obtained by removing the half-space response would not be valid.

The galvanic response provides information on the position of the conductor but does not provide much information on the conductor's geometry. At early times, the profiles in Fig­ures 7 and 10 are similar for vertical and horizontal conduc­tors: compare the profles at - 140 and 140 m at 0.5 and 1 ms. The best information on the geometry of the conductor comes from the vortex response, which is observed after the galvanic response in these figures. Employing multiple-source loops is important for determining the position and geometry of a conductor (Dyck and West, 1984).

Numerical modeling of 3-D borehole TEM responses for applications in mineral exploration requires a large amount of computer time. On average, each model presented in this paper required 12 hours of CPU time on a VAX-11j780 com­puter. In spite of these demands, numerical models have con­tributed significantly to our understanding of 3-D TEM fields.

While 3-D scale models also have contributed to this under­standing, numerical models are more versatile. For example, the study of vortex and galvanic currents presented in this paper could be carried out only with a numerical solution. Since computing the models in this paper, we have developed a new version of the solution for bodies with two vertical symmetry planes but still with arbitrary conductivity contrasts and transmitter-receiver configurations. This new version is up

607 Borehole Transient EM Responses

WEST awl Source loop

Near edge of source loop

\a:::ea.aIiWaGl£'ijiiigiilWAfJJ:ijijMiiamWi 'I- 4f3Ol3Ai!iA\ARA~tiLS4gei fA 5 T

Depth (m)

FIG. 11. Inferred geometry for the borehole TEM data of Figure 12 (based on description in Lane, 1987).

20

40

60

...,. 80 ~

t 100 III

a 120

140

160

180 o 0' ..., 0' '" - - n o 0 0 ~

,I I '"

-10 -5 0

pVIA

FIG. 1-2. Bor-ehole TEM- pr-ofiles- due to pyrite-pyrr-hotite- con­ductor in the Kanmantoo trough, South Australia (after Lane, 1987). As in our simulations, the response appears to be gal­vanic at early times and inductive at later times.

to ten times faster and requires only a fourth of the memory for the same model.

FIELD EXAMPLE

Lane (1987) presented borehole TEM data that show fea­tures similar to those of our computer simulations. The field data were collected in a borehole passing through a massive pyrite-phyrrhotite body in the Kanmantoo trough. South Aus­tralia. A surface Sirotem survey indicated the presence of a tabular conductor with a steep westerly dip. A hole drilled to test the anomaly intersected two sulfidic horizons within a sequence of quartz-biotite schists and quartzites. Based on

J( It: lIll .IC .. lit Vl

Lane's E\1 conductivity and magnetic susceptibility logs and his interpretation, we infer the geometry shown in Figure 11. The EM conductivity log indicates that the main conductor responsible for the TEM anomaly occurs between depths of 150 and 165 m in the drill hole. Because the drill hole is inclined and intersects a complicated, dipping body, we pre­sent the field data not as an exact match to our results but to show how the insight produced by our simple computer models can be applied to understand anomalies encountered in practice.

Borehole TEM profiles are shown in Figure 12; the trans­mitter is a 200 by 200 m loop located as shown in Figure 11. In the earliest time channel, a broad negative anomaly is ob­served (the coordinate convention here is z negative down­ward), which is very similar to the galvanic response of Fig­ures 7 and 10 (inspect the 0.5 ms profile at station 140 m). Consistent with this interpretation is the report by Lane (1987) of a significant host response in the early-time channels of the profiles, even though the host rocks are reported to have re­sistivities in the hundreds of ohm-meters. If the early-time anomaly in Figure 12 is galvanic, we interpret between the borehole and transmitter one or more conductors. where host current is being channeled.

At later times, two crossover anomalies become apparent in Figure 12. It is likely that the upper. stronger crossover anom­aly is a vortex response from the upper part of the main conductor shown in Figure t 1. Vortex currents are con­centrated in the upper part of the conductor because it is closer to the transmitter. The deeper anomaly occurs at a depth of 150-160 m and hence probably is due to the fact that the borehole intersects the main conductor at that location. Because the shape of the vortex response did not change with transmitting loop position, the conductor must be a tabular body, with eddy currents always flowing in its plane (Lane. 1987).

In summary, the galvanic response of the conductor is ob­served at early times followed by the vortex response at later times. This behavior is similar to that of our model simula­tions. Further. the geometry of the conductor is difficult to resolve at early times based on the galvanic response; only the approximate location of the conductor can be interpreted. This result also is consistent with the conclusions drawn from our model simulations.

