Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

621.3.011.21:621.395.73:621.392.2 Monograph No.

28
SUPPLY SECTION

THE MUTUAL IMPEDANCE OF EARTH-RETURN CIRCUITS


By L. J. LACEY, B.Sc.
{The paper was first received \3th September and in revised form 23rd November, 1951. It was published as an INSTITUTION MONOGRAPH
\5th March, 1952.)
SUMMARY
Formulae are developed for the mutual impedance of two finite-
length earth-return circuits, either parallel or at an angle. The new
formulae are shown to be generalizations of the Carson-Pollaczek
formula and of the more recent work of Carter. Numerical results are
compared with calculations performed by the methods at present
recommended by the Comite Consultatif International Telephonique.

Fig. 1.—Notation for lines at an angle.


(1) LIST OF PRINCIPAL SYMBOLS
Distances of end points of lines from their point of flowing over power lines are based on the Carson-Pollaczek2'3
intersection or from an arbitrary reference line. formula, augmented by semi-empirical rules to cover the case of
Distance of arbitrary points on the lines from same non-parallel lines.
reference as alt a2, b{, b2. It is assumed in the derivation of this formula that one line is
Distance between two points, one on each line. infinitely long. In 1931 Foster4 published an expression for the
mutual impedance of two infinitesimal segments from which the
Distance between end points of lines. impedance of circuits of finite length can be obtained in the form
12'r21
'*0 Distance between parallel lines. of a double integral. The only evaluation of Foster's integral
Srs ar bs published to date is due to Carter,5 who has obtained an expres-
Par ar sin a sion for short parallel lines at wide spacings. The object of the
Pbr br sin a \ r, s — 1, 2. present paper is to present generalized evaluations of Foster's
Qars ' a
r ^.v C 0 S a integral from which the mutual impedance of any circuits com-
Qbrs br a COS
s a
J posed of straight sections may be found.
A - V(* 2 + ?2).
£, 17 — Transformations of *, 7 defined as used. (3) SUMMARY OF FORMULAE
a Angle between lines measured in the positive Foster's integral for the mutual impedance of two straight
directions. wires on the surface of a semi-infinite homogeneous earth may be
>,</>' Angle between wires and a line joining two arbitrary put in the form
segments.
cr = Earth conductivity, mhos/m. 3 sin <f> sin <f>' — c o s a
[xQ — Permeability of earth (assumed to be that of free Zl2
~ 2W 'a,
space- 1-257 x 10-6 H/m). J
b
y (1 + yr) exp ( - yr
- cos a dxdy . (1)
M
0 '0V'
u) - Angular frequency = 2nf. Carter's formula for short parallel wires at large spacings is
Z{2 Mutual impedance of circuits 1 and 2. obtained from this integral by neglecting the term involving
Rl2 ----- Mutual resistance of earthing points. exp (— yr) and putting a = 0.
K0(z) = Modified Bessel function of the second kind of This gives
order zero. 1
Zn = *« - '" ' r"22 -,2-^21) • • (2)
£• "A X )• Various special functions defined as introduced,
c ----- 0-577 216 . . ., Euler's constant.
in which Ji2 = _L(J_ r
- - - - - ) . . . (3)
The notation relating to distances is illustrated for lines at an 22
angle in Fig. 1. jR12 represents the mutual resistance of the earthing points of
Rationalized M.K.S. units are used throughout. the wires and will appear in each of the formulae to be given.
For reasons discussed in Section 4 it will not generally be
(2) INTRODUCTION included in the calculation.
The methods at present recommended by the Comite Con- The first generalization of this formula is to extend it to lines
sultatif International Telephonique1 (C.C.I.F.) for calculating at an angle, still assuming that the exponential term is negligible.
the voltages induced in communication circuits by earth currents Substituting for sin <f> and sin <f>' in eqn. (1) we find,
•iJbi
Correspondence on Monographs is invited for consideration with a view to
publication.
1 (3xy sin2 a cos
Mr. Lacey is with the Hydro-Electro Power Commission of Ontario, Canada, and Zl2 = • • (4)
2TTCT
was formerly with the British Electricity Authority.
[ 156]
LACEY: THE MUTUAL IMPEDANCE OF EARTH-RETURN CIRCUITS 157
2 2 2
with r — x + y — 2xy cos a (4) MUTUAL RESISTANCE TERM
In the derivation of Foster's integral it is assumed that the
In Appendix 10.1 it is shown that this integral gives
current flows into the earth through point contacts. Hence in
1 cos < the integrated forms we have the term
'12 2TT(T sin2
I\L. a2b2
(5)

Since either o, or 6, may be zero it is important to observe that 2iT(j\r r22 r2l
n l2

b representing the mutual resistance of these point electrodes.


