Professional Documents
Culture Documents
8 August 1981
THE COMPLEX GROUND RETURN PLANE
A SIMPLIFIED MODEL FOR HOMOGENEOUS AND MULTI-LAYER EARTH RETURN
Abstract - For modelling current return in homo- are numerically evaluated. The method of the "complex
geneous ground, the paper introduces the concept of an ground return plane" is, inaddition, extended for the
ideal (superconducting) current return plane placed be- modelling of multi-layer ground return.
low the ground surface at a complex distance p equal to
the complex penetration depth for plane waves. This THE COMPLEX GROUND RETURN PLANE
"complex"' plane appears as a mirroring surface, so that
conductor images can be used to derive very simple Fundamental Equations
formulae for self and mutual impedances under ground
return conditions. Such equations, without proofs, were The ideal conductor/ground return (self and mutual)
originally proposed by Dubanton and published by Gary.1 impedances, according to Carsonand in a formadapted to
In this paper, plausibility arguments serve to initially our purpose, are:
justify the procedure, then the equations are analyti-
cally related to those of Carson and, finally, the h + V
errors, which in most cases are less than a few percent, Zs =
5
jw Zs
- _2
2iT
n--+
r
w
uJPo
iT 5 =11
J
X +
j -dX
+j
( "t)
accepted. It is valid for homogeneous earth and for o
frequencies low enough so that the capacitive displace-
ment currents in ground can be neglected and the wave- and
length is sufficiently large compared to transversal co
geometrical dimensions.
The calculation of line impedances according to .-(hk+hk) X
Jm = P=P + jQ = -j e ~~ cosXd k x dX (2")
Carson is based on equations which contain infinite
integrals with complex arguments. For their evaluation
m
+jQ~~O
m
x+v7 T r 1
Z
s
= ii 2Tr°-O n 2(h +
r
p) (3)
+dk2k
(hk+hg+2p)
2r Qn - (4)
Z =j&
m 2 2
WM 222-9A paper recommended aiid approved by the
IEEE Transmission & Distribution Committee of the
( hk-hI) + dkQ
IEEE Power Engineering Society for presentation at
the IEEE PES Winter Meeting, Atlanta, Georgia, These are the Dubantonm equations. They provide simple
February 1-6, 1981. M.anuscript submitted August 22, and remarkably accurate substitutes to Carson's equa-
1980; made available for printing December 8, 1980. tions (1) and (2) over the whole range of frequencies
1981 IEEE
3687
for which Carson's equations are valid. at any depth, and an ascending wire. For sucha contour:
It should be mentioned that the convenience of
equations (3) and (4) results from the fact that they f(Edz) = -jwf (5)
can be evaluated by electronic hand held calculators,
while the evaluation of Carson's equations requires a where 4 is the flux crossing the loop. The line-inte-
computer program or reading of numerical values from gral in (5) is
graphs.
The next section contains a plausibility argument
for the validity of these equations.
f(EdQ) = -V + Z cI + IJ (6)
where Z is the impedance (per unit length) of the con-
Plausibility Argumentation ductor, and J the current density at the depth of the
return line of the contour.
The following concepts will be introduced in what From (5) and (6):
follows:
.Equivalent return distance V = Z c I + ar1 J + jw4 (7)
*Complex depth
-Complex ground return plane This equation shows the voltage components of V. If we
The equivalent return distance is defined with the aid
of Fig. 1. The conductor is assumed to be a plane at put Zc = 0, we have
conductor plane conductor plane V = a1 J + jw4 (7')
+
Z V =
= (8)
(a) (b)
Its calculation requires the determination of the flux
Fig. 1 To the definition of equivalent return produced by I.
Now we recall the definition of equivalent return
distance
distance (see Fig. lb). Let us assume h = 0. Then we
obtain the complex depth p as the distance between the
two planes of Fig. lb for which the flux between the
height h above ground. The magnetic field is parallel
to the ground. The impedance of the conductor/ground planes equals the total flux in earth, from its surface
to infinity.
return loop is Z = V/I, as shown in Fig. la. In Fig.lb
the earth is replaced by a perfectly conductive plane We will calculate the complex depth p for three
cases, all with p =
at distance D from the plane conductor. The equivalent
return distance is defined by the requirement that the
- Homogeneous earth
impedance Z of the conductor/return plane loop of Fig. lb
- Multi-layer earth
be the same as of the loop in Fig. la.
- Continuously stratified earth (here, of
course, strata do not have the customary
Let D be the equivalent return distance. It turns
out to be a complex length since the inductance of the significance of discrete layers)
loop has to account for the resistance of the earth re-
turn. Therefore p = D h represents the complex depth
-
Homogeneous and Multi-Layer Earth
of the ground return plane below the earth surface.
Because p is complex, this plane will be calied the com- If I denotes the current per unit width (perpendi-
cular in Fig. 1 to the plane of the paper) then
plex ground return plane. For the plane field of our H = H(O) = I, where H(x) is the magnetic field inten-
discussion the complex plane replaces effectively, and Q
without loss of accuracy the actual ground return path. sity at depth x. The pertinent differential equations
are:
The complex ground return plane is significant be-
cause the whole current returns through it. It is re-
dE
tained for the case of actual thin conductors, to
dx
-jyp
0
H
(9 1)
construct image conductors. This isa heuristic approach
to extend theusability of the concept of complex ground
return plane for conditions different from those for
dH
dx (9")
which it has been defined. It is assumed that a plane
which serves as an ideal current return path is, by this
and their solution is, for a layer numbered by K (see
fact, an effective mirroring surface, which can be used
even in the case of real conductors. It appears thus Fig. 2):
plausible that equations such as (3) and (4) will be
good substitutes for Carson's formulae, (1) and (2).
