Again the authors state that it is assumed ments to contend with because our tests question are so much
s question are so much alike that one is forced
that all repairs can be made on prearranged
interruptions and that work will be per­
formed under favorable conditions, but it
must be recalled that many of the lines in
question are single circuit, which means
there is no such thing as a favorable time
for interrupting the service of the line.
In fact, to get one of the lines out of service
frequently m a y mean preparations lasting
days to get old or uneconomic plants
properly manned and on the line, and
even then the service* would be subjected
to a considerable hazard. The fact of the
matter is that conductor burning is one
of the few things that cannot be tolerated
in the operation of an important line, if
the entire investment (not alone the invest­
ment in a very minor item like a ring
assembly) is not to be nullified b y its un­
reliability.
The discussers again are in error in
assuming a cost of $6.00 to change an insu­
lator on a smashed insulator string. For
work carried out in quantity, the figure
might be reasonable, but for sporadic and
spasmodic repairs the actual cost is more
likely to be 4 or 5 times that. Again, in
figuring cost of investment, the authors
assume a necessary return of 15 per cent.
This would be excellent if it could be o b ­
tained, but it isn't obtainable today to
utilities in general on the entire plant
investment, and there is no reason why
such a cost figure has to be assumed in
carrying insulator protective equipment.
Another item of expense the discussers
failed to consider is the more frequent line
inspection required on a line not equipped
with arcing protection, if any attempt is to
be made t o keep the line conductors and
insulators in a condition even approximately
that existing on the protected line.
The above are only some of the many
errors of omission and commission that
render valueless the elaborate tabulation
presented.
The remaining conclusions of the dis­
cussers, such as that the amount of money
that is warranted for arcing protection
increases as the expected number of flash-
overs increases, or that each transmission
project should be discussed individually,
are of course t o o axiomatic to need any
further comment.
Theory and Tests
of the Counterpoise
Discussion a n d author's closure of a paper b y
L. V . B e w l e y p u b l i s h e d in t h e A u g u s t 1 9 3 4
issue, p a g e 1 1 6 3 - 7 2 , a n d presented for oral
discussion at t h e lightning session o f t h e
Pacific Coast c o n v e n t i o n , Salt L a k e C i t y ,
Utah, September 5 , 1 9 3 4 .
C. L. Fortescue (Westinghouse Elec. and
Mfg. Co., EJPittsburgh, P a . ) : This paper
is particularly interesting to me because it
reports a series of tests similar to those
made at Trafford b y M r . Fielder and my­
self, and similar troubles and disappoint­
ments were encountered on account of the
low resistivity of the soil on the site of the
test. In addition to these troubles, how­
ever, in making our tests we had the ele­
228
were made during the months of January,
February, and March, and the winter in
Pittsburgh was unusually severe.
In general the tests reported in this paper
lead to the same conclusion as those re­
ported in our paper. Neither set of tests
will enable the engineer contemplating a
future transmission line or considering im­
proving an existing line to determine
whether it is more economical to use coun­
terpoises or t o reduce the tower footing
resistances b y other means, such as driven
grounds. In other words, to obtain a logi­
cal answer he will be compelled to make
some kind of comparative tests, for which
he probably is not properly equipped, or
else use his judgment based on the experi­
ence of those who have used counterpoises
(of which there is very little statistical data
at the present time) and wait for years to
have his judgment confirmed or contra­
dicted.
How much better it would be if similar
tests to these could be made jointly b y all
interested utilities to determine the behav­
ior of counterpoises in representative loca­
tions. I feel sure- that the problem could
be solved satisfactorily b y half a dozen such
tests carried out on sites having different soil
conditions. The state of California offers
exceptional opportunities as it presents a
great variety of soil conditions. Transmis­
sion lines are no doubt available and in
process of construction where such tests
could be made covering a large variety of
soil conditions at very little expense. T h e
first tests on system stability were made in
California and, as a result, the solution of an
important problem of system design was
achieved. T h e lightning problem is just
as important from the system design stand­
point as the stability problem, indeed it is
an important element in the general prob­
lem of system stability as lightning proof
lines will enable transmission lines to be
