Section II Bare Aluminum Wire and Cable

Chapter 6

Operating Performance and Problems

Operating problems occurring in installations of bare Short-Circuit Performance

overhead conductors are of several kinds. Only those re
lated to the conductors themselves are considered herein. The ampacity data in Chapter 3, Figs. 3-11 to 3-15,
Such matters as voltage drop', system regulation, transi apply to steady-state normal operation for bare ACSR
ents and calculation of probable short-circuit currents are and all-aluminum conductors for temperatures up to
in the province of the system electrical engineer and be 1000C (60 0C rise over 40 0C ambient). This temperature
yond the scope of this book. Subjects covered' in this is frequently used for 1350-H19 conductors since the
chapter include the ability of the conductor to withstand strands retain approximately 90 percent of rated strength
short circuits and their related mechanical forces, the ex after 10,000 hours at temperature. (See Fig. 6-3.) For
tent that emergency overloads may be carried without ACSR the strength is even less affected because the steel
serious damage and the effects of arcing-burndown. Ref core is essentially unaffected at these temperatures.
erence is also made to aeolian vibration and conductor Short circuits in a power system can result in extremely
galloping with a brief description of devices that reduce large currents in conductors from the time of fault initia
their effects. tion until its interruption by the protective device, such
as circuit breaker or fuse. With modern relaying, the dura
The terms used herein relating to overload matters are tion of the 60 Hz fault current is usually only from 3 to
as follows:
20 cycles for transmission circuits but may be longer
Thermal Limit (as associated with steady-state overload for distribution lines. If the circuit is immediately re
conditions): The maximum temperature at which a con established by automatic reclosure and the fault has not
ductor can operate continuously yet maintain the mini cleared, the total fault-current time will be the sum of
mum tensile properties established by the manufacturer the interrupting times.
or the user. Heating will generally be more rapid than cooling, and
Arc-Current Burndown: Rapid failure caused by the heat loss-of-strength estimates would require integration of the
of an arc on the surface of the conductor, accompanied temperature-time curve for temperatures above the arbi
by the heat effect of current. trary "damage" level. However, as temperature is not
measured, a useful and practical alternative is to use the
Fault-Current Burndown: Failure caused by overheating current-time product and neglect the temperature slopes.
as a result of a current overload. The conductor strength When limits have been established, the time in which the
decreases sufficiently to cause tension failure. fault must be cleared can then be determined.
Fault-Current Limit: The current (temperature) and In establishing suitable fault-current limits, 340'C has
time combination which produces the maximum accep been selected as the maximum temperature for all-alumi
table loss in conductor mechanical strength. num conductors since momentary exposure to this tempera
Current Values: Unless otherwise stated, all current values ture does not result in a significant loss of strength. For
used in the discussion of overload conditions are in terms ACSR or A WAC conductors with sizeable steel content
of rms symmetrical amperes. (not the 18/1 or 36/1 strandings) an upper limit of
645'C represents the threshold of melting for aluminum
with the sleel expected to supply the needed mechanical
>I< Applying to bare transmission and distribution circuits only. The strength. The curves of Figs. 6-1 A, B, and C apply this
critical voltage-drop limitations of the National Electrical Code relating criteria using an average specific heat and assume no
to circuits under NECjurisdiction are mentioned in Sections 21O-19(a) heat loss from the aluminum strands during the short
and 21S-2(b) of the NEe, and the methods of computing drop or
obtaining it from industry-supplied tables are described, applying to duration of the fault current. Figs. 6-2 A, B, and C do
conductor sizes used mostly for interior circuits. the same for ACSR conductors.