HEK7 rC BKDK Hr 1 (10)
Calculation of Complex Depth
-1
For this latter case (14) yields
(11) x
H = H e P (16)
o:
Co
and the total flux in the earth is
K =-- KK 2
=
K
sinh
PK co
where f = P H dx =P
VIH0 (17)
1
PK 12')
The complex depth p is defined by means of Fig. lb as
resulting in the same flux 4. Consequently, for homo-
geneous earth and =1=po p-is the complex penetration
depth
12") pe h1
(18)
related to the real penetrat ion depth
6 = 1 (18')
= e PK (122f") by VI 'f p
or, eliminating Hn :
EKHK~~~~~~~
n+1 no no
(21)
°
n+l ~n+1 Cn
Fig. 2 Notations for layer no. K since H = I, E can be calculated from (21) and all
EK, HK will be obtained from (20). Consequently, the
Note, that solution (10) is analogous to that of trans- flux in each layer K is obtained by integrating (9'):
mission lines connected in a chain.
If there are n + 1 layers, in the last one (ex-
AK
tending to infinity) there will be no backward wave.
Therefore
E
k = PO Hdx = (EKi - ER) (22)
n
(5 0)
H n+1 (15)
n
3689
The total flux is from (3) and (4) that the impedances will be (approx-
imately) the same.
f = E¢K (23)
Complex Inductance for
This permits to calculate the complex depth of equation Homogeneous Earth and } $ po
(17):
K For ' ipo the concept of complex plane can still
P p H (24) be used for calculating the conductor/ground return
0 0 impedances, but the pertinent equations will be differ-
ent from (3) and (4). Instead complex inductances
Continuously Stratified Earth
L = R- J (28')
Quite naturally, the multi-layer case can be re- s 7r s
duced by a limiting process to the continuous problem
formulation. So, for instance, equation (10) reproduces L
m
= kEIt Jm (28")
(9), which is available in the first place:
can be calculated for equations (1') and (2'). Indeed,
it will be shown later (see equations (40) and (47))
d E 0 -jwpo EF' (9) that
dx H -a 0 H_
The integration of (9), where a is a function of j
5 Sf 1k
2 h (29')
x, is not possible in closed analytical form 1912. But
the numerical integration is straight forward, for
specified starting values: E(o) = Eo, H(O) = Ho. Since,
however, Eo is not known, one has to use several trial
values until the following objective is satisfied: J a J J kn
i k9n(hkh+2p)2+dk2
at a sufficiently large value of x, denoted by x.,, the
m 2 (h+h2 +d2 Q
(291")
backward wave
=vrjwtc
1
must vanish. Then, from (9'): (30)
I
x
01
(2)
(3)
-equivalent
homogeneous x
earth: Heq-Hoe P
matter what is the shape of H(x), as long as the centre Equation (31) is valid for homogeneous earth and
of gravity is given by the same depth p, itis expected any -p (and so is the correspondent expression for Zs).
3690
ANALYTICAL DERIVATION Indeed, using (19), equation (3) becomes, for |p|>>h:
,k
1Im(w)
_1
V
Self Impedance
We intend to derive equation (3) from (1'). First 1/117
we substitute (30) into (1"):
I -2hX
J= 1 Je dX (33)
s + /X2+ /p2
Fig. 5 To the integration of (35)
where the limits indicate integration along the real
axis. Next we introduce a new dimensionless complex
integration variable w, using
w
X =_- (34')
p
and
h
q=-
p
(34")
We obtain
-2qw
J (q) = Je dw (35)
s
J
I
w + VW'+-
7
(L)
where (L) is the line, at angle -450, shown in Fig. 5
Since the integrand in (35) has no poles in the half-
plane at the right of the dashed line and is zero
around the arc LCL' at infinity, the integration can Fig. 6 To substitution (37)
be conveniently performed -along the real axis L'. It
is convenient then to consider the derivative of (35)
with respect to q:
Zs T27° r 2¶( 4 r
J'(q)
00
f ~-2qw
J (j w2w=w_+W dw (36) or
6
Jm ( q) - 1 J-9,n
2 (46)
/
4
0
(hk+ht+2p)2 + d 2
(47) - 2
m 2n h )2 + dk
- 4 --
Expression (47) can be used for any p as the correction -3 -2 -1 gOo
log1
term in (2'). If p = po then (2') yields precisely the -I
approximate expression (4) which we wanted to derive. lHz lOOHz lOkHz f IMHz
ERROR EVALUATION
Expressions (1) and (3) for the self impedance, Fig. 7 Relative error in the absolute value of the
and expressions (2) and (4) for the mutual impedance total impedance, including fluxes in air
of an ideal conductor/ground return loop have been (p=1OOQm, hk=hk=20m)
3692
Manuscript received February 23, 1981. The most important advantage of equation (50) is perhaps that it
can be easily integrated numerically for obtaining ZO, even in the case
of continuous stratification, starting at a large enough value of x. Then
too, (51) gives directly the value of the complex equivalent depth.
A. Semlyen: I would like to thank Dr. El-Hawary for his comments on Finally, I would also like to thank Dr. El-Hawary for the informa-
our paper. We agree that for the calculation of the complex equivalent tion related to a procedure similar to ours in exploration geophysics.
depth, in the case of a multi-layer earth representation, a simpler ap-
proach could be developed. This could be achieved (according to Dr. Manuscript received April 13. 1981.