loaded close to their static stability rating.
For long transmission distances it is hard to
contemplate a line over some portion of
which counterpoises will not be needed.
Bewley brings in his artificial lightning
surge over a short length of line (3,077 feet),
a method which is to be highly recom­
mended. For a section of experimental line
which does not extend in 2 directions from
the point of attachment of the surging line
the effect should be equivalent to a surge im­
pedance of the lightning channel of about
250 ohms. A t Trafford we would have
liked to do the same but space limitation
prohibited it. The method of obtaining
the velocity of propagation in the buried
counterpoises is extremely interesting and
the results obtained are about what would
be expected. There never has been any
doubt in m y mind that the velocity of propa­
gation of the main surge in the counter­
poise is relatively slow as compared to that
in the line and our tests indicate that the
velocity in the line is considerably less than
the speed of light. T h e effect of both
leakance and the high dielectric constant of
the earth is to slow up the velocity of propa­
gation.
On one point I d o not agree with the
author. In figure 11, part II, he shows the
same surge potential for ei and e , although
s
d applies to the overhead line and d to the
buried counterpoise. Again in figure 12 he
shows similar results. T h e 2 surges in
to the conclusion that they are identical and
the assumption has been made that e\
and e* have the same value. Repeated
tests have been made b y us which demon­
strate that such an assumption is not valid.
In our tests at Trafford we found differences
of potential due to the down drop lead from
line to counterpoise of appreciable magni­
tude. In fact, in making measurements of
surge impedance of grounds and counter­
poise we found it necessary to place our
measuring terminal right at the counter­
poise because even a short length of line
(about 70) between the counterpoise gave
incorrect results.
L. V . y B e w l e y (General Electric C o . ,
Pittsfield, M a s s . ) : Hagenguth's thorough
analysis of our test data has resulted, I
think, in 2 outstanding contributions to our
knowledge of counterpoise behavior; first
with respect to front of the wave phenom­
ena, and second with respect to the simpli­
fication of counterpoise calculations. I will
discuss these 2 aspects of the problem
briefly.
F r o n t o f t h e Wave P h e n o m e n a
In this and 2 previous papers ("Attenua­
tion and Distortion of W a v e s , " L. V . Bewley,
E l e c t r i c a l E n g i n e e r i n g , December 1933,
page 876-84. " T h e Counterpoise," L . V .
Bewley, General Electric Review, February
1934) it was shown that the earth current
must penetrate at a rapid rate immediately
following the passage of the wave front,
and then at a more gradual rate. After
the "readjustment transient" at the wave
front the effective current depth m a y be
considered as constant as far as induction
on the conductors is concerned, because
the inductance coefficients change very
little with an increase of current depth.
From this consideration of the character
of the earth current transient, several im­
portant effects could easily be explained
in a qualitative way, but Hagenguth,
from his careful and detailed study, has
obtained actual quantitative data on the
subject. In our original tests separate
oscillograms of voltage and current were
taken, and from them the surge impedance
computed as a function of time b y taking
the instantaneous ratios of e/i. In the case
of the overhead wires and the insulated
counterpoise the impedance function as
thus computed appeared to start at a high
initial value, quickly fall to a minimum
value, and then slowly increase. T h e
explanation for this behavior was given
in figure 13 of m y paper. However, in the
case of the buried counterpoise it was not
at all clear whether the impedance started
at a high value, fell to a minimum, and then
slowly increased; or started at a l o w value,
rose abruptly in a short interval, and then
continued slowly to rise. Either interpre­
tation could be arrived at b y a very slight
mismatching of the fronts of the voltage
and current waves. Hagenguth there­
fore suggested that oscillograms be taken
with voltage and current, respectively, on
the 2 plates of the oscillograph. W h e n
this was done there was n o doubt that the
impedance of the buried counterpoise starts
at a very low value and then rapidly rises
t o its path of relatively slow increase.
ELECTRICAL ENGINEERING
This is n o t inconsistent with the explana­
tion for the insulated counterpoise given
in figure 13 of m y paper. In the case of
the insulated counterpoise 1 foot above
the surface of the earth, the earth current
could get no closer than within 1 foot
of the counterpoise, and therefore the
G R O U N D X W I R E
minimum inductance was comparatively
large, so that the change of capacitance Z| Zl