6-1
bare aluminum wire and cable

Adjustments for 6201-T81 such cases. For example, a /1;0. 4/0 AWG 6/1 ACSR
and ACAR Conductors under 17001b tension has arcing burndown time of 10 to
14 cycles (.167 to .233 sec.) at 15,250 amp, whereas the
Values from Fig. 6-1 may be adapted to 6201-T81 and jaultcurrent limit time (there being no local arcing) is
ACAR conductors by applying suitable multiplying fac 1.6 sec for that current, under assumptions applying to
tors. Usually the value that is specified as the estimated Fig. 6,2. Also see Table 6,1. For the usual transmission
fault current is the known quantity, and the corresponding line, or those at the higher distribution voltages, relay
time is found that wiU cause the upper temperature limit co,ordination on the basis of fault current limit time
to reach 340C over 40'C ambient for 61.2 percent lACS usually is satisfactory, but for lower distribution voltages
conductor, thereby enabling the current-limiting devices in metropolitan environments consideration should be
to be properly set. For other conductors, the time for given to arcing burndown.
the 1350-H 19 conductor is multiplied by factors as below:
Table 6-1 contains representative data from arcing tests
For 6201-T81 conductor, multiply by 0.903 conducted with the conductor under tension.
For ACAR conductor, see the applicable portion of
the following example: While arcing failure times are so short that little if any
change in tension can occur prior to failure, high fault
Examples: Assume 500 kernil conductor and 20,000 rIDS 60 Hz fault currents can heat the entire line. The reSUlting increase
current. As this conductor size is not sh.own by Fig. 6~1, the time is in sag can establish contact with ground or other conduc
obtained by interpolating between values for 417 kcmil and 566.5 kcmH
to 2.80 sec for 1350 H19. Then for 6201~T81 it will be 2.80 x 0.903, or tor, initiating an arcing problem. Clearances can, there
2..53 sec.
fore, be as sigoificant a constraint on maximum acceptable
For 241 13 ACAR, the time wiJt be
current as is conductor strength.
(2.80 X 0.65) + (2.51 X Q.J5) ~ 2.71 sec.
Arcing Effects
Adjustmentjor Upper Temperature Limit Aluminum conductors resist damage by arcing better
than conductors of other metals because the arc tends to
Whereas the upper-limit temperatures specified in Figs.
cause less pitting and surface metal melting. When
61 and 6,2 are suitable for bare overhead conductors,
there are conditions where a lower temperature, limit is subjected to arc currents, the aluminum conductor sUlface
advisable, such as when the bare cable is confined in frequently shows only a removal of sheen, slight roughen
switchgear or in switching compartments. Other con, ing, and change of color over a considerable area. The
dilion" such as the use of soldered, copper terminal pads; effect described applies to arcs of less intensity than those
also may warrant a lower temperature limit. Multiplying that produce arc-current burndown. However, the ad
factors for these conditions are as follows: vantage of aluminum in this respect aids measurably in
reducing operating costs, particularly for the smaller sizes
of ACSR, in the many instances where small arcs result
Multiply time from Fig. 6-1 by from flashovers, lightning, momentary contact with a tree
For 1350-H19 6201,T81 limb, and the like.
Upper Limit From one group of tests, aluminum's resistance to
3000C 0.903 0.814 surface damage from such minor arcing was evident with
250'C 0.771 0.691 arcs ranging up to about 78 cycles duration.'
200'C 0.621 0.559
and multiply time from Fig. 6-2 by Loss of Strength
For ACSR The loss of conductor strength due to time at tempera
Upper Limit ture is a cumulative effect. Heating due to short circuit
500'C 0.845 occurrence should therefore be added to heating due to
400"C 0.721 other circumstances to estimate the condition of the con
300C 0.556 ductor. In actual practice, however, the total time of fault
currents is usually very small relative to emergency oper
For 6201-T81 and ACAR, apply these factors after ating time and is therefore igoored as an effeet on con,
applying those as listed in the preceding section. ductor strength. The temperature-time strength loss re
lationship is covered in more detail in the section on
emergency loading (Chapter 12).
Arcing
Caution must be exercised in applying the fault-current
.. C. A. Martens, "Power Arc-Over On Overhead Distribution
times, as described, for relay settings of protective devices Lines and New Developed Equipment for Protection Against
on distribution lines that may be subject to arcing buro Conductor Bumdown From That Cause," AlEE Technical Paper
dovvn. Arcing locally cuts into the conductor quickly in No. 4145, 1941.

6-2
 operating performance and problems
3

r\ " 1\\ l\
1\ 10

\ \ \ 6

~ -" \ 5

_\ ~ 4

\ ~ \ 3

1\ \ \ 1\ 1\ 2

1\1\
\ /\ \ ~
1\
~.I2/0AWG I
9

1.0

\ 1\ 6
en
\ 0

~ / 5

I/O AWGJ z
\ \ ~~ li. 0
\ \ \/ 4.<l 2 AWG L
4

'-'
L&J
en
1\ 1\ ~ 4.1 4 Awtl 2

1\
1\
1'0\0 Lil Y.
6AWG I
0.1
I
L&J
:IE

BARE STRANDED ALUMINUM CONDUCTOR

I\:
L~
A
~ ~
, -.l, 8

\/ 5

/\
\ \
\
\ , 4

'--
t = ( O.0671m
I
)2;t
m
= seconds.
= area, cmils
\ \ \ '\ '\ 2

~ r\
1\
I = amps, rms \ ~
1\
1\ 0.01
9
8
7
6
0.005
2 3456789 2 3456789 2 3 4 5 6 7 8 9
100 1,000 10,000 100,000

CURRENT IN AMPERES
Fig. 6-1A. Maximum fault-current operating limit for Note 1. Time plotted is that required for a given rms
stranded aluminum conductor. Upper temperature limit fault current to cause conductor damage due to annealing.
340C, ambient temperature 40C. 2. Graphs asSUme there is no heat loss in the conductor.

The curve for all aluminum conductors may be applied to alloy 6201-T81 and ACAR conductors by computing the
equivalent 1350-H19 cross section. The current may then be determined by extrapolating for the computed cross
section USing Figs. 6-1, A, B, and C.