In all practical problems this term must be given special
(6) consideration.
In the first place it should be observed that when a complex
exposure is split into a number of segments for calculation pur-
poses, Rl2 needs to be calculated only for the end points of the
It is worth noting that as the angle a between the lines tends complete lines. For it follows from the form of Ri2 that the
to zero a,, a2, bx, b2 all tend to infinity in such a way that contributions from the end points of intermediate segments will
ax sin a, a2 sin a, b{ sin a, b2 sin a all tend to r0. Thus in the occur twice with opposite signs and will therefore cancel out.
limit we obtain the formula for parallel lines. Secondly, the validity of assuming that the earthing systems
In order to extend the formulae to smaller spacings for which may be regarded as points must be checked in each case. The
the exponential term is not negligible, we have to evaluate the assumption is entirely adequate as far as it affects the general
second part of Foster's integral. For the parallel-line case the earth-current distribution along the length of the lines, but it
required integral is may be far from the truth if used to calculate the mutual resistance
of normal earthing systems placed reasonably close together.
When this occurs Rl2 must either be measured or be calculated
• • (7)
from formulae applicable to the particular case. Again, if the
earthing points are far removed from one another, the point-
with r = V(*2 + y2 ~ 2xy + r2) electrode assumption is reasonable, but the term is in any case
negligible for circuits long enough to be of practical importance.
This integral is discussed in Appendix 10.2. It is shown that In general it is recommended that Rl2 be omitted from the
on combining with the terms for wide spacings we obtain
calculation of impedances, and that, where necessary, the mutual
resistance of the earthing systems be added separately.

V2 V2" (5) LINES ABOVE THE GROUND


All the formulae of Section 3 assume that the wires are on the
(8) surface of the ground.
V2
If one or both circuits be above ground, it is possible to
where ip(x, >0 is a special function defined in Appendix 10.5. take account of their heights by replacing the horizontal separa-
Formulae and tables from which i/»(x, y) may be computed are tion appearing in the formulae by the radial separation. It is
also given in Appendix 10.5. In Appendix 10.4 it is shown that known from calculations using the Carson-Pollaczek formula
eqn. (8) reduces to the Carson-Pollaczek formula when the that, if one line is infinitely long, this approximation gives very
length of one line is allowed to tend to infinity. good results for all but the closest spacings.6 There appears to
The final and most general formula is that for lines at an angle be no reason for supposing that the approximation will be any
and any spacing. To obtain this, eqn. (7) has to be evaluated with worse when both lines are of finite length. For any circuit
heights likely to be encountered in practice the correction itself
y2 — 2xy c o s a) (9) is unimportant when the spacing exceeds 100 m. For small
In Appendix 10.3 it is shown that this gives spacings the length of the lines will in general be several times
the separation and the Carson-Pollaczek formulae can be used
without significant errors.
cos
12 R 27rcr\sin 2 - ( - ' 1 2 + In - ILK -1 2 2 -) (6) CROSSINGS
sin 2 ocKa^bj a2bl a
J>\ a.2b-J
When two lines cross one another, the formulae of Section 3
cannot be applied as they stand, owing to the infinities which
occur in the original integrand. In addition, at the small
spacings, the heights of the lines cannot be ignored. The treat-
-f E3(ypbl, yqa2l) - E3(ypbl, yqall) + E3(ypa2, yqbl2) ment to be employed will depend to some extent on the individual
characteristics of each case. If the crossing is via a short span
*- E&VPav Wbn) - E3(yPai> Y9bn)]\ almost at right angles to the second circuit, its contribution to
the total impedance may be ignored. If the lines cross at a small
. . . . (10) angle, a reasonable approximation is obtained by using the
in which Carson-Pollaczek formula and a mean spacing.

(7) COMPARISON OF NUMERICAL RESULTS


The impedances obtained from the equations of Section 3
Formulae and tables for computing E3(yp, yq) are given in have been calculated for a number of special cases in order that
Appendix 10.6. It is shown in Appendix 10.4 that eqn. (10) the results may be compared with the established methods of
reduces to eqn. (8) for parallel lines in the limit as a -> 0. calculation.
158 LACEY: THE MUTUAL IMPEDANCE OF EARTH-RETURN CIRCUITS

100 1000 10 000 10 100 1000 10 000


Si'parat lon.m Separation, m
ft)
Fig. 2.—Comparison of formulae.
(a) Lines of length 600 and 400 metres.
(6) Lines of length 2 000 and 1 000 metres.
Carson-Pollaczek.
Carter (omitting .R12).
Equation (8).

