Notes 10: Conductor sizing & an example

10.0 Conductor sizing

Since it is impractical to manufacture an

infinite number of wire gauges, standards
have been adopted for an orderly and simple
arrangement of such sizes.

The American Wire Gauge (AWG) sizes

conductors, ranging from a minimum of no.
40 to a maximum of no. 4/0 (which is the
same as “0000”) for solid (single wire) type
conductors. The smaller the gauge number,
the larger the conductor diameter.

For conductor sizes above 4/0, sizes are

given in MCM (thousands of circular mil) or
just cmils.

1
What is a cmil? A cmil is a unit of measure
for area and corresponds to the area of a
circle having a diameter of 1 mil, where
1 mil=10-3 inches.

The area of such a circle is πr2= π(d/2)2, or

π(10-3/2)2=7.854x10-7 inches2

1 kmil=1 inch.

1 cmil=(1 mil)2 and so corresponds to a

conductor having diameter of 1 mil=10-3 in.

1000kcmil=(1000 mils)2 and so corresponds

to a conductor having diameter of 1000
mils=1 in.

To determine diameter of conductor in

inches, take square root of cmils and then
divide by 103:
Diameter in inches= cmils
10 .3

2
Stranded conductors (multiple strands of
wire) for distribution purposes usually range
from a minimum of no. 6 to a maximum of
1,000,000 cmil.

It is quite common to specify conductors in

terms of the size (or gauge), the strands, and
the layers. For example, 4/0 6/1 indicates
AWG size 0000, with 6 strands and 1 layer.

The table below shows conductor sizes

commonly used in distribution systems.

3
10.1 Example

4
Determine the phase impedance matrix [Zabc]
and the sequence impedance matrix [Z012] in
Ω/mile for the 3-phase pole configuration
shown in Fig. 1. The phase and neutral
conductors are 250,000 all-aluminum.

Fig. 1

From the table of conductor data, we see

that for 250,000 cmil AA conductor, the
diameter is 0.567 inches, the GMR is 0.0171
ft, and the resistance is 0.41Ω/mile.

From eqs. (39,40) of Notes 9, we have:

5
  1 1 
Zˆ ii  Ri  0.00158836 f  j 0.12134ln  7.6786  ln 
 ri 2 f

 1 1  
Zˆ ij  0.00158836 f  j 0.12134 ln  7.6786  ln 

 Dij 2 f 


With f=60 and ρ=100, these equations are:

 1 
Zˆ ii  Ri  0.0953016  j 0.12134ln  7.93401
 ri 

 1 

Zˆ ij  0.0953016  j 0.12134 ln  7.93401

 Dij 


So we need to compute the self terms for the

four conductors (which will all be the same
since they all have the same GMR). We also
need to compute the mutuals, which will be
the harder part because we need to obtain a
mutual for each pairwise combination. This
is 4 things taken 2 at a time, which is
4!/2!(4-2)!=6.
So we have 6 calculations to make.
First, lets do the self term, since it is easiest.
 1 
Zˆ ii  Ri  0.0953016  j 0.12134ln  7.93401
 ri 
 1 
 0.41  0.0953016  j 0.12134 ln  7.93401
 0.0171 
 0.5053  j1.4564

To get the mutuals, we need to determine the

distance between conductors.
a to b: sqrt(42+22)=4.4721

6
  1 
Zˆ ab  0.0953016  j 0.12134ln  7.93401
 4.4721 
 0.0953  j 0.7810

a to c: 2+2=4
 1 
Zˆ ac  0.0953016  j 0.12134ln  7.93401
 4 
 0.0953  j 0.7945

a to n: sqrt(42+62)=7.2111
 1 
Zˆ an  0.0953016  j 0.12134ln  7.93401
 7.2111 
 0.0953  j 0.7230

b to c: sqrt(42+22)=4.4721. Same as a to b.
b to n: 2+2=4. Same as a to c.
c to n: sqrt(42+22)=4.4721. Same as a to b.
Therefore, the primitive impedance matrix is

 0.5053  j1.4564 0.0953  j 0.7810 0.0953  j 0.7945 0.0953  j 0.7230

 0.0953  j 0.7810 0.5053  j1.4564 0.0953  j 0.7810 0.0953  j 0.7945
Zˆ prim 
 0.0953  j 0.7945 0.0953  j 0.7810 0.5053  j1.4564 0.0953  j 0.7810
 
 0.0953  j 0.7230 0.0953  j 0.7945 0.0953  j 0.7810 0.5053  j1.4564

The abc matrix is then found by Kron

reduction based on the formula we
developed:
Vabc     Zˆ pp    Zˆ pn  Zˆ nn   Zˆ   I   Vabc
 
1
np abc

where we define:
      Zˆ 
Z abc  Zˆ pp  Zˆ pn Zˆ nn
1
np

In our problem, the necessary matrices are:

7
  0.5053  j1.4564 0.0953  j 0.7810 0.0953  j 0.7945
 
Z pp  0.0953  j 0.7810 0.5053  j1.4564
ˆ 0.0953  j 0.7810
0.0953  j 0.7945 0.0953  j 0.7810 0.5053  j1.4564 
0.0953  j 0.7230
 
Z pn  0.0953  j 0.7945
ˆ

0.0953  j 0.7810

 Zˆ    0.5053  j1.4564  0.2126 - j0.6129

nn
1 1

 Zˆ    0.0953  j 0.7230 0.0953  j 0.7945 0.0953  j 0.7810

np

Plugging in to get Zabc, we obtain

      Zˆ 
Z abc  Zˆ pp  Zˆ pn Zˆ nn
1
np

 0.5053  j1.4564 0.0953  j 0.7810 0.0953  j 0.7945

 0.0953  j 0.7810 0.5053  j1.4564 0.0953  j 0.7810
0.0953  j 0.7945 0.0953  j 0.7810 0.5053  j1.4564 
0.0953  j 0.7230 
 0.0953  j 0.7945  (0.2126 - j0.6129) 
0.0953  j 0.7810 

 0.0953  j 0.7230 0.0953  j 0.7945 0.0953  j 0.7810

 0.5300 + j1.1123 0.1269 + j0.4038 0.1256 + j 0.4235 

  0.1269 + j0.4038 0.5448 + j1.0429 0.1333 + j0.3743 
0.1256 + j0.4235 0.1333 + j0.3743 0.5418 + j1.0565 

We will refer to Zabc as the phase impedance

matrix. Note carefully that the self terms are
all different, and there are three different
mutual terms. This is because of the
difference in their positions relative to each
other and relative to the neutral conductor
and the fact that there is no transposition.

8
However, the matrix is symmetric, i.e., what
the a phase sees looking at b-phase is the
same that b-phase sees looking at a-phase.
Now let’s compute the sequence matrix Z012.

This is given by Z012=A-1ZabcA, which is

Z 012 
 0.5300 + j1.1123 0.1269 + j0.4038 0.1256 + j 0.4235 
 A  0.1269 + j0.4038 0.1333 + j0.3743  A
1
0.5448 + j1.0429
0.1256 + j0.4235 0.1333 + j0.3743 0.5418 + j1.0565 

where
1 1 1 1 1 1
1
A  1 a 2 a  and a 2 
1
A  1 a
3
1 a a 2  1 a 2 a 

resulting in
Z 012 
 0.7960 + j1.8716 - 0.0164 + j0.0327 0.0029 + j0.0352
  0.0029 + j0.0352 0.4103 + j0.6700 0.0078 - j0.0054 
 - 0.0164 + j0.0327 - 0.0072 - j0.0052 0.4103 + j0.6700 

Clearly the sequence networks are not

decoupled. Important point to be made here
is that, because sequence networks are not
decoupled, they offer no real advantage over
the abc network in terms of analysis.

9
10.2 Two phase and single phase lines

If one is given a single phase or two phase

configuration, with or without neutrals, then
one may apply the same procedure as for the
3-phase configuration, i.e.,
1. Determine resistance per mile and GMR
of each conductor.
2. Determine distance between conductors.
3. Compute primitive impedance matrix
using eqs. (39,40) of notes 9 with
appropriate value of resistivity ρ.
4. Perform Kron reduction to eliminate the
presence of the neutral and thus obtain the
phase impedance matrix.

Notes:

10
 The primitive impedance matrix of step 3
will be square with dimension equal to the
number of conductors that you have in the
configuration.
 The phase impedance matrix of step 4 will
be square with dimension equal to the
number of phase conductors that you have
in the configuration.

So a single phase configuration, with

neutral, will result in a 2x2 primitive
impedance matrix and a 1x1 phase
impedance matrix.

If by chance a single or two-phase

configuration extends from a three phase
configuration without changing the positions
of the remaining conductors, then one may
just pull out the appropriate elements from
the 3x3 phase impedance matrix
corresponding to the 3 phase configuration.

For example, if

11
  Z aa Z ab Z ac 
Z abc   Z ba Z bb Z bb 
 Z ca Z cb Z cc 

is the phase impedance matrix for the 3

phase configuration, if we run a 2-phase
lateral using phases a & c without changing
the a & c phase & the neutral position, then
 Z aa 0 Z ac 
Z abc   0 0 0 
 Z ca 0 Z cc 

Likewise, if we run a single phase lateral

using phase b with without changing the
phase b position relative to the neutral, then
0 0 0
Z abc   0 Z bb 0
 0 0 0

If positions do change, then non-0 elements

of above matrices will need to be
recomputed. However, we will be able to
use 3x3 matrices for any line configuration.

12