6-3
bare aluminum wire and cable
3

2
I

~
\ \ 1\ \ i\
.0
\ " \. i\
9
B

H
7
\\\\ \
6
5
\ \ \\\ ,\ \
\. \\.\." ,\\ 954 000 em ~
4

3
\ \ \' ~~ I / 195.000 em'l t
2
, \\\' r-.. / 1/ 115.500 emil I

1,0
9
B

t\\\ ~ ~II IL
I 636.000 emil

605.000 emil

556.000 emil I
I
I

7
6
\ \ 1\ \\
1\
\
/
/
/ /
411.000 emil t
'"
0 5 \ 1/ '/

'~
z
4
\. ~/. / / / 391,500 emil ~
0
'-'
w 3
\ i\/ / / 336.400 emil t
'"
z 2
i\ I\r'\ K>(/ V 265.800 em , L,

\ l\ \ 1\\
~ V
w 4/0 AWG I
::E
>
01
r\ 1\ 1\ 1\ 3!Q AW I
,
9
B
7
I
, "
\ 1\', \ 1
6
5
!x \/.\\' \
4 BARE STRANDED ALUMINUM CONDUCTOR "- 1\ X\ \\\ ,\
3 , \\ \~\\ j\
2 J.
I I I I 11111
( O.067Jm ) 2 ; t
I j\\\' ~~ ~' 1\
=
t\"0~~ ~
t seconds \ I
I m = area, cmils
I = amps, rms 1\
0.0'
9
a
7

0.005
2 2 3 4 6 789
1,000 10,000 1,000,000

CURRENT IN AMPERES
Fig, 6-1 B. Maximum fault-current operating limit for Note: 1. Time plolted is that required for a given rms
stranded aluminum conductor, Upper temperature limit fault current to cause conductor damage due to annealing.
340C, ambienltemperature 40C. 2. Graphs assume there is no heat loss in the conductor.

The curve for all aluminum conductors may be applied to alloy 6201-1'81 and ACAR conductors by computing the
equivalent 1350-H19 cross seclion. The current may then be determined by extrapolating for the computed cross
section using Figs. 6-1, A, B, and C.

6-4
 3

! 10

\\' 8

\'
6

!
5

:
4

;
'\ , 1

I
!
2

!
1:590,000 em,r I
i 1/ 1.0
1,510,500 em,
, 9

,
7 8

,1,431.000 emi; I
r

/ I
6 '"
<:>
17 V 5 z
/ 1,351.000 emi I 4

0
<>
!
I~ [7 V / I .....
3

1,272.000 emi '"

!
1/ / /1 V -1
2

i,}92,500 emir
1/1/ 1/(,,vI I .....
!
i
1/ 1/
LlI),OOll em,r I
0.1
::IE
I
,
,
1.033.SOIl cmll . 9

, ,
7

I I ;
4

I I
BARE STRANDED AlUMINUM CONDUCTOR
D ,
1

;
2

!
t = ( O.0671m ) 2 ; t = seconds !

I m= area. cmils
- I = amps, rms 0.01

. 6

5 6 I 4 5 9 1 4
100,000

CURRENT IN AMPERES
Fig. 6-IC. Maximum fault-current operating limit for Note: 1. Time plotted is that required for a given rms
stranded aluminum conductor. Upper temperatllre limit fault current to cause conductor damage due to annealing.
340C, ambient temperature 40C. 2. Graphs assume there is no heat loss in the conductor.

The Curve for all aluminum conductors may be applied to alloy 620J-T81 and ACAR conductors by computing the
equivalent 1350-HJ9 cross section. The current may then be determined by extrapolating for the computed cross
section using Figs. 6-1, A, B, and C,

6-5
bare aluminum wire and cable

(I)

o
Z ,"
o
<.>
.....

.....
:E
....

21--->-

o.oo~ 2 3456189 23456789 2 3456789

.00 ItOOO 10,000 100,000

CUR R EN T IN AMP ERE S

Fig. 6-2A. Maximum Fault-Current Operating Limit for Note: 1. Time plotted is that required for a given rms
Bare Stranded ACSR conductor. Upper temperature limit fault current to bring aluminum strands to the threshOld
645 0 C, ambient temperature4QoC. o/melting.
2. Graphs assume there is no heat loss in the conductor.

66

 operating performance and problems

3

\ i\\1\
!\~i\ ~\ 9

10

1,\ '\ '\ '\ 6

\ \ \ ,\\ \' \ 5

1\ \\ '\ ,\ l\ \ I 954,000 emil 4

\ \\ \\~ ~\ \ L I I
79! ,000 emil. 3

\\\ \\( ~ ~\ / / II .500 emil

2

\\' ~~ ~~ ~ ~-
I 631.000 emil

1605.000 emil
i\ 1.0
9

I 556,500 emil 8

V / 7

'\ '\ / / / I 47: ,000 emil 6

,1\ L7. / en
\ \ \ 5 c
\ \ \ ~ Y. / / 1397 .500 emil 4
Z
0
\ \ \
~~/ L I 336.400 emil 3 '"
'"
\ \ \\ ~~X / 2&.800 emil
I
2

<J>

z:

1\\\ ~y.~ D( I 4/0 AWG II

'"
::IE
\ \ i\ . 3/0 AWG
0.1
I
9

X c\ 7

\ \ A. \_ ,,\ \
BARE STRANDED ACSR CONDUCTOR
~/\ V ,\, \
6

1\/\\ \\' \ 4

1\ \ \\ \'\\ ~\ 1\ 3

I = (O.0862m ) 2 'I = seconds

!\\ \ \\ ~ ~ 1\ 2

\\; ~ ~~ l\
~
I m = area, cmils \
I amps, rms
~'

m
I I I 0,01
9

0005

4 5 S 789
!IQOO 00,000

CURRENT IN AMPERES

Fig. 62B. Maximum FaultCurrent Operating Limit for Note: 1. Time plotted is that required for a given rms
Bare Stranded ACSR conductor. Upper temperature limit fault current to oring aluminum strands to the threshold
645 0 C. am~ient temperature 400C. of melting.
2. Graphs assume there is no heat loss in the conductor.

