Professional Documents
Culture Documents
Lecture05 PDF
Lecture05 PDF
P
y
h d
D
x
-h
The earth plane (at y 0 ) is at zero potential. For determining the electric
field in the space above the plane ( y 0 ) we replace the plane with an image
line charge.
equipotentials
lines of force
Let:
By Gauss Law, the electric field vector at the radial distance r from a line
charge in free space is:
E r (5.3)
2 0 r
a
Vab E dl (5.4)
b
d dr
V (5.5)
h 2 0 r
h
d
h V=0
Using superposition, we can add the potential due to the negative image charge
to obtain the potential at point P with respect to the earth plane:
d dr D dr
V
h 2 0 r h 2 0 r
ln d ln h ln D ln h
2 0
D
ln (5.6)
2 0 d
1 1
Ex sin sin
2 0 d D
1 1
Ey cos cos (5.7)
2 0 d D
Hence:
E x x, y
x x
2
2 0 x y h 2 x 2 y h 2
yh yh
E y x, y 2
2 0 x y h 2 x 2 y h 2
(5.8)
At y 0 :
E x x ,0 0
h
E y x ,0
0 x 2 h 2 (5.9)
plane:
h
s 0 E y x ,0 2 2
Cm -2 (5.10)
x h
s dx 2 s dx
o
2 hdx
o x 2 h 2
- Cm -1 (5.11)
Thus the actual induced surface charge per unit length ( ) equals the
fictitious image charge per unit length, and exactly balances the inducing line
charge per unit length .
It can be shown that the equipotential surfaces near the line charge in (a) above
are cylinders, and when the radius of the cylinder is much smaller than the
distance h, the axis of the equipotential cylinder very nearly coincides with the
line charge.
Hence the line charge model is valid for a long cylindrical conductor of radius
R, and when R h the line charge and the conductor may be assumed to be
coaxial. In these cases, applying Eq. (5.6) to a single isolated conductor we
obtain the conductor potential:
2h
Vcond ln (5.12)
2 0 R
2 0
C Fm -1
V 2h (5.13)
ln
R
V P (5.14)
where:
1
P A F1m (5.17)
2 0
where:
Dij
A matrix with elements Aij ln (5.18)
d ij
Conductors
j
dij
i
i'
Images j'
(5.20)
Dii 2hi double height of conductor i
C 2 0 A 1 Fm 1 (5.21)
Example
Solution:
10 11
A11 ln 5.298 A12 ln 2.398
0.05 1
12
A21 A12 A22 ln 5.481
0.05
C 2 8.8542 10 12 A 1 Fm -1
13.09 5.73
pFm-1
5.73 12.66
2
5.73 pF/m
6.93 pF/m
1
7.36 pF/m
Assume the total charge density a is divided equally between the m sub-
conductors (a fair approximation in most cases). Then, from the first row, we
obtain:
a
Va P11 P12 P1m
m
a m D1 j
ln
2 0 m j 1 d1 j
1
m m m
D D1 j
a
1j
a
ln
j 1
ln j 1 (5.23)
2 0 m m
2 0 1
d 1j m
d1 j
m
j 1
j 1
D (5.24)
Va a ln aa
2 0 d aa
where:
1
m m m2
Daa Dij
(5.25)
i 1 j 1
1
m m m2
d aa d ij (5.26)
i 1 j 1
Daa = mutual GMD between the bundle conductor a and its image
Dij = mutual GMD between the bundle conductor i and the image of j
Example
Solution:
1 1
Equivalent radius d aa d11d12 d 21d 22 4 d11d12 2 0.05 1 0.2236 m
1 1
Daa D11D12 D21D22 4 10 112 12 4 10.98 m
2 8.8542 10 12
C Fm -1 14.29 pFm -1
10.98
ln
0.2236
We will assume a simple transmission line with just three conductors labelled
a, b and c. We assume all conductors to be identical, except for their position
in space. Image conductors at a depth equal to the conductor height are
assumed.