In Fig. 2 are given curves which show the mutual impedance 8 000 m
of a pair of parallel lines at varying separations, as calculated 500m 250m-
from the Carson Pollaczek formula, from Carter's formula and
from eqn. (8). The curves relate to a frequency of 50 c/s and 2 000m
an earth resistivity of 5 000 ohm-cm. Two pairs of line lengths
are given, the first pair being 600 and 400 m and the second
2 000 and 1 000 m. It will be seen that at small spacings
eqn. (8) agrees well with the Carson-Pollaczek formula and at , 3 875m
large spacings it agrees with Carter. For medium spacings it '1000m X1125m 250m'
gives the correct connection between the two limiting cases. Fig. 3.—Hypothetical exposure used as example in Section 7.
It is of particular interest to note that the error in the Carson-
Pollaczek formula increases rapidly for the shorter lines; at a
spacing of 1 000 m it is almost 4 times too large. formula taking account of phase angle gives results more than
As a second example we take the hypothetical exposure between 30% too high.
a power line and a communication circuit shown in Fig. 3. (8) ACKNOWLEDGMENTS
The 8 000-m (5-mile) straight length is assumed to be the power The author wishes to acknowledge his indebtedness to Mr.
line. To calculate the impedance by the methods laid down in J. P. Holland for assistance during the preparation of the paper
the Directives of the C.C.I.F. the communication circuit is and for proof reading, and to the British Electricity Authority
divided into segments. The length of each segment is assumed for permission to publish the paper.
to be its projection on the power line and the separation is
assumed to be the geometric mean distance from the power line (9) REFERENCES
of its ends. Proceeding in this way it is found for the complete (1) "Directives concernant la protection des lignes de tele-
exposure that communication contre les actions nuisibles des lignes
Zl2 ••---• 0 109 +./0 0676 electriques industrielles," C.C.I.F., Edition de Roma 1937,
This figure is obtained if account is taken of the phase angle revisee a Oslo 1938.
of the component parts of the total impedance. If the C.C.I.F. (2) CARSON, J. R.: "Wave Propagation in Overhead Wires with
curves are used, the magnitude only of the impedance is obtained Ground Return," Bell System Technical Journal, 1926, 5,
giving p. 539.
Z I 7 = 01335 (3) POLLACZEK, F.: "Ober das Feld einer unendlich langen
wechselstromdurchfiossenen Einfachleitung," Elektrische
The correct figure obtained from eqns. (8) and (11) is Nachrichten-Technik, 1926, 3, p. 339.
Zl2 0 070 • yO-0646 (4) FOSTER, R. M.: "Mutual Impedance of Grounded Wires
lying on the Surface of the Earth," Bell System Technical
|Z 12 | 0095 Journal, 1931, 10, p. 408.
It will be seen that in this case direct application of the simple (5) CARTER, R. O.: "The Mutual Impedance between Short
C.C.I.F. method, intended for general use, gives results more Earth-Return Circuits," Journal I.E.E., 1947, 94, Part I,
than 40% too high, and application of the Carson-Pollaczek p. 275.
LACEY: THE MUTUAL IMPEDANCE OF EARTH-RETURN CIRCUITS 159
(6) SUNDE, E. D.: "Earth Conduction Effects in Transmission r i i Pycosa x cos 2a
Systems" (Van Nostrand, 1949), p. 112. ' ~dv3,dy = —3 2 - I
yr .Y sin a|_ r
(7) "Tables of the Generalised Exponential Integral Functions,"
Annals of the Computation Laboratory of Harvard • -. , (r + x - y cos a\ ~]
(17)
University, 1949, Vol. XXI. - sin2 a log ( . )
\ y sin a / J
1 1 y - x cos a
• (18)
(10) APPENDICES r3 x2 sin2 a r
(10.1) Formulae for Lines at an Angle and Large Spacing
An inspection of the terms appearing in these integrals shows
We require that they may be collected in the form
f
3xy sin2 a cos a^
">5""~ ~ " 7 : ) )xy
AT
where r -••• V X * 2 -f- y2 — 2;cy c o s a )
-r / 4 (a) - + / 5 (a) log ( —. ) . (19)
This integral can be evaluated in terms of elementary functions V x \ ^ sm a /
by successive integration with respect to x and y. The work is Selecting the appropriate coefficients and simplifying, we obtain
straightforward but complicated.
The substitution /,(a) - 1 2 cot2 a

x y cos a = ^ sin a tan 0 cos a


2 4
sin2 a
converts the integration with respect to x (treating y as constant) (20)
2 cos a
into integration of standard forms with respect to 6 and results in 3
sin2 a
3xy sin 2 a cos a and / 5 (a) - 0
/•s r3
Hence we obtain
3 cos a r
3xy sin2 a cos a>
J . - sin 6 cos2
y2 sin 2 ,—
a cos o ~r >'2 sin a
cos a /A,/), 1 — 2 cot2 a cos a / y * !
- -.—.-;-cos 0)dddy
y2 sin2 a / = • — • — • + . _ — ( • •
2
-f

r sin <x\xr yr
3 cos a , . 3 cos3 ^ 1 r cos a
-.----(sin i sin ^)
y2 sin 2 a sin a r xy sin2 a
(21)
cos a . Q~]
- 2, . — - sin 9 Denoting this function by (f>(x,y) we have for the definite
y sin2 a J integral
^ sin2 a , 2 cos a y cos a)
r3 /7 2 sin2 a
= -L -i- - I - A _ J . _ COSCC.(32 + / l l r
12 _ '21 \
cos a ( x - ^ c o s ^ O 3 / % i ' rii ri2 r^
2
sin a U 2 6 2 a1bl axb2 a2b{)
yW sin2 a J ' " * . . . " . (22)
The component terms in this integrand can all be evaluated (10.2) General Formula for Parallel Lines
separately. Omitting constants and treating x as constant while
integrating with respect to y, we have
We require / =

where r = V(*2 + y2 - 2xy -j r2)