lip
exerted a dominating effect at the wave C O U N T E R P O I S E
front. With the buried counterpoise the
earth current is initially very close t o the
conductor, the inductance is practically
L
U
Z 3
7 ' 3 | R

nil, and consequently the surge impedance

f(",A)
starts at nearly zero. T h e great influence
of the initial inductance on the shape of
the impedance curve is illustrated in 1*3
figure 1 of this discussion.
The penetration of the current into the ..e \ — v _ a > \ / : — , |f 5 L « 2*CZ -R ) 3 3

earth changes the inductance, and in conse­

quence changes the surge impedance, veloc­
ity of propagation, and coupling. It ap­
pears permissible t o ignore the change of INITIAL S U R G E IMP.

capacitance in the case of the buried — V - Z J ^ ^ ^ i ^ . — R 3 «L E A K A G E RESISTANCE!

counterpoise. W e m a y then summarize
••2 - Z a ^ ^ ^ - ^ ^ S ^ -
the effect of the current penetration in
table I of this discussion.
From these considerations it is easy to
visualize the flattening of the wave front, 0 .1 . 2 . 3 . 4 . 5 . 6 .7 . 8 . 9 1 . 0 I.I 11 . 22
lengthening of the wave tail, decreased <r= T 2a» ( W A V E F R O N T S X 1 0 0 0 X R3/Z3
coupling of a short wave, etc. Hagen- " 6 l 2 3 6 ( L E N G T H O F C O U N T E R P O I S E ^
guth in his discussion has given curves and
tables for the effect of the current pene­ Fig. 2 . Lightning voltage at t o p of tower permitted b y a counterpoise
tration. H e has also suggested that the
reduced coupling in figure 12 of the paper, 2 E -Aa3t/T~l

as compared with that in figure 11, is n o 2

1 + Z/r ^ A ( ^ X " / T
( A ^ ) J
doubt due to the smaller depth of current 2 E
penetration associated with the shorter e » i . = ^ _|_ 2 / ' ' K / A ) <r
m r
wave of the higher voltage surge t o which
A = ( 1 + Z/r0/(1 + Z/r) T = Front o f incident lightning w a v e
figure 12 applies. It appears, therefore,
that in regard to the front of the wave 1/r = ( N / Z , + 1/R + 2 / Z O N = Number of counterpoise wires
phenomena, we know a great deal more in 1/r' = ( N / R , + 1/R + 2 / Z i )
a quantitative way than heretofore.

from 4 t o 8 per cent, a n d t h e o r y suggests t h a t it of transition requires a p p r o x i m a t e l y one c o m p l e t e

SIMPLIFIED COUNTERPOISE CALCULATIONS will n o t exceed 10 per cent. reflection on t h e counterpoise, a n d since t h e v e ­
2. T h e initial surge i m p e d a n c e o b t a i n e d b y e x t r a p o ­ locity on t h e counterpoise is close t o /z t h e v e l o c i t y 1

From our own theory and tests, as well lating t h e i m p e d a n c e c u r v e t o zero t i m e appears of light, this m e a n s t h a t t h e transition requires
6
as tests made in Michigan in 1930 and tests t o b e of t h e order of from 1 0 0 t o 2 5 0 o h m s , with an
I = ^ (length of counterpoise in f e e t ) .
a v e r a g e v a l u e of 1 5 0 o h m s . T h e M i c h i g a n tests
by Fortescue and Fielder, 3 facts stand out
g a v e 2 2 0 o h m s , t h e Pittsfield tests from 1 2 0 t o 1 6 0
rather clearly: o h m s , and t h e Trafford tests 1 5 0 o h m s . Hagenguth suggests in his discussion
1. T h e additional coupling effect on the line wires 3. T h e i m p e d a n c e falls in a r o u g h l y exponential that the additional coupling due t o the
due to a counterpoise is rather small in comparison fashion f r o m its initial surge i m p e d a n c e v a l u e t o its counterpoise be ignored entirely and that
with t h a t of the ground wire. W e found not more final leakage resistance (which m a y be s o m e w h a t simplified calculations be based on regarding
than 4 per cent while F o r t e s c u e and Fielder show less t h a n t h e measured d-c resistance^. T h e time
the counterpoise as an impedance varying
from an initial value Z t o a final value R* 3