b
Conductors c
a
Earth
a'
Images c'
b'
Dab , Dbc , Dca = mutual GMDs between conductor and image of another
conductor
Daa , Dbb , Dcc = mutual GMDs between conductor and its own image
Cond. in pos. a
a b
Cond. in pos. b
b c
Cond. in pos. c
c a
For the first (LH) section of the barrel we obtain the potential coefficients
from Eq. (5.17):
The values in the matrix are rotated for the other two sections. Thus, for the
complete transposed line:
Ps Pm Pm
Pabc Pm Ps Pm (5.29)
Pm Pm Ps
1 1 Dab D D
Pm ln ln bc ln ca
2 0 3 d ab d bc d ca
1 3 Dab Dbc Dca
ln (5.31)
2 0 3 d ab d bc d ca
P0 0 0
P012 H 1Pabc H 0 P1 0 (5.32)
0 0 P2
where:
P0 0 0
P012 H 1Pabc H 0 P1 0 (5.33)
0 0 P2
and:
P1 P2 Ps Pm
P0 Ps 2 Pm (5.34)
P1 P2 Ps Pm
1 3 Daa Dbb Dcc 3 D D D
ab bc ca
ln ln (5.35)
2 0 d aa 3 d d d
ab bc ca
1 3 d ab d bc d ca 3 D D D
ab bc ca
ln ln
2 0 d aa 3 D D D
aa bb cc
2 0
C1 C2
d eq 3 D D D (5.36)
ln ln ab bc ca
d aa 3 D D D
aa bb cc
where:
(5.37)
d eq 3 d ab d bc d ca = equivalent spacing
When the conductor spacing is very small compared to the height from the
3 Dab Dbc Dca
ground, then ln is very small, and the capacitance per unit length
3 Daa Dbb Dcc
approaches:
2 0
C1 C2 (5.38)
d
ln eq
d aa
We could find the zero sequence capacitance per unit length C0 by using
earth of all three conductors in parallel 3C0 , hence the zero sequence
2 0
C0 (5.39)
D
3 ln 3c
d 3c
where:
1
D3c Daa Dbb Dcc D D D 2
ab
2
bc
2 9
ca
(5.40)
= mutual GMD between all three conductors and their images
and:
1
d 3c d aa d bb d cc d d d
2
ab
2
bc
2 9
ca 3 d aa d eq2
(5.41)
y
Conductor
h
-h
Image
From Ampres Law, the magnetic field vector at the radial distance r from the
conductor is directed clockwise and has a magnitude:
I
H (5.42)
2r
2I Ih
Hx cos , Hy 0
2r x h
2 2
(5.43)
perfect earth conductor (whatever the permeability might actually be). The
discontinuity has to be accounted for by a surface current density equal to H x .
2 Ih dx
H x dx 2 H x dx
0
0 x2 h2
I
(5.44)
Thus, there is a return current in the earths surface under the conductor, and
our rather dubious adoption of the method of images appears to be vindicated.
0 I hdr 0 I 2h dr
2 R r 2 h r
(5.45)
0 I
ln h ln R ln 2h ln h (5.46)
2
I 2h
0 ln Wbm -1
2 R
Since L I , the inductance per unit length of the conductor with perfect
earth return is:
0 2h
L ln Hm -1 (5.47)
2 R
Assume a second conductor, not carrying any current, at point P. The first
conductor at 0, h carries current I.
P
y
h d
-h
The flux per unit length linking the second conductor is:
0 I h dr 0 I D dr
21
2 d r
2 h r
(5.48)
I
0 ln h ln d ln D ln h
2
I D
0 ln Wbm -1
2 d
0 D
L21 ln Hm -1 (5.49)
2 d
The similarity of form between the formulae for inductance per unit length and
potential coefficients is obvious, and we can now set up the inductance per unit
length matrix for the ideal multiconductor transmission line:
(5.50)
L 0 A Hm 1
2
where A is identical to the matrix used in the calculation of potential
coefficients.
With practical conductors the current is not confined to the surface, but
diffuses exponentially some distance into the conductor. We define the skin
depth:
(5.51)
2
m
where:
= resistivity ( m ) (5.52)
The flux inside the conductor, only partially linked to the current, contributes
an additional self-inductance to each conductor. Mutual inductances are not
affected. It can be shown that for sold cylindrical conductors and uniform
current distribution this internal inductance per unit length is a small fixed
value equal to:
0
Linternal 0.05 Hm -1 (5.55)
8
regardless of size (for a proof, refer to the prescribed textbook, or others).
0 Dii
Lii ln (5.56)
2 d ii
where we have Dii 2h and d ii R for an ideal conductor (zero skin depth).
For the low frequency real conductor with uniform current distribution:
0 0 Dii
Lii ln (5.57)
8 2 d ii
0 1 Dii
ln
2 4 R
0 Dii 14
ln e
2 R
Therefore Eq. (5.56) is also valid for the real conductor if we use:
1
d ii e R 0.778 R
4 (5.58)
Note that there is a difference in the meaning of the GMDs used
in inductance and capacitance calculations. In capacitance
calculation we need the GMD between surface charge densities,
while in inductance calculation we need the GMD between the
current densities. The numerical difference is quite significant
for the self-GMDs, but usually insignificant for mutual GMDs.