1 Put
ry2
C
x
T
sm
(14) so that r =
y
Then
f 1 , 1, fr + x — y cos a\
_. c/y = - - log ( ~ •) . . (15) 2
]ry x \ y sin a / + '•o) [ + 'o~)

1 1 y0 Si 2 a !)
2 3
/-
dv -- x*- sin2- l
a[
"r ~ where D is the transform of the rectangle bounded by .v -- alt
x = a2,y = bx,y = b2, and
, A:(COS3 a — 5 sin2 a cos a) x2 sin2 a
r yr 1 - 1
• (24)
,i • •> i //* + x — y cos a\ "1
(16) 1 1
4- 3 cos a sin2 a log ( ~ .) Tixliy
\ y sin a / J
160 LACEY: THE MUTUAL IMPEDANCE OF EARTH-RETURN CIRCUITS
It follows from the transformations that D is a rectangular which on rearranging gives
domain bounded by the lines T _

- yrl2) exp ( - yr2l) _ exp ( - yrn) _ exp (-- yr21)


(25) r
i) + £--2a2 Tj- £---2b2j 12 21

- 1 *(s f+S
If we now suppose for convenience that
bl<al<b2< a2
"I" H . . . . (29)
we obtain for D the domain shown in Fig. 4. The full expression for Zl2 may then be written as:

] * (30)
We have now to arrange the formula in a form convenient
for computation using as far as possible functions for which
published tables are available.
Since rn = -\/(^u + ro) w ^ tn s ^ ^ r expressions for r22, r12,
r2], we may write:
e x p ( - yrn)

11

. . . . (31)
and . . . . (32)
'0 '0
Also by splitting into real and imaginary parts it may be
Fig. 4.—Transform of domain of integration—parallel-line formula, shown that

Dividing the integral into three parts we then have V(€2 + ro)

V2
4- i f - <r
where ^ "

. . . . (26) Ec(x,y)=
.(34)
If we now perform the integration with respect to r/ and f^exp \— \/(x2 — f2)! sin A '(.r2 + /2)
observe that Es(x,y) = ' dt

Tables of Ec and £, are available.7


..^ +gfr The expression for Z 12 may now be put in the form
(27) z _ o Wo
12 12
"

we may write
exp [
where
and
exp [•- yy'(f2 + / o)]T : '
''Sa

exp [ (37)

. . . . (28)
LACEY: THE MUTUAL IMPEDANCE OF EARTH-RETURN CIRCUITS 161
(10.3) Formula for Lines at an Angle

ex
(-—T^) P ( - yr)dxdy

where r = VC*2 + y2 - 2xy cos a)


Put £ = /• = V(* 2 + y2 - 2JO> cos a)
77 = j / / x

Then r ,
(38)
-a,sin«-| \ J i r- 4
—- 4-
where D is the transform of the rectangle bounded by x — ax,
x = a2, y — bu y = b2 and Fig. 5.—Transform of domain of integration—oblique-lines formula,

Now exp ( - • • (46)


2 (39) J £2 £
x
The integration with respect to £ may therefore be performed
and the limits may be inserted. On rearranging this gives
Now + 772 — 277 cos a)
6,/ai r -1

and so i exp
L
- £_V(l + T?2 27? cos a)
J.
1 + 772 — 277 COS a
J
6/« 2 2 - 2 7 7 cos a)3/2 "
Thus
!>2/a2 f
1 •— 1 exp T?2 - 277 cos a)
- drfdg . (40)
^ 2r) cos oc ' y L V
, 62(1 + 772 - 2,7 cos a)3/2
We have now to consider the form of the domain D.
The transformations convert the line x = a (a > 0) into the exp [— axy\/(l + 772 277 cos a)]
curve 2
$ = flV(l + r)2 - 2r) cos a) . . . . (41) '- 277 cos a)V2
"W«i
which is that part of the hyperbola
exp [— a2y\/{\ + 772 277 cos a)]
£2 _ (77 - cos a) 2 _ '*» (47)
2 2 (42) 772-277 cos a)3/2
a sin a sin2 a
which lies in the 1st and 4th quadrants. By making the substitution
The line y = b (b > 0) transforms into
1
-- - yb(u cos a) (48)
£77 = fr\/(l + 772 — 277 cos a) (43) ^ u
which may be written in the first two of the integrals and
(44) t — 0(77 cos a) (49)
77 77
in the second two, it is then found that
.- fx(a\ - bz cos a) pv( u i' b\ cos :
This curve is asymptotic to 77 = 0 as £ ->. 00 and to £ — 2
' = y [* 2 J f<f, f>2)dt - bx I / ( / , b{)dt
as 77 -> 00. It touches £ = b sin a when 77 = JSOt)
cos a
If it is assumed for convenience that
f{t,a2)dt \. (50)
ax<bx<a2< b2 - -y j ^ l — <J 2 COSQ

the domain D is as shown in Fig. 5. + v2p2 sin2 a)l


where
The integral may now be expressed as the sum of three repeated
integrals by dividing D into three parts. We have
The function £3(y/?, yg), in which y ^ y 0 e x P C/77"/4) a n c i yo»
a 2 VU+7)2-27] cos a) r,2—2/)cosa) and 5 are real, may now be defined by
7|2—2ri COS a)
+ U? ;2 -2r, cos a) f exp [— y\/(/2 + /?2
7)
• (51)