in time t = 6/. H e has calculated the

impedance curve b y considering half that
portion of R uncovered b y the advancing
s

wave front t o be in parallel with the in­

Z ^ _ _ — = = stantaneous Z . • 3

400
40 In this connection I propose an equiva­
lent circuit for the counterpoise consisting
of its leakage resistance R* in series with
BURIED COUNTERPOISE
INSULATED COUNTERPOISE ( Z — i? ) shunted b y an inductance L such
3 3

300 300 that the time constant of the circuit is

z z r z t £
-+ essentially the same as that of the actual
counterpoise. On this basis a complete
.„ > ^ ,

>X<2
r , „ . ,
IT : set of estimating curves m a y b e obtained
as shown in figure 2 of this discussion, so
200
I L
| i i [ . • i that for any assumed or measured values
1 ij"m I I I I ^ z of the voltage of the lightning surge, the
surge impedances of the stroke, ground
" J / L J M O 1 ~ — — — —
^^^=———— 1 1 /LXI'OS —
wire, and counterpoise (initial), and the
100
MIN. / L X I 0 6
100
— A^- /cx ip , 8
I I — tower footing and counterpoise resistances,
I I j-l I p H the corresponding voltage o n the ground
wire m a y be readily found. These curves
urn. J j ' j - - - j c x i o - - - - -
8 8

f^MIN V L X IQG are computed from the equivalent circuit

.6 .8 1.0 1.2 1.4 2 4 .6 .8 1.0 1.2 1.4 of the counterpoise, choosing L so that the
M I C R O S E C O N D S M I C R O S E C O N D S time constant of the equivalent circuit
Fig. 1 . Effect o f initial inductance o n t h e shape o f t h e c o u n t e r p o i s e i m p e d a n c e curve matches that of the actual transient im-

FEBRUARY 1 9 3 5 229
 Tabic I—Effect of Current Penetration the maximum number of counterpoise wires
which may be used in parallel before the
mutual surge impedance between wires be­
Effect on
comes excessive. H e also asks what shape
Penetration
the transient surge impedance curve has
of Current Inductance Surge Impedance Velocity Coupling
when the leakage resistance is higher than
the surge impedance.
Initially close t o surface A m i n i m u m , and A m i n i m u m , and H i g h , a n d practi- A minimum and In accordance with the a b o v e criterion,
of counterpoise very small practically zero cally t h e v e l o c i t y very small
if from 200 to 300 feet are not sufficient t o
of light

Initial d o w n w a r d dis- R a p i d l y increases R a p i d l y increases R a p i d l y decreases Rapidly increases

placement

R e l a t i v e l y slow d o w n - S l o w l y increases S l o w l y increases S l o w l y decreases Little change

ward trend

R e t u r n s u p w a r d to S l o w l y decreases S l o w l y decreases S l o w l y increases Slowly decreases

w a v e tail

TRANSIENT IMPEDANCE IN OHMS

pedance where a unit function current is decreases from 150 ohms t o 10 ohms in
applied. It is not imperative that the about 10 microseconds. If, however, the
exact value of Z be known, since the results
3 1,000 feet is used as 2 500-foot counter­
are not greatly affected b y a departure from poises, the impedance starts at 75 ohms
an average value of Z = 150, but the leak­
3 and reaches 10 ohms in about 5 micro­
age resistance R* must be known from field seconds. Further subdivision reduces the
measurements, either with direct current initial impedance and shortens the time of
or a bridge. transition, b o t h highly beneficial effects.
As an example of the application of these However, it is evident that a region of
curves, take: diminishing returns is approached b y 4
wires* and this is accentuated b y the fact
Z = 400 ohms surge impedance of light­
that for very short wires the end effects
ning stroke
become pronounced, reflections are not
Zi = 500 ohms surge impedance of ground
wiped out, mutual inductance between 2 3 4
wire
wires plays a more important part and, MICROSECONDS
Zz = 150 ohms surge impedance of counter­
most important of all, the leakage resistance Fig. 3 . Effect o f t h e n u m b e r o f wires o n t h e
poise
may exceed the surge impedance for very counterpoise impedance
Rs = 40 ohms leakage resistance of coun­
short wires. The proper length of each
terpoise
counterpoise wire should be such that its
/ = 200-foot length of counterpoise
leakage resistance is substantially less than reduce the leakage resistance below the
N = 2 counterpoises per tower
its surge impedance, but otherwise its length surge impedance, say 150 ohms, then longer
R — 200 ohms tower footing resistance
in thousands of feet need not exceed /« the 1