Furthermore, with relatively large spacing between conductors
the mutual GMDs may often be replaced with the corresponding
centre-line distances.
The problem of self and mutual impedances of parallel conductors with earth
return was solved independently, and almost simultaneously, by Pollaczek and
Carson in 1926. The finite resistivity of the earth causes the earth return
currents (surface currents in an ideal earth) to penetrate well below the surface.
The effect of this penetration is equivalent to increasing the image distances
Dii and Dij and thereby increase all inductances (self and mutual). At the low
frequencies and typical conductor heights of power lines the effective values of
Dii and Dij are much larger than double conductor height and approach the
The self and mutual inductances per unit length of the conductors thus become:
0 DE
Lij ln Hm -1 (5.59)
2 d ij
where:
(5.60)
DE 659 (m)
f
(5.63)
d ii = self-GMD of current distribution of conductor i
(5.64)
dij = mutual GMD between conductors i and j
DE 1.31 (5.65)
Values of DE
Typical values of DE are seen to be much larger than the ideal earth image
distances ( 2h ).
Assume we are dealing with real earth, so that the image conductors may be
placed at the equivalent depth of earth return DE . Also assume line height and
all lateral dimensions are DE , al three conductors are identical, and the line
is transposed.
b
Conductors c
a
Earth
Distant images
0 DE
Ls ln Hm -1 (5.66)
2 d aa
0 1 DE D D
Lm ln ln E ln E
2 3 d ab d bc d ca
0 DE
ln
2 3 d ab d bc d ca
0 DE (5.67)
ln
2 d eq
L1 L2 Ls Lm
L0 Ls 2 Lm (5.68)
Hence:
0 d eq
L1 L2 ln Hm -1 (5.69)
2 d aa
and:
0 DE D
L0 ln 2 ln ln E
2 d d eq
aa (5.70)
0 DE3 0 DE
ln 3 ln
2 d aa d eq2 2 3 d aa d eq2
Hence:
0 DE
L0 3 ln Hm -1 (5.71)
2 d 3c
where
1
d 3c d aa d bb d cc d d d
2
ab
2
bc
2 9
ca 3 d aa d eq2
(5.72)
for capacitance calculation in Eq. (5.39), but the numerical values are different
because different values of d aa are used.
1
n
xi
n
i 1 (5.73)
(5.74)
d eq 3 d ab d bc d ca = equivalent spacing
We can calculate the mutual GMD between two cross-section areas by dividing
the first area into n elements, and the second areas into m elements, then
calculate the geometric mean of the n m possible distances. The required
GMD is then the limit value when both n and m .
The mutual GMD between two areas with circular symmetry is their centre-
line spacing.
Previous equations Eq. (5.25) and Eq. (5.26) cover the calculation of the self-
and mutual GMDs of bundle conductors.
Examples of Self-GMD
Geometry Self-GMD
Note of caution: The above concept of GMDs breaks down when there is a
significant proximity effect, or a significant skin effect.
Proximity effect causes both current and surface charge
densities to be non-uniform. Skin effect causes non-
uniform current density in conductors of large cross-
section.
Example
6r
1 2 3 4 5 6 7 Answer:
1 a 2 b 4 b 2 2
1
2 2 a 2 b 4 b 2
r a 7 2 24 b12 46 49
3 b 2 a 2 b 4 2 1
4 4 b 2 a 2 b 2 74 24 12 6 12 49
r e 2 2 3 2
5 b 4 b 2 a 2 2
1 48 6
6 2 b 4 b 2 a 2
re 28
2 3
49 49
7 2 2 2 2 2 2 a
2.177 r
Distance matrix
Example
d
r 5 mm d 500 mm
r d
0 0 D
C L ln
ln
D ds
R
where:
R = radius of conductor
1
ds e R 4
(Prove the above formula using Eq. (5.13), Eq. (5.56) and Eq. (5.58))
1
2
4
1
2
1 1
d s e R
4
2d e 8 2 4 Rd 1.0495 Rd 52.47 mm
For capacitance:
1
ds R2 2d 2 4 Rd 1.1892 Rd 59.46 mm
2
4
Then:
8.8542
C 13.06 pFm -1
500
ln
59.46
4 10 7 500
L ln 902 nHm -1
52.47
Example
Calculate the positive and zero sequence inductances per metre of a three-
phase 50 Hz transmission line with conductors spaced at 2.7 m in a flat
horizontal arrangement. Each conductor consists of seven strands of 3 mm
diameter copper wire. The equivalent depth of earth return is given as 1300 m.