It may then be shown that


1 + 772 — 277 cos a
. . . . (45)
162 LACEY: THE MUTUAL IMPEDANCE OF EARTH-RETURN CIRCUITS
in which E3r, E3i are real functions of real variables given by Now from the way in which the limit is approached it can be
seen that ?•, = y{a2 — b2 cos a) is independent of a.
T (,• v\- v f exp [ \(x2 -f / 2 )] „ , ,, ,
Also Lim v{ = — Lim v2, Lim u{ — Lim u2 •(60)
a->0
(53)
and E3(u, v) is an odd function of v.
V *» + It follows that
T. D£3 _.
L i m - - - = - Lim
In terms of E3 it is now possible to write
and that
7
" v'2 L = Lim («, - «2) + —3 2
-f- E3(yphl, yqu2l) E3(yphl, ygall)

= Lim y(b2 - a2)--^ + ya2 sin a--^


(54)
and the full expression for Z 1 2 as

ZP flP H * r C 0 S , a (-7*12- -f- -- 21 ' n - /"22-) - /1(55) where S 22 = a2 - b2 in the limiting position of the lines.
27TCTl_sin- oc\alb2 a2bl albl a2b2/ J Straightforward differentiation and rearrangement now gives
gi

(10.4) Consistency of Formulae


It will be shown that the formula for lines at an angle agrees
with the parallel line formula in the limit as the angle tends to V2 d_ f "exp [ - yy/(r\ ±J2))dt , ^2 2 V 2 exP (^ y/22)
zero, and that the parallel-line formula agrees with Carson and r 2
yyo o^oio "VH+'t ) ~" YYo' ~r\2~~
Pollaczek as the length of one line tends to infinity.
Starting from the formula for lines at an angle, the way in . . . . (62)
which they shall tend to become parallel must be defined. Let it
be agreed that in Fig. 6 Av A2 and B2 shall remain fixed and and -*— E (yr yS ) == r ^ 1 QXP
Sl-^22}
that B{ shall move along a fixed line perpendicular to AyA2 to B[. i(yS22) 3 °' 22 yoy r\2
Hence, since S\2 + r% = r\2,

t2 r
S/^ V(^ + ) YO 22
. . . . (64)
Similarly, corresponding expressions are obtained for the other
three remaining pairs of E3 terms.
For the remaining terms in Z ] 2 it is only necessary to note that
B1, «j sin a, a2 sin a, bx sin a, and b2 sin a all tend to rQ.
Finally collecting terms the expression for the limit of Z 1 2 is
Fig. 6. —Notation for proof of consistency of formulae.

Now consider the limit of two £"3 terms. 0) —

LetL f e x p C - yru) exp ( - yr22)


£,[y/) 2 sin a, y(a2 b2cos a)] -- £3[ycr2sin a,; a 2 cos a)] L /-j, " " r 22
Lim
sin a
. (56) exp (•- yrl2) _ e x p j - yr 2 J )l
r
21
Since numerator and denominator both tend to zero,
1 D exp[^
s n a
L •- Lim [^[y^ ' > y("-> ^2 c o s a ) ]
-f E3[ya2 sin a, y(6, — a2 cos a)]} (65)
(57)
which is the formula previously obtained for the parallel line
Putting y/>2 sin a = Wj, ya, sin a case, eqns. (29) and (30).
This can be shown to be a generalization of the Carson-
y(a 2 •- b2 cos a) — /•,, y(b2 — o 2 cos a) -•= ?-2 J Pollaczek formula by taking one line to be of unit length and
letting the length of the other tend to infinity.
then L Lim [£ 3 (//,, rj) • E3(u,, i\) Then
r21 — r2 2 -
Lim (66)
Sl2 S{ '21
f
22
LACEY: THE MUTUAL IMPEDANCE OF EARTH-RETURN CIRCUITS 163
1
taken in interpolating this table as second- and higher-order
and all the S and r quantities tend to infinity. It follows that
differences are not negligible. If the method is to be used for
z K r routine work, the table requires to be extended by tabulating at
n (a = o) -> 2^A ~i + — ^r o(y o) • • (67) closer intervals of both x and y. It would be possible without
excessive labour to prepare tables in which interpolation in the
in which use has been made of the integral x-direction was not justified by the accuracy of the initial data
and interpolation in the ^-direction could be performed at sight.
f A-2)] Such tables would not have to be extended to cover the largest
~—dt = K0(yr0) . . (68) values of* and y that could occur in practice as a series expansion
can be obtained which converges rapidly for large values.
By making the substitution \Q = yQr0 — ro\/(a)[MQa) and using It can be shown that
the formula
(69) exp (— A) cis ( - A)
we obtain
+ y\/2[ker' x^/2 -f j kei' x\/2]
z (ker kei
i 2 U - o) -> — 1-2 + ' Wo + J ' Wo) r' x\/2 ; j bei' xy/2]Ej[\\/2 exp O'TT/4)]
/kei AQ y ker Ao\
, — •( . J . (70) + exp ( A) cis ( A)~[ber x\2 + j bei xy/2]
TIAQ TT \ AQ J

which is Pollaczek's formula for medium spacings in rationalized - 2 exp ( - A) cis ( - A)>'"2( I)" 4 lSn
M.K.S. units. )2
XA i