lengths should be used, and this will usually

T = 1 microsecond wave front
shortest front in microseconds of any lightning mean continuous tower-to-tower counter­
E = 10,000-kv lightning stroke voltage
surge to which it may be subjected. This poise wires. If the leakage resistance ex­
(incident wave)
criterion ensures that the full length of the ceeds the surge impedance, then the tran­
1/r = (N/Zi + 1/R + 2 / Z O = (2/150 +
counterpoise will have been reduced to sient impedance initially decreases, reaching
1/200 + 2/500) = 0.0223
substantially its leakage resistance b y the a minimum approximately in time t =
1/r' = (N/R + 1/R + 2 / Z i ) = ( 2 / 4 0 +
z
time the shortest front wave has reached 6//l,000, then increasing and reaching a
1/200 + 2/500) = 0.0590
its crest, and that the voltage will n o t maximum equal t o the leakage resistance.
1 + Z/r' 1 + 400 X 0.0590 exceed that permitted b y the surge im­ An exact determination of the mutual
A = = 2.5
1 + Z/r 1 + 400 X 0.0223 pedance only. If a longer counterpoise is surge impedances between radial counter­
used its minimum resistance will not be poise wires is rather awkward, but at least
Rz T _ 40 1 X 1,000
= 0.22 available for the shortest front waves, al­ 8 wires may be used without much of a
Z 3 6/ ~ 150 6 X 200 though the additional length would prove boost in surge impedance. If the wires are
Therefore, the maximum voltage on the beneficial for longer fronts, but not so much run parallel along the right of way, a
ground wire is so as is obtainable b y using more of the minimum spacing of 10 feet appears per­
shorter wires. These considerations sug­ missible.
2E . . . . 20,000
gest that the standard counterpoise length Fortescue suggests that additional field
24.6 should be from 200 t o 300 feet (if this is tests be carried out in soils of different
X 1.55 = 1,260 kv sufficient to reduce the leakage resistance characteristics. While these tests would
Allowing 30 per cent coupling between below 150 ohms) and that as many wires be very desirable as confirmation of the
ground wire and line wire and 5 per cent be used as necessary to secure the desired existing data and theory, yet I d o not at
additional coupling due t o the counter­ reduction in voltage as given b y the curves all believe that such tests are necessary
poise, the voltage across the insulators of figure 2 of this discussion. before a rational counterpoise analysis can
would b e : Waldorf points out in his discussion be made. Since m y paper was written,
that in many actual installations a length Hagenguth has made a careful analysis
1,260(1 - 0.35) = 820 kv. of from 200 t o 300 feet is n o t sufficient t o of our records and reported his findings in
reduce the final leakage resistance below his discussion of this paper; and I have
M u l t i p l e C o u n t e r p o i s e Wires the surge impedance, and he inquires as t o extended m y calculations in this discussion

Since it has been shown that the addi­

tional coupling due to the counterpoise is Table II
of secondary importance as compared with
its impedance effect, it is at once apparent Waves leaving starting point AM AM AM
that a given length of wire is more efficiently Time of arrival at transition point l/Vi l/Vi l/Vi
employed as a number of radial counter­
Reflections of velocity (v\) tfi'M Ji'M
poises than as one continuous wire. Figure Time of arrival at starting point 2l/vi 1/V2 + l/Vi l/v - f l/Vi
3 of this discussion shows 1,000 feet of 3