CONCLUSIONS

3-D models have provided new insight into the effect of a conductive host in borehole TEM surveys. Vortex and galvan­ic currents were studied and related to their corresponding borehole TEM responses. In free-space. the galvanic response vanishes; however, for the models considered in this paper, galvanic response can be important for host resistivities as high as 1000 n· m, particularly at early times. In general, the relative magnitudes of the vortex and galvanic responses are determined by the conductivity contrast and the geometry of the conductor. The controlling parameters are the induction and channeling numbers (West and Edwards, 1985). Based on our results, it is evident that the traditional method of inter­preting borehole data using free-space models must be altered if the host rock is conductive. We have provided some of the

608 Newman e' at.

necessary information for understanding borehole data in these instances.

Naturally, this study is limited in scope by tile number of models discussed. A more extensive study should include a wider selection of 3-D models, such as cubes and thick prisms. The effect of loop position should be in vestigated ; e.g., current channeling is expected to be much less important for bodies located directly beneath the transmitting loop (Spies and Parker, 1984). The influence of conductive overburden also needs to be addressed. Finally, more work is needed to find a means of interpreting borehole data collected in conductive environments. We did not address these issues.

ACKNOWLEDGMENTS

Financial support was provided by Amoco Production Co., ARea OiL and Gas- co.. Chevron. Resources, Company" eRA. Exploration Pry Ltd, Standard Oil Production Co., and Unocal Corp. Additional support for one of the authors (G. A. N.) was provided by the Alexander von Humboldt Foundation (AvH), West Germany. We thank Associate Editor Brian Spies, Dave Fitterman, Phil Nelson, and two anonymous reviewers for suggestions that improved the paper.

REFERENCES

Augustin. A. M., Kennedy. W. D., Morrison, H. F., and Lee, K. H., 1989, A theoretical study of surface-to-borehole electromagnetic logging in cased holes: Geophysics, S4, in press.

Barnett, C. T., 1984, Simple inversion of time-domain electromagnetic data: Geophysics, 49, 925-933.

Boyd, G. W., and Wiles. C. J., J984, The Newmont drill-hole EMP system-s-examples from eastern Australia: Geophysics, 49, 949-956.

Dyck, A. V., and West, G. F., 1984, The role of simple computer models in interpretations of wide band drtll-hole electromagnetic surveys in mineral exploration: Geophysics, 49,957-980.

Ead ie, T.. and Staltari. G., 1987, Introduction to downhole electro­magnetic methods: Expl. Geophys., lH, 247-351.

Eaton, P". A.. and' Hohmann, G. W:, 1984; The influence ofconductive host on two-dimensional transient electromagnetic responses: Geo­physics. 49, R61-869.

Kaufman, A. A.. and Keller, G. V., 1985, Inductive mining prospect­ing: Part I: Theory: Elsevier Science Publ, Co.

Lamontagne. Y., 1975, Applications of wideband time-domain elec­tromagnetic measurements in mineral exploration: Ph.D. thesis, Univ. of Toronto.

Lane, R. J. L.. 1987, The downhole EM response of an intersected massive sulphide deposit, South Australia: Expl. Geophys., 18, 313­31 ~.

Lewis. R., and Lee. T.. 1978, The transient electric fields about a loop on a halfspaee: Bull., Austral. Soc. Expl. Geophys., 9,173-177.

McNeill. 1. D.. Edwards. R. N .. and Levy. G. M., 1984, Approximate calculations of the transient electromagnetic response from buried conductors in a conductive half-space: Geophysics, 49.918-924.

Nabighian, M. N.. 1979. Quasi-static transient response of a conduc­ting 11<llr~spHce-An approximate representation: Geophysics. .w~ 1700-1705.

Newman. G. A.. and Hohmann, G. W., 1988, Transient electro­magnetic responses of high-contrast prisms in a layered earth: Geo­physics, 53. 691- 706.

San Filipo. W. A., Eaton, P. A., and Hohmann, G. W., 1985, The effect of a conductive half-space on the transient electromagnetic r..:srOI1SI: of a three-dimensional body: Geophysics, SO, 1144-1162.

Spies. B. R.• and Parker, P. D., 1984. Limitations of large-loop tran­sient electromagnetic surveys in conductive terrains: Geophysics, 49.902 912.

West. G. r., and Edwards, R. N., 1985. A simple parametric model for the electromagnetic response of an anomalous body in a host medium: Geophysics, SO, 2542--2557.

West. R_ C. and Ward, S. II., 1988. The borehole transient E\1 response of a three-dimensional body in a conductive half-space: Geophysics, 53, 1469-1478.

Woods, D. V., 1975, A model study of the Crone borehole pulse electromagnetic system: M.S. thesis, Queen's Univ., Canada.