(10.5) Formulae and Tables for *\*(x, y) XA i (74)


The function tp(x, y) is defined by the equations
in which+ 2£",(«)
exp (is A)
thecisclassical
Q - A)exponential integral function
I/J(X, y) = *fjr(x, y) , y) defined by

(71) =(yAV-t
2 2 • (75)
_ x exp [- V(* + sin ±y_ )
(2u

Extensive tables of where —-— is the largest integer

and A - v V + y2)
and The majority of terms are negligibly small for the ranges in
^ ' which the series would be used.
E(X exp sin For y > 3 only the first two terms of the series of Sn and crH
need be included. These may be expressed in finite terms by
I/A
have been published by the Harvard Computation Laboratory.? >A
Their differentials with respect to x can be obtained by finite
difference methods from these tables. The appropriate formula is jK*--(y_
(76)
1 2A2
f'n =
2a A- j '
t-log ( yA
n(3n - 4)/ 0 - n)f{ + nV2]
1 r For y > 9 and 3 only the first and third terms are
-i + 3/ 0 _3/ 1 +/ 2 ] . (73) important and

in which / _ , , / 0 , / „ f2 are successive tabular values, a is the (77)


interval between them, and n is the fraction of the interval
between To and/, at which the differential is required. When y < 3 and x > 6 the formulae for large spacings may
A short table of i/rr and ifjt is given in Table 1. Care must be be used with negligible errors, i.e. we may take ifj(x, y) — Xjx.
164 LACEY: THE MUTUAL IMPEDANCE OF EARTH-RETURN CIRCUITS
Table 1
</U*. y)
X
y
0 1 2 3 4 5 6 7 8 9

0 0 0-8012 10563 1-0493 10120 0-9981 0-9976 0-9993 10001 10001


1 0 0-9384 1-1673 1-1129 10462 1-0183 10112 1-0094 1-0078 10062
2 0 1-2766 1-4592 1-2869 1-1426 1-0765 10510 10390 10308 1-0245
3 0 1-7113 1-8564 1-5354 1-2867 1-1666 11141 10865 10679 1-0542

4 0 2-1860 2-3053 1-8278 1-4638 1-2818 1-1969 1-1498 11179 1-0945


5 0 2-6774 2-7793 2-1455 1-6632 1-4158 1-2955 1-2264 11791 1-1442
6 0 3-1769 3-2665 2-4785 1-8778 1-5640 1-4068 1-3141 1-2498 1-2021

0 3-6809 3-7617 2-8214 2-1030 1-7228 1-5279 1-4107 1-3285 1-2672


8 0 4-1878 4-2621 3-1712 2-3361 1-8894 1-6567 1-5146 1-4139 1-3383
9 0 4-6967 4-7661 3-5259 2-5747 2 0621 1-7916 1-6244 1-5048 1-4146

X
y
0 2 3 4 5 6 7 8 9

0 0 0-3096 01231 0-0070 -00139 -00065 -00007 + 00006 + 00003 + 00001


1 0 0-4922 01789 00161 -00146 -0-0074 -00009 + 00006 + 00004 + 00001
2 0 0-8749 0-3094 0 0373 -00171 -00101 -00016 + 00006 + 0-0004 + 00001
3 0 1-3044 0-4636 00619 -00217 -00140 -00026 + 00007 + 00006 + 00001
4 0 1-7401 0-6202 0 0839 -00279 -0-0185 -00036 + 00008 + 0-0007 + 0-0002
5 0 2-1765 0-7762 0-1087 -00347 -00240 -0-0047 + 0-0009 + 00009 + 00002
6 0 2-6119 0-9317 01265 -0-0417 -00279 -00056 + 00011 + 00011 + 00003

7 0 3-0472 10867 0-1475 -00486 - 0-0325 -00066 + 00013 + 00013 + 00003


8 0 3-4825 1-2419 01685 -00556 -00372 -0-0075 + 00014 + 00014 + 00004
9 0 3-9178 1-3972 0-1896 -00626 -00418 -00084 + 00016 + 00016 + 00004

X
y
0 0-2 0-4 0-6 0-8 10

0 0 0-1976 0-3826 0-5471 0-6870 0-8012


0-2 0 01999 0-3865 0-5520 0-6925 0-8071
0-4 0 0-2064 0-3978 0-5666 0-7090 0-8244
0-6 0 0-2167 0-4160 0-5901 0-7357 0-8526
0-8 0 0-2302 0-4401 0-6216 0-7718 0-8910
10 0 0-2464 0-4694 0-6602 0-8160 0-9384