Reflections of velocity (v ) 2h'M sA'M

wire having a final leakage resistance of
2

Time of arrival at starting point 2l/vt

10 ohms and an initial surge impedance Reflections of velocity (i» ) 3
l/Vl
ift'M
+ 1/V2 2f*'M
l/vz + l/vt
zA'M
of 150 ohms per wire. If this is used as a Time of arrival at starting point
single long counterpoise the impedance
1/V2 4- l/v% 2l/vz

230 ELECTRICAL ENGINEERING

to include successive reflections into the Table I V
second and third generations. The agree­
ment with test results is so g o o d in all
Condition I Condition II Condition III
respects that I no longer have any doubt
that our theory is sufficient and adequate
for making routine design calculations once e\ + ei' - ft' i\ + i\ = 0 ei + e\ = ei'
the leakage resistance is known, and this i\ + i\ H + H* = 0 ex + ez = e,'

can be measured quite easily.

Fortescue calls attention t o the fact
H
h
+ H'
+ i%
=
=
0
0
i%
ez +
+ iz
e' z
-
= ft'
*'/Zi
it
ii
+ *Wi ' +=
+ ii
0
+ tV = eS/Z
that I used the same oscillogram for the
ground wire surge and the counterpoise
surge in part III, figures 1 1 and 1 2 of m y
paper. The height of our pole was only 8 4 per cent of the velocity of light) move and solving these simultaneous equations
3 0 feet and as our wave front was about out on the ground wire, line wire, and for each condition, there finally results:
0 . 4 microsecond we could observe no counterpoise. These waves arrive at the
difference. Furthermore, it must not be transition point (grounded end of the
thought that connecting the oscillograph to counterpoise) at different times, and as V o l t a g e s D u e to R e f l e c t i o n s o n
Re-
the overhead wire b y a separate down lead each wave arrives it generates reflected
C o n - flection Ground Line Counter­
gives any truer reproduction of the surge waves of all 3 velocities. Thus 9 individual dition G r o u p Wire Wire poise

on the overhead wire than does connection waves start back toward the origin, leaving

1
to the tie between the ground wire and the transition point at different instants in A - 0 . 0 0 3 1 £0 - 0 . 0 2 0 5 Eo 0.0007 Eo

counterpoise as used in m y tests because 3 separate groups, each group consisting of B 0.0225 Eo 0 . 0 3 9 7 Eo 0.0574 Eo
C - 0 . 0 2 0 7 £0 - 0 . 0 3 6 3 Eo - 0 . 1 2 1 1 Eo
there are successive reflections in both cases. 3 separate wave velocities. See table I I

11
A 0 0 0
of this discussion. B 0.0783 Eo 0 . 0 7 8 3 Eo - 0 . 0 3 4 7 Eo

CALCULATION O FSUCCESSIVE R E F L E C T I O N S
In view of the fact that v\ and v are so 2
C - 0 . 1 4 7 9 £0 - 0 . 1 4 7 9 £0 - 0 . 2 7 4 0 £0
III A 0.0076 Eo - 0 . 0 3 2 8 Eo 0.0076 Eo
nearly equal we can greatly simplify the
B 0.0329 £0 0 . 0 9 8 9 Eo 0.0330 Eo
In connection with figure 4 of m y paper work b y dealing with only 3 wave groups C - 0 . 2 8 4 6 Eo - 0 . 2 0 1 0 Eo - 0 . 2 8 3 3 Eo
it was pointed out that the reflections o b ­ with respect to their time of arrival at the
viously showed the chara teristics of multi- starting point:
velocity components, and in the text the
initial multivelocity wave components were Group A: (if/ - i / ' - / i ' ~ 2 2 2 / 2 ' ) returning The contribution of these reflection groups
calculated. Since the paper was written approximately at (21/vi ^ 2 / / v ) a
is easily identified on the oscillograms of
the calculations have been extended to Group B: ( 1 / 3 ' - 3 / 1 ' - 2 / 3 ' - 3 / 2 ' ) returning figure 4 of m y paper. For example, con­
include the reflections, and the numerical approximately at (l/vi + l/vz) sider the voltage e on the line wire under
2

agreement with the oscillograms in all Group C: ( 3 / 3 ' ) returning at (21/v ).