Then:
3402
L1 2 10 7 ln 1.39 Hm -1
3.266
We also have:
Then:
0 DE 1300
L0 3 ln 3 2 10 7 ln 4.957 Hm -1
2 d 3c 0.3356
At low frequencies this earth return resistance has been shown to be:
0
RE 2 f 10 7 m -1 (5.75)
8
49.3 mm -1 at 50 Hz
Eq. (5.75) is a low frequency approximation valid over the range in which the
equivalent depth of earth return is much greater than the height or lateral
spread of the conductors. The surprising feature is that the earth return
resistance is independent of the resistivity of the earth! The inductance depends
on the resistivity, but the resistance does not. The qualitative explanation is:
the skin depth increases proportional to the square root of the resistivity, and so
does the lateral spread of the earth return current. Therefore the effective cross-
section of the earth return path increases in proportion with the resistivity, and
the resistance remains constant. At high frequencies however the lateral spread
is fixed, with the net result that the earth return resistance is proportional to the
square root of frequency at high frequencies.
Rii RE Ri (5.76)
Rij RE (5.77)
where:
(5.78)
Ri = internal resistance per unit length of conductor i
A transmission line can have both series and shunt components of resistance.
For the purpose of network analysis the shunt components (insulator leakage,
corona, etc.) can usually be neglected. The series components however are
significant.
The effective series resistance must account for all losses of real power that are
proportional to the square of the current. Thus power dissipated by induced
currents in earth, earth wires, structural steelwork, etc. as well as power
dissipated in the conductor itself would have to be accounted for in the
effective series resistance. In case of practical power lines carrying only
positive or negative sequence currents these stray effects are generally
negligible, but not necessarily so in case of zero sequence currents.
The earth return resistance RE cancels out for positive and negative sequences.
Thus the positive sequence resistance is near enough just the conductor
resistance, allowing for skin effect when necessary.
R1 Ra
(5.79)
R0 Ra 3RE
Eq. (5.75).
Summary
The method of images is used to determine the capacitance and
inductance per unit length of infinitely long overhead transmission lines.
This method assumes an ideal earth plane.
The formulae for inductance and capacitance per unit length utilise the
concept of a geometric mean distance (GMD). The GMD can handle the
cases of bundle conductors and multiple circuits (e.g. three-phase).
Inductance calculations need to take into account skin depth and the finite
resistance of the earth, giving rise to the concept of equivalent depth of
earth return. In addition, there is a fixed contribution to the overall
inductance by the internal inductance. These complications are not
present for capacitance calculations.
References
Carmo, J.: Power Circuit Theory Notes, UTS, 1994.
Exercises
1.
Calculate the capacitance per metre of a long horizontal conductor suspended
at a height of 10 m above the earths surface for the following options:
2.
A double-circuit three-phase transmission line is constructed so that its six
conductors are placed at the corners of a regular hexagon with side length of
two metres. The conductors are cylindrical and 20 mm in diameter. There are
no transpositions.
b b b b
c a c c
Figure A Figure B
(a) Calculate the line inductance (positive sequence) of circuit 1 with circuit 2
open-circuited.
(d) Figure A is the better arrangement. Why? (There are two reasons.)
3.
A 132 kV three-phase urban transmission line has conductors arranged as
shown. The line is transposed, and the conductors have a radius of 10 mm.
4.
An earth fault occurs on a three-phase transmission line. The fault current is
5000 A, the frequency is 50 Hz. The line is not provided with any earth wires.
A wire fence runs for a distance of 1 km parallel to and 30 m from the line.
Calculate the longitudinal induced voltage in the fence, assuming the earth to
have a uniform resistivity of 250 m. Is the answer realistic?
5.
A telephone line runs parallel to a three-phase power line carrying 500 A
symmetrical currents. The frequency is 50 Hz. Neither line is transposed.
Relevant dimensions are as shown:
untransposed
power line
untransposed
telephone
6m 6m line
5m
20 m
1m
Calculate the voltage induced into the telephone line (loop voltage).
6.
An underground three-phase high voltage cable consists of three conductors,
each 12 mm diameter. Each conductor is surrounded with a coaxial metallic
sheath, with a 24 mm inside diameter. The insulation between the conductor
and the sheath has a relative permittivity of 2.5. The centre-line distances
between all three conductors are 50 mm. The sheaths are connected to earth,
and are cross-bonded (transposed) so that no positive sequence currents are
induced in them.
r = 2.5
50 mm 50 mm
outer sheath
shield
50 mm