67

bare aluminum wire and cable

3

~ ~\

10. ~ I
8

5
, ,
:

4 ~

I
3

2
.780.000 emil I
V 1500.000 cmil
V
1.0
V I
9
1.510.500 emil ,
6
L.
en 7 ,\
0 6 .431.000 emil
z 5 I.e':
0
~
~\V :.35,.000 cmil
~\ V V
u
"-' i I
3
en i.272.000 emil
z
2
V. / V V I
I
I
"-'
:IE
/1/ 1.191,500 cmil
I
I
0 0.1 , , / 1,1I3,OQO 'mil
9

6
.033.500 emil
7

I
6

'\ K
,
5

4
BARE STRANDED ACSR CONDUCTOR , i I
3

2 I I, I I, I III! ! ! !
,

~
( O.0862m ) 2 ; t
t
'" '" seconds
I m'" area, cmils I

I amps, rms I
0.01

f= ,
7

~
6 /
0.005 I I
2 3 4
100,000

CURRENT IN AMPERES

Fig. 6-2C. Maximum Fault-Current Operating Limit for Note: I. Time plotted is that required for a given rms
Bare Stranded ACSR conductor. Upper temperature limit fault current to bring aluminum strands to the threshold
645 0C, ambient temperature 400C. oJmelting.
2. Graphs assume there is no heat loss in the conductor.

68

 operating performance and problems

Fault-Current Electro-Magnetic Forces Between this condition, with fallit current lagging nearly 900 , the
Parallel Bare Wires and Cables fault-current wave will be symmetrical if the fault is
initiated at the peak of the voltage wave, but it wi!! be
Fault currents are more likely to cause thermal damage offset (similar to Fig. 6-5) if the fault is initiated at a
to bare overhead conductors than mechanical damage. zero crossing of the voltage Wave.
However, the high electro-magnetic forces of fault cur
rents sometimes can be an important factor in line design For a fully offset wave, the instantaneous peak value,
and equipment selection. i, approaches value OA, which for zero power factor ap
proaches 2.828 I,m. as a maximum. However, because the
The electro-magnetic lateral force between long par inertia of the conductor prevents an instantaneous deflec
allel current-carrying conductors is proportional to the tion response to the applied force, some designers and
product of the instantaneQus values of current in each con
test authorities consider that a suitable current value for
ductor and inversely proportional to their distance apart.
computing maximum short--circuit force is the root-mean
See Eq. 6-1. For three-phase circuits, the vector direction
square value of current in the first current loop, which
of the three forces as well as their instantaneous values
must be known. approaches the value represented by the line OR of Fig.
65 (at zero power factor = 1.732 X 1=,), designated
The heat effects of short-circuit currents, as previously maximum rms asymmetrical current.
mentioned, are stated in terms of root-mean-square sym By similar analysis, equivalent values are obtained for
metrical amperes (I,m,) for alternating current (shown currents that provide electro-magnetic forces between the
in line CT in Fig. 6-5). However, for calculating the force conductors of a balanced three-phase circuit.
between parallel conductors under fault conditions, the
higher instantaneous value is normally used. The point of The following equation shows the relation between the
initiation of a fault is usually referred to the voltage wave short-circuit current expressed in various ways, spacing
because this is the non-variable: the current in both mag between conductors and lateral force.
nitude and phase angle is dependent on the load while the
voltage magnitude is practically constant and the phase 5.4 X 1, Ie
F=G--- (Eq.61)
angle is fixed in time.
dlO T
Transmission-line faults are practically limited in magni
tude only by the reactance of the faulted circuit. Under

TABLE 6-1

Arc-Current Burndown Times on 60 Hz Basis

For Bare Conductors Under Tension

From Tests

Conductor. I I :
!
I
tension Min Min Min !
i Min
Description Ib Amp cycles
! Amp cycles Amp cycles Amp cycles
! , ..

2 AWG7/1 ACSR 1456 1350 3Jil

,
i
4800 10 9800 4.5
,
15,750 1
i
310 AWG6/1 ACSR 1326 4550 26 9100 14 15,500 8 18,600 7
i
410 AWG611 ACSR 1701 4450 53 8580 21 15,250 10 18,700 8
336.4 kcmil1811 1701 8425 25 15,200 12 18,800 10

ACSR

350kcmil19stt. 1076 4800 19 9600 11.3 15,200 12 18,200 8

1350
500 kcmil37 str. 1456 4800 42 8800 22.5 15,400 14 18,450 11
1350 I I
: !
The arC-<lUrrent burndown times are reported from tests at Baltimore Gas & Electric Co. in a paper by W. B. Goode and G. H.
Gaertner (Edison Electric Institute, Oct. 1965). These values, with those obtained from Figs. 6-1 and 6-2. show the com
parative times for arc~current burndown and normal fault"'CUrrent limit when there is no arcing to conductor sides.

6-9
bare aluminum wire and cable

where:
F = Pounds per linear foot of conductor
G = Multiplying factor, as in Table 6-2
I, and I, = Short-circuit current in each conductor
a-c symmetrical rms amp, or in d-c amp
10,000
d = Spacing between centerlines of conductors in
inches
Example: Assume a flat 3-phase circuit of 210 AWG-6/1 ACSR I
5,000
on 7-ft spacing. subjected to a fault current of 20,000 amp rms
I I

ft
symmetrical (line CT of Fig. 6-5). What is the average lateral force
exerted on the center conductor caused by an rms symmetrical
r-:q~/ oc../
fully inductive fault current in the first offset loop (line OR of Fig. 6-5)
-~
....
~
without allowing for mechanical damping, caused by inertia, elasticity.
and side-sway friction?
From table 6-2(d) the applicable multiplying factor G is 4.17.
/ 1,000
Applying Eq. 6-1, the average force F during the first current loop,
assuming zero power factor is
I I 500
4.17 X 5.4 X 20,000 2 L I
F =---------- 10.7 Ib per It
7 X 12 X !O, II /
Under fault conditions, the mechanical action of

/ /
stranded conductors, which usually have very long span
distances compared to separation distances, is different
from the action of more rigid bus conductors described 100 en
in Chapter 13. The conductors can slap together violent ""
==
<=
ly-especially the subconductors of bundled conductor =
lines-and traveling waves move longitudinally along the
line. Experience and testing have shown that this action is
II
I
I
50 ...
1
::IE
not damaging to the mechanical strength of conductors or
insulators, but it must be carefully considered in the de / ...;:::
sign and selection of spacers and dampers. II ""
==
en

Emergency Loading 10
...
<=
"

Transmission and distribution conductors are oc 7

casionally subjected to current overloads, due to emer

I
5

gency conditions, which produce temperatures beyond the

normal thermal limit. Coincidence of peak loads with high I 3

summer ambients, shifting of additional loads to an already

loaded conductor, and use of high loadings to prevent 2

icing are some reasons for such overloads.

The question of what maximum conductor tempera / 1
tures should be permitted for emergency operation de
pends on how much loss of strength is allowable and how
long the emergency-load temperature continues. The effect 0.5
of heating is cumulative. As an example, if a conductor
is heated under emergency loading for ten hours each year
'/
for a period of ten years, the total effect is nearly the
same as heating the conductor continuously at that tem
perature for 100 hours.
Fig. 6-3 delineates the effect of time on 1350-H19 alu
0.1
minum strand strength at three temperatures which are
100 90 80 70 60
of interest to power engineers. The curves permit estimates PERCENT REMAINING OF INITIAL STRENGTH
of the change in strength of conductors which have carried
emergency overloads. Fig. 6-3. Time-temperature percent strength remaining in
An example is cited, based on the following assumptions IJ50-HI9 wire. Tensile tests made at room temperature
which should not be considered typical. after wire exposure to the indicated temperatures.

6-10
 operating performance and problems

( 3) Maximum temperature for emergency condition,

ISOoC (302 0 F)
(4) Conductor: 795 kcmil-37 str. 13S0-H19
9000 At the end of 30 years, the conductor will have been
!
! ! heated to 1500 C for 720 hours. Using Fig. 6-3 as a guide
for the estimate. the strength of 13.900 Ib would be re

8000
\i i
duced to approximately 10,600 Ib-about a 24"1. loss.
If the conductor were of the same size. but 2617 ACSR.
the strength would be reduced from 31.500 Ib to 28.200 lb.
~ 246.9 kcmil or a 10 percent reduction. The advantage for ACSR is
6201-T81 due to the steel core, which is essentially unaffected by the
temperature range considered for emergency overloads.

7000
: I Short time exposure to even higher temperatures can
: occur, and Fig. 6-4 shows the effect of 112 hour of heat
ing on similar conductors of three different aluminum
alloys. Strength loss is rapid at temperatures above ISOoC.
, For momentary exposure to elevated temperature. there
is much less reduction in strength. The cumulative effect
'""
~
6000
:
of a succession of short-time fault-currents during short
J:
:
circuits where high temperatures are possible plus emer
I.? gency operation at lower temperature can cause conductor
z
w strength loss which is of concern. However, knowledge
....
'" I i of the actual conditions-current, time. ambient tempera
'"';";:; 5000 i
I ture, wind velocity, conductor emissivity and the resulting
.
'w"
actual conductor temperatures is seldom very precise. The
"damage curves," Figs. 6-3 and 6-4, are also drawn from
'"cow ,
data having inherent variability. They therefore may be
~
u; used only as a basis for a very approximate estimate of
Z
W

4000 the actual condition of the conductor.