3
condition I I :
details has been found to be excellent.
The complete calculations are entirely too The transition point equations applying
long to be reproduced here, but a brief at the grounded end (resistance R) of the c 2 Condition I I Calculated Test
resume and summary of the results seem counterpoise are
desirable, since, as far as I know, this con­ Original W a v e 0 . 0 6 4 0 . 0 6 8

stitutes the first evidence of the exact

ei + e'
x = ft' + Groups A and B 0 . 1 4 2 0 . 1 3 6

quantitative agreement obtainable b y multi-

e 2 + £2' = e2
ez + ez = R(iz +
velocity theory, and should therefore put
i\ + i\ = ii
at rest those intimations raised during the
discussions of m y first paper (Attenuation ii + i% = ii
and Distortion of Waves, L. V. Bewley, Substituting the multivelocity components,
E L E C T R I C A L E N G I N E E R I N G , December 1 9 3 3 , these 5 simultaneous equations are then
and the discussions at the 1 9 3 4 winter con- solved for the reflections. Corresponding
Table III
+ Group C - 0 . 0 0 6 0 . 0 0
*V)
All points on these oscillograms check very
well with the above calculations, within the
limits of accuracy of this kind of work. It
has, therefore, been shown that the multi-
velocity theory of traveling waves gives
correct quantitative results.

Reflected Lightning Investigation

Wave Group Condition I Condition II Condition i n
on Transmission L i n e s — I V
I/i'M A 0 . 0 1 4 5 EO 0 0 . 0 1 3 5 £ 0 Discussion o f a p a p e r b y W . W . L e w i s a n d
2/1'M A 0 0 0 C . M . Foust p u b l i s h e d i n t h e A u g u s t 1 9 3 4 is­
*fi'M B 0 0 0 sue, p a g e 1 1 8 0 - 6 , a n d presented for oral dis­
1/2'te) A 0 0 0 cussion a t t h e lightning session o f t h e Pacific
MM A - 0 . 0 0 2 0 EO 0 . 0 0 0 9 EO - 0 . 0 0 1 0 £ 0 Coast c o n v e n t i o n , Salt L a k e C i t y , U t a h , S e p ­
3/2 '(v2) B 0 . 0 1 1 3 £ 0 0 . 0 4 4 0 £ 0 0 . 0 5 2 6 £ 0 tember 5 , 1 9 3 4 .
if*'M B 0 0 0

ih(vz) B 0 . 0 1 1 2 EO - 0 . 0 0 5 1 £ 0 0 . 0 0 5 7 £ 0

3/3'fe) C - 0 . 0 2 0 6 EO - 0 . 0 8 0 5 £ 0 - 0 . 0 9 6 4 £ 0 C. L . f o r t e s c u e (Westinghouse Elec. and

vention published in the March and April
1 9 3 4 issues, page 4 7 1 - 3 and 595-8) on
multi-velocity theory to the effect that the
mathematics was not a true description
of the phenomenon.
The insulated counterpoise used in our
field tests was 9 2 5 feet long, and at its far
end was grounded through driven pipes
having a ground resistance of 2 4 0 ohms.
When the surge is applied at the starting
end of the counterpoise, multivelocity waves
of 3 velocities ( 1 0 0 per cent, 9 8 per cent, and
Mfg. Co., E. Pittsburgh, P a . ) : Lewis and
Foust have reached the conclusion that the
tower top potential is given b y the product
to the 3 conditions I, II, and I I I of figure 4 of tower footing resistance and crest current.
of m y paper, the numerical values are as W e are asked to believe that no matter how
given in table I I I of this discussion. severe the stroke may be, it will cause no
N o w regarding these reflections as incident flashover on a poorly insulated line pro­
waves impinging on the transition condi­ tected with ground wires if the tower footing
tions at the starting point of the counter­ resistance is very low, of the order of 1 or 2
poise, and assuming all components of any ohms. Y e t in our laboratory work using
group to arrive there simultaneously, new artificial lightning surges on actual trans­
transition point equations can be written mission lines, we have observed potentials
for each condition. These equations are of considerable value in 3 0 or 4 0 feet of
given in table I V of this discussion. down drop leads to ground. A large tower
Substituting multivelocity components has, of course, a much lower surge imped-

FEBRUARY 1 9 3 5 231