X
y
0 0-2 0-4 0-6 0-8 10

0 0 01627 0-2610 0-3099 0-3223 0-3096


0-2 0 0-1787 0-2755 0-3222 0-3325 0-3179
0-4 0 0-2182 0-3151 0-3571 0-3621 0-3423
0-6 0 0-2708 0-3731 0-4106 0-4084 0-3810
0-8 0 0-3313 0-4435 0-4781 0-4682 0-4318
10 0 0-3967 0-5224 0-5557 0-5383 0-4922
LACEY: THE MUTUAL IMPEDANCE OF EARTH-RETURN CIRCUITS 165

Table 2

y
0-5 10 1-5 20 25 30 3-5 40

0 ooooo 0-0000 0-0000 OOOOO 0-0000 0-0000 0-0000 0-0000


01 0-2077 00197 0-0007 -00014 -0-0011 -00005 -00002 -0-0000
0-2 0-3879 0 0383 00013 -00028 -00021 -00011 -00005 -0-0001
0-3 0-5267 00551 00018 -00042 -00031 -0-0016 -0-0007 -00002
0-4 0-6250 0 0693 0 0020 -00056 -00041 -00022 -00009 -00003
0-5 0-6910 0 0808 0 0021 -00070 -00051 - 00027 -0-0011 -00004
0-6 0-7339 0 0898 00019 -00083 -00061 - 0 0032 -00013 -00004
0-7 0-7614 0 0965 00014 -00097 -0-0070 -0-0036 -0-0015 -00005
0-8 0-7787 01011 00008 -00110 -00079 -00041 -00017 -00006
0-9 0-7896 0 1042 ooooo -00122 -00088 -00045 -00019 -00006
10 0-7962 0 1060 -00009 -00134 -00096 -00050 -00021 -00007
1-2 0-8024 0 1072 -00030 -00157 -00110 -00057 -00024 -00008
1-4 0-8042 01067 -00051 -00176 -00123 -00064 -00027 -00008
1-6 0-8042 0 1054 - 00070 -00193 -00134 -00069 -00029 -00009
1-8 0-8036 0 1040 -00087 -00207 -00142 -00073 -00031 -00009
20 0-8029 01027 -00102 -00218 -00149 -00077 -00032 -00010
2-5 0-8015 01005 -00125 -00237 -00161 -00083 -00034 -00010
30 0-8009 0 0994 -00136 -00245 -00166 -00085 -00035 -00010
3-5 0-8006 0 0990 -00140 - 0 0248 — 00168 -00086 — 00035 — 00009
40 0-8005 0 0988 -00141 - 0 0249 -00168 -00085 -00034 - 00009
4-5 0-8005 0 0988 -00142 - 0 0249 -00168 -00085 -00034 -00009
50 0-8005 0 0988 -00141 - 0 0249 -00168 -00085 -00034 -00009

y
0-5 10 1-5 20 2-5 30 3-5 40

0 0-0000 0-0000 0-0000 ooooo OOOOO 0-0000 0-0000 0-0000


0-1 01144 0 0308 0 0099 00031 00008 00001 -00001 -00001
0-2 0-2181 0 0606 00195 0 0061 00016 00002 -00002 -00002
0-3 0-3048 0 0885 0 0288 0 0090 0 0023 00002 -00003 -00003
0-4 0-3734 01139 0 0376 00118 0 0030 0 0003 -00003 -00003
0-5 0-4258 01364 0 0458 00145 0 0037 00003 -00004 -00004
0-6 0-4651 01559 00532 00170 0 0043 00004 -00005 -00005
0-7 0-4944 01726 0 0599 00192 00049 00004 -00006 -00006
0-8 0-5162 01865 0 0658 00213 0 0054 00004 -00007 -00007
0-9 0-5325 0-1980 00710 00231 0 0059 00005 -00008 -00008
10 0-5447 0-2075 0 0754 0 0247 0 0063 00005 -00009 - 0 0008
1-2 0-5607 0-2213 0 0824 0-0273 0 0070 00004 -00011 -00010
1-4 0-5699 0-2301 0 0872 0 0292 0 0074 00003 -0-0012 -00011
1-6 0-5751 0-2356 0 0905 0 0304 0 0076 00002 -00014 - 0-0013
1-8 0-5782 0-2390 0 0925 00312 0 0077 00001 -10 0016 -00014
20 0-5799 0-2410 0 0938 00317 0 0077 -00001 -00017 00015
2-5 0-5816 0-2430 0 0950 0 0320 0 0075 -0-0005 -0-0021 -0-0017
30 0-5818 0-2432 0 0950 00318 0 0072 -00008 -00023 -00018
3-5 0-5818 0-2431 0 0948 00315 0 0069 -00010 -00024 -00019
40 0-5817 0-2430 0 0947 00314 0 0068 -00011 -00025 -00019
4-5 0-5817 0-2429 0 0946 00313 0 0067 -00011 -00025 -00019
50 0-5817 0-2429 0 0945 00312 0 0067 -0-0011 -00025 -00019
166 LACEY: THE MUTUAL IMPEDANCE OF EARTH-RETURN CIRCUITS