~i\ \
I
The creep rates at l50'C of the all-aluminum and
aluminum alloy conductor are considerably higher than
those of corresponding sizes of ACSR at the same tempera
3000
I ture. As was noted in Chapter 5, the creep rate used for
predicting 10-year final sags and tensions is based on the
1--0 creep rate at 60F.
(211.6 kcmi!) The analysis of the interaction of the thermal expansion
""AWG "\: rates, component stress levels and differing creep rates at
1350-H19
2000
I elevated temperatures to determine the effect of high tem
~ peratures on final sags is very complex. High temperatures
!
!
for time periods whleh may seem short in terms of the
life 01 the conductor can result in significant changes in
sag--especially for the conductor constructions which do
1000 I . i not have significant proportions of steel. A method of
o 100 200 300 practical calculations is presented in IEEE Paper TP
TEMPERATURE DEGREE C
69-674-PWR by J. R. Harvey and R. E. Larson.
A t}'Pical practice is to limit emergency load tempera
tures to a maximum of 125'C.
Fig. 6-4. Reduction of breaking strength of aluminum
and aluminum alloy stranded conductors of equivalent Vibration and Fatigue of Overhead Conductors'
conductance. Breaking strength tests were made at room
temperature after Vz hour exposure to elevated tempera An unprotected or improperly protected overhead con
tures. ductor may undergo wind-induced vibrations under cer
tain conditions to such an extent that fatigue failures of
strands will develop at points of restraint or support.
Similar failures have been observed at or near splices and
(1) Emergency conditions exist for 24 hours each year.
,. EPRI Handbook, "'Wind Induced Conductor Motion." contains an
(2) The uselullife 01 the conductor is 30 years. excellent treatment of this subject.

611
bare aluminum wire and cable

TABLE 6-2

Multiplying Factors for Maximum Short Circuit Lateral Force Acting Upon Suspended Parallel Wires and Cables in various Arrange
ments Assuming Balanced Loading, in Terms of Direct, or of Symmetrical RMS Alternating Fault Current, Amp (Line CT,
Fig. 6-51 = I,m,

Type ot circuit
and designation
of location on Conductor
currentwave of upon which
Arrangement of fault-producing force is Multiplying
circuit current applied factor G

d
(al A B Direct current* Aor B 1_0
0 0
'-phase a-c
symmetrical Aor B 2.0
d
A B
(bl 0 0 1-phase a-<:
asymmetrical Aor B 8.0

1phase ac
rms of first loop Aor B 5.55

3-phase a-c
d A~
0 asymmetrical A, B, or C 6.93
(cl '< i /A',B'Y 3-phase a-<:
0--- -0
l-d~1
rms of first loop A, 8, or C 4.17

3-phase a-c
asymmetrical B 6.93
(dl
l_d_l_d_1
same Aor C 6.45
A B C
0 0 0
3-phase a-<:
rms of first loop B 4.17

same Aor C 3.89

I
Although steadystate direct-current implies that a multiplying factor of 1.0 is satisfactory, the transient and overshoot at fault
initiation renders it common practice to use a factor of 2.0.

NOTES: All values assume a fully offset current wave in a fault of zero power factor without damping, or resonance effects from
support vibration.

See NEMA BU-' for adjustment factors if fault-current power factor differs from zero, as determined by XlR ratio.

This arrangement of factors differs from that of ANSI (37.32 because it is usual practice to designate fault currents of
apparatus and lines in terms of rms symmetrical amperes (I,m').

6-12
 operating performance and problems

A
1\
~
Rr
'-, 0
DQ.
S
"- ~ B B
, - i :... - I __
~" ~ I=:::'~' -===f> - T__
,-
,,-- - - f-. r-- _-1-' I-J -- r~ -- - T
f"

/\ 1\0 0

-
V'-~ c

\,
F F

Distances represent comparative current values as follows: OS = I,,,,, asymmetrical of ac component

CT == It'lll" symmeticaJ~ CB = l lx'lll> symmetrical EF = Minimum peak current values
OR ::: I,"" asymmetrical; limit to which value approaches
OA I""" asymmetrical; Hmit to which value approaches Norc: A value slated as closely approaching a designated limit
is considered as cOlnciding with that limit for computation pur
OD = Peak of de component; limit to which value ap poses. An osciJloscope trace shows that the difference is slight in
proaches most cases,

Fig, 6-5. Typical curve 01 alternating current wave during offset short-circuit (X/ R aboUl 15).

other discontinuities, and damage may also occur to sup ported to have been seen infrequently even with the
porting structures and hardware. conductors free of ice,
These phenomena have been extensively studied at out Aeolian vibration and galloping present the most serious
door test sites m which virtually any type of overhead problems, since either of them may lead to failure of eon
conductor operating condition can be duplicated. The d uctor strands at points of support or at other discon
results of many years of sueh research have been made tinuities. The most common types of damage are actual
available to the utility industry by cooperating manufac failures of the conductor, the hardware, or components
turers and technical institutes and universities. of the supports or towers, In addition, there might be
Conductor vibration and oscillation may be divided into damage and service interruptions caused by phase-te-phase
or phase-to-ground contacts during severe galloping.
three general types;
l. Sway or side swing is the most obvious and simplest
form of conductor movement in an entire span. It Aeolian Vibration ot Conductors
is caused by crosswinds or short-circuit forces, The accepted explanation of the wind-induced phe
2. Aeolian vibration is a resonant vibration, It is the least nomenon known as aeolian vibration is as follows: When
readily observed and usually the most damaging type. a comparatively steady wind blows across an overhead
It is caused by steady crosswinds. The conductors vi conductor under tension, vortices are detached at regular
brate in much the same way as any string under intervals on die lee side of the conductor-alternately
tension. Frequencies range from 2 to 200 Hz. from the top and bottom portions. The conductor is thus
3. Gal/oping or dancing is the movement that sometimes repeatedly subjected to forces that are alternately im
results when the interrelation of wind direction and pressed from above and below. The frequency of these
velocity. as well as of moisture and temperature, is forces increases with increasing wind velocity and with
such that the conductor becomes eccentrically glazed decreasing conductor diameter.
or ice-coated, A movement pattern develops in which If the frequency of the forces corresponds approxi
the entire span oscillates as a whole or in a few loops, mately to the frequency of a mode of resonant vibration
with amplitudes of several feet and at low frequency, of the span, the conductor will tend to vibrate in many
largely in a vertical direction, The envelope of mo loops in a vertical plane, As the amplitude of vibration
tion usually is an inclined ellipse, Galloping is re increases. the vortices tend to be detached in synchronism

613

bare aluminum wire and cable

with the vibration to increase the amplitude. The forces rubbed repeatedly against each other or against an armor
impressed by the wind on the conductor produce traveling rod or clamp. Micrographic studies show that the surface
waves that move away from the points of application of the layer of a strand is severely disturbed by the fretting.
forces toward the ends of the span. Each wave, i.e., each Cracks appear within the disturbed layer and-under the
crest and trough, stores part of the energy it receives from vibration stresses present in the conductor-may pene
the wind during the course of its travel, in the form of in trate into the undisturbed metal below the fretted region.
creased amplitude-the crest becoming higher and the The probable explanation of the phenomenon of fret
trough deeper. ting is as follows: Flexing of the conductor at the point of
When a wave reaches the end of an undamped span support results in a small amount of movement between
and is reflected, neither its amplitude nor the energy adjacent strands in the conductor or between strands and
stored in it is significantly diminished by the reflection. adjacent members. At the microscopic level, the contact
During its subsequent travel, the wave acquires more between metal surfaces is not a plane contact but rather
energy and greater amplitude until an equilibrium ampli a contact between asperities (minute projections). The
tude is reached where dissipation in the conductor matches intimate contact between asperities, aided by the wiping
input energy. At the ends of the span the reflected traveling action-which removes surface films-results in micro
waves are superimposed on incoming traveling waves, scopic welds between the asperities. Further movement
thereby producing standing waves. The standing-wave between strands, however, breaks these welds or the
loops thus formed have frequencies that are multiples of metal adjacent to the welds. When movements between
the fundamental frequency of the entire span. the strand surfaces are repeated a number of times, many
welds are made and broken, and a disturbed layer is
The observed relative absence of vibrations at higher
formed on the strand surface. Debris produced by the
wind velocities can be attributed in part to wind turbu
fretting can be seen as a fine dust surroun<!jng the fretted
lence. Conductor vibration is usually not observed at wind
velocities above 15 mph, although where high tensions area. Cracks are graduallY opened in the disturbed
are used and where there are steady winds of up to about surface layer by the forces involved.
30 mph, conductor vibration has been observed. Another
reason why vibration of significant amplitude does not Vibration Dampers
generally occur at high wind velocities is that these cause Perhaps the first device of any value for reducing 'ibra
high vibration frequencies, and the self-damping or tion was the festoon damper, with one or more somewhat
internal dissipation of energy in a stranded conductor loose auxiliary conductors from 4 to 12 ft. long clamped
increases rapidly with frequency. to the tensioned conductor at each side of a suspension
The tendency of a conductor to vibrate increases point. It was not until about J930 that successful damp'
rapidly as conductor tension is increased. Conductor vi ing control was achieved by the introduction of the Stock
bration is almost never observed at low stringing ten bridge damper, Fig. 6-7. This device consists of two
sions; i.e., less than about 10 to 12 percent of ultimate weights attached rigidly to the ends of a resilient steel
strength. Hence, even with dampers, limitations of 25 cable, which, in turn, is attached to the conductor by
percent final tension and 33 percent of ultimate strength means of a clamp at the midpoint. Because of the rela
initial tension with no ice or wind at the design loading tively large mass of the damper weights, the steel sup~
temperature were established for controlling aeolian vibra~ porting damper cable is not stiff enough to force them
tion, and are nOw widely accepted. to follow accurately the motions of the cable clamp, and
No exact tension limit can be defined which will assure this causes flexure of the damper cable, which results in
complete self-damping protection, but only rarely has slipping between its strands with consequent dissipatior.
fatigue damage been observed when tensions have been of mechanical energy from interstrand friction. If the
12 percent of rated strength or less. damper and conductor span can dissipate energy at a
greater rate than that at which the wind imparts it. :he
In certain areas where local wind turbulence caused by vibration of the span is suppressed to harmless prop."rj,,,lS.
broken terrain Or trees reduces the power input of wind,
somewhat higher tensions have been used on otherwise The selection of damper sizes and the best placement
unprotected spans without resultant vibration difficulties. of them on the spans are determined by the tension,
In exposed areas with steady winds, however, a few lines weight, and diameter of the conductor and ,he expected
with tensions as low as 11 percent of ultimate have suf range of wind velocities. With new efficient damper de
fered damage. signs and usual conductor tensions and span lengths, one
damper is installed near one span support point. For
Fatigue of Conductor Strands long spans, additional dampers may be required. Tension
is normally taken as that for "final condition" at about
Close inspection of fatigue failures has shown that
60'F. It has been found that protection from damaging
cracks begin at fretted regions where the strands have
vibration is most evenly balanced over the range of ex
$: IEEE Transaction on Power Apparatus and Systems, Vol. pected frequencies of line vibration when the damper is
PAS-87, No.6, June 1968, pp. 1381,1384, Fricke and Rawlins. spaced so it is approximately 70 percent of a free-loop

6-14
 operating performance and problems

(6250 X 32.21 1.094) Vl

From EQ . 6) : FLL = = 7.J ft . free loop
2 X 29.4

length (from crest 10 crest on the same side of conductor), hence

the spacing would be approximately 0 .70 X 7.3 = S. ll ft from suppon .
Normally. the spacing is increased 0 . 2 ft to allow fOT approximately
one half lenglh of Ihe suspension clamp Or insulalOr groove. Allhough
dam per spacings usually are given (rom the center or Ihe suspension
clamp or ins ulalOr groove Ihe fixed end is more nearl y the point of
tangency nea r the end of (he clamp or groove . A t dead end s. spacing
is measured from the mouth of the clamp . Precise data in th is regard
should be obttti ned from the damper supplier.

Values from Eq. 6-3, modified a s noted above , are plot

ted on Fig. 6-8 for a maximum steady w ind velocity of
15 mph . For other maximum stead y wind velocitie s,
fac tor the spacing by multiplying the JS mph di sta nce by
(IS/ prefe rred veloc it y in mph). Fig. 6-9 shows a s imil ar
solution where armor rod s are used . Armor rod s shorten
the end loop by II percent. When armor rods are used.
th ey s hould be of such length that d a mpers can be
mounted at proper s pacings ju st be yond th e rod end s .
Dimensions of Stockbridge-type dampers, weights, and
recommendations as to the number to be used for various
span lengths are obtainable from the manufacturers.
Fig. 6-6. Installing a 735,000-voll line of aluminum across Other types of vibration dampers ha ve been used in
the St . Lawrence River. cluding torsional , impact, spiral, dash-pot, visco-elastic,
and variations of the Stockbridge with extra weights and
eccentric weights. The most popular system, however, is
length from the fixed end of the span for the highest ex the one described.
pected frequency, though this distance may vary with the
design of the damper. Determination of the free-loop Spacers and Dampers for Bundled Conductors
length is as follows: Undamped horizontally bundled conductors used on
long-span high-voltage lines with spacers at the customary
f = 3.26 V/ d (Eq . 6-2)
250- to 3oo-ft intervals typically vibrate with about half
and the amplitude of a single conductor of the same size under
identical conditions. It has been confirmed that the leeward
(Tg/w) y, conductor of the pair usually vibrates at greater amplitude
FLL=-- - (Eq. 6-3) than the windward conductor.
2f
Stockbridge-type dampers are used on the individual
where: conductors of a bundled line . Spacer-dampers, designed to
f = Frequency of conductor vibration, cycles per dissipate vibration energy, are also used frequently. They
are popular on lines employing three or four subcon
sec
ductors per phase. and provide vibration control as well
v= Wind velocity, mph as the spacing function.
d = Conductor diameter, in Figs. 6-10 and 6-11 depict typical types of spacer
dampers. The spherical configuration of the end clamps
FLL = Free-loop length between amplitude peaks of spacer-dampers used on EHV lines reduces surface
of conductor vibration, ft gradients, thereby avoiding corona.
g= Acceleration of gravity, 32.2 ft/sec'
T = Conductor tension, Ib
w = Conductor weight, Ib per ft

Example: Assume a span o f 795 kcmil2617 ACSR at tension of

6250 Ib (20070 of rated strength) ("posed to a steady transverse wind
of up to 10 mph . Substituting values from the conducto r tables.
From Eq . 6-2: f = 3.26 X 10/ 1.108 = 29.4 Hz . conductor vibration . Fig. 6-7. A Stockbridge damper.

615

bare aluminum wire and cable

OAM~1l: SPACINC -INCHES

l'.IotO . . ..
,
,.,"" , 0000
'000 \ \ ,"0
I \ \ , \ \1 \ \ 1\
"'" '" l \ \ \ \ \ \ \ '" .

'"'"
(1)00

,...
i\
\
\
\ 1\
\. \
\
\ \. \ \ \ \ "'''
\ \. \ 1\ \ 1\
" ..
lOOO ~
\
1\
\ \ \ 1\ \ \ 1\ ,
lOOO5
~
\
\
\ \\ \\
\ 1\ 1\

\\\~ 1\1\,
\
\

lOOO
!
\
I
.~ 000

~
1\
\\~
\1\

\ \
\
~\

0.2
I ",e 1.0
,
CO"-iDUcrOIl: OIAMliTEIl: INCHES CONO\J(TOP DIAMETeR - INCHES.

(Use rhis graph when armor rods are not employed,) (Use this I(raph when armor rods are employed.)

T = Conductor tension Ib at average temperature. T = Conductor tension Ib at average temperature.

W Conductor weight Ib/rr. W = Conductor weight lb/ft.
Fig. 68. Spacing between damper and tangent support Fig. 69. Spacing between damper and tangem suppOrt
center to center or to mouth oj dead end. 15 mph maxi center 10 center or to mouth oj dead end. 15 mph maxi
mum vibration inducing wind velocity assumed. mum vibrationinducing wind velocity assumed.

Fig. 610. EHV 21e bundle/phase spacerdamper. Fig. 6-11. EHV 3/e bundle/pnasespacerdamper.

6-16