Overhead Line Thermal Rating Calculation

Based on Conductor Replica Method

Daniel C. Lawry, Member, IEEE, and J ose R. Daconti, Member, IEEE
Abstract- .4 new methodology for determining the dynamic
thermal rating of overhead transmission and distribution tines is
described. This includes using a simple monitor and a calculation
algorithm based on the tested and widely accepted IEEE 738
standard. The monitor, called the ThermalRate Monitor,
consists of a eonductor replica, which does not require physical
cnntact with the line itself. Determining the actual rating of
overhead tines allows the transfer of more power while still
operating safely. This new technology b a low-cost way for
electric utilities to nprate power lines, while avoiding expensive
physical modification or mbnilding ofthe lines. The new monitor
and method of dcnlation described in this paper have
advantages over the existing technology including accuracy, ease
of installation, low cost, and durability.
Index Terms-. Power transmission tines, power transmission
meteorological factors
I. ADVANTAGES OVER EXISTING METHODS
here are five main dynamic line-rating techniques:
T weather monitoring, tension monitoring, sag monitoring,
line temperature monitoring, and the new PTI ThermalRate
technique. There is no overall best method. The most
appropriate method is a function of the particular application
and is based on various issues including accuracy, cost, and
installation factors. The new monitor and method described
in this paper comes as the result of experience with the
present monitoring technology and methods of upgrading
lines. Some of the drawbacks with the existing technology
have been resolved with the ThermalRate monitor. The
following addresses four main problems encountered with the
existing technology.
A. Accuracy at N o m 1 Electrical Loading
Tension, sag, and line temperature monitoring have a
common disadvantage in that they cannot determine ratings
during times of low electrical load. For example, if a tension
monitor is placed on a line with normally low load (less than
about 1 amp/mm2 or 35% of the lines static rating), it is
physicaIly impossible to determine the rating, since the
conductor temperature rise above air temperature is very
small. During a contingency or otherwise high electrical load
resulting in high conductor temperature, tension and sag
D. C. Lawq and 1. R. Daconti are with Power Technologies, Inc
monitors are able to determine the most accurate thermal
rating. It is often not acceptable that only after the
contingency initiates can the rating be determined and
provided to the operator. The new method described in this
paper uses a different approach, which allows determining the
line rating without dependence on the actual line load. With
this approach, the rating can be provided to the system
operator before as well as during the contingency. Also,
ratings can be consistently calculated, saved, and used as a
basis for real-time prediction and for off-line analysis.
At present, most lines have fairly low load during normal
system operating conditions. The possibility of high line load
occurs during contingency loading scenarios. For example, it
may be determined that if a certain backbone 345 kV line is
lost in a contingency, the surrounding lower voltage lines ( 1 15
or 69 kV) may become overloaded. Therefore, the load on the
345 kV line is limited to avoid this. Under normal conditions,
the static rating of the lower voltage lines limit the total power
transfer capability even though they may have minimal load.
B. Accurate Measurement of Weather Conditions
Weather monitoring can be used to effectively determine
overhead line ratings. At low wind speeds, though, where
knowing the rating i s most important, rotating wind speed
sensors stall. This will worsen over time due to hearing wear.
Also at low wind speeds, the wind direction can become
variable (turbulent) and therefore its effect on conductor
rating becomes hard to measure. The ThermalRate Monitor
has no stall speed and in fact becomes more accurate at low
wind speeds. Its design intrinsically performs an average of
the wind speed and, direction. In addition, the new monitor
measures the important effects of solar radiation,
precipitation, and other minor weather effects. Measuring
each of these parameters independently would require
extensive weather equipment.
C. Effort and Cost of Installation
Monitor installation can be very expensive if it requires
physical modifications of the line. In addition, if the
installation requires a line outage, this outage might be vefi
hard or impossible to obtain. Sag monitors and line
temperature monitors can usually be installed live by a line
crew without much hassle. Although a tension monitor is
theoretically possible to install live, many utilities have rules
preventing such work. Installation of a tension monitor
involves selecting a favorable deadend structure and
0-7803-81 10-6/03/$17.00 02003 IEEE 880
disconnecting one of the insulator strings in order to insert a
strain gauge. The tension monitor therefore has the highest
installation cost. The PTI rating monitor has low installation
cost, since it does not require an outage to install, does not
need to make contact with the actual conductor, and is
flexible in where it can he located.
D. Purckse Cost
The new monitor was designed to