Table 3
<c

ker tdt
Jo

X 0 01 0-2 0-3 0-4 0-5 0-6 0-7 0-8 0-9

0 00000 0-3419 0-5456 0-6977 0-8170 0-9125 0-9896 10521 1•1027 1•1433

1 1-1757 1-2010 1-2206 1-2351 1-2455 1-2523 1-2563 1-2577 1-2571 1-2549
2 1-2513 1-2466 1-2411 1•2350 1-2285 1-2216 1-2187 1-2075 1-2004 1-1934
1-1866 11800 1-1737 11677 11619 1-1565 1-1514 1•1466 1-1422 1-1381
3
4 1-1343 1-1308 1-1277 1-1248 1-1222 1-1199 1-1178 11160 1-1144 11129
5 11117 11106 1•1097 1•1089 1•1083 1•1078 1•1074 1-1071 1•1068 11067
6 1•1066 1-1065 1•1065 1•1066 1•1067 1•1068 1•1069 11071 1•1073 1•1074

7 1•1076 1-1078 11080 1•1082 11084 1-1086 1-1088 11090 1-1091 1•1093
8 1•1094 11096 11097 1•1098 11100 1-1101 1-1102 11103 11103 11104
y 11105 11105 11106 11106 11107 11107 11108 11108 11108 11108

kei tdt

X 0 01 0-2 0-3 0-4 0-5 0-6 0-7 0-8 0-9

0 00000 0 0782 01550 0-2296 0-3015 0-3703 0-4358 0-4978 0-5562 0-6110

1 0-6623 0-7101 0-7544 0-7953 0-8331 0-8677 0-8994 0-9283 0-9545 0-9782
•y
0-9996 1-0187 1-0358 1-0511 10645 10764 10867 1-0957 1•1034 11100
3 1-1156 11203 1-1241 1-1272 11297 1-1316 11329 11338 11343 1-1345

4 1-1344 1-1341 1-1336 1-1329 11321 11311 11301 11291 1-1280 11268
5 1-1257 11246 1-1235 1-1224 11214 11204 11195 1-1186 1-1177 11169
6 11162 11155 1-1148 1-1142 1-1137 1-1132 1-1127 1-1123 11120 11117

7 11114 11111 11109 1-1107 11105 1-1104 11103 11102 11101 11101
8 11100 11100 11100 11100 11100 1•1100 1-1100 11100 11100 11100
9 11101 11101 11101 11102 11102 1-1103 11103 11103 11104 11104

In computing from the series the following are useful in For values of A beyond the range of the tables the asymptotic
evaluating £,[A\/2 exp (/JT/4)]. series

exp ( - /)
E,[\\/2 exp O'TT/4)] dt
may be used.
xd j) For small values of x and y the following series can be used.
exp (— t) cos t exp (— /) sin
dt
*7.
c (78)
exp ( - /) cos /
(80)
= x[l - A + fa _ yl) _ JL(3x4 _ yA)

These series give five figure accuracy for x < 0-2, y < 0-2.
where Ec and Es are the functions, previously defined, which are For x < 0-1, y < 0-1, the range in which they will normally be
tabulated in Ref. 7, and c = 0-577216 is Euler's constant. used, some terms become negligible.
LACEY: THE MUTUAL IMPEDANCE OF EARTH-RETURN CIRCUITS 167

(10.6) Formulae and Tables for E3(yp, yq) . y xy(X 4)


tan ^ - 12
The function E3(yp, yq) is defined by the equations
5.v4
3 360

(x4 (84)
exp cos 630 V
(81) .Y^A /ll 13 ,
2 520V16 24 '
'exp [•• V( sin V(> 2 + X2)
(X6 .:- V4V2 v 2y4
22 680

A table of E3r and E3i is presented in Table 2. It may be E3i(x, y) ^ arc tan xy 10
interpolated without difficulty in the ^-direction, but in the 30
jc-direction second- and higher-order differences are not neg-
ligible. For values of the argument beyond the range of the ~~ 304) 1 O g ( .v
table the following may be used with negligible error.
For y > 4 and any x, the integrals may be taken to infinity. _^ (85)
Then 636
xy
Eir(x, y) jE3i(x, y) x'y'y!
22 680
^A / 1 3 7 6
kei udu
400V384 A % ***
The number of terms given above ensures five-figure accuracy
for ^ < 1, y < 1, and could be used in preparing tables. For
kei -- V 2 udu (82) y a l u e s o f x a n d ^ le§s t h a n ^ t h e m a j o r i t y o f t e r m s g i y e n a b o y e

are negligible and the computation is particularly simple.


Since negative arguments can occur in E,
J in a practical calcula-
Values of the integrals in these expressions are given in Table 3. t i o n it snoui(j ^ note£j ^
For x > 4, and y < 4
E3r(x, y) — E3i(x, y) = 0 . . . . (83) (86)
and
When both x and y are small interpolation in a table of E3 is It is also of importance to observe that, although E3 -•» oo as
difficult owing to the rapid fluctuation of the differences. The x -> 0, whenever a zero argument occurs it appears in two £"3
following series facilitate computation in this range. terms and the limit of their difference can be shown to be zero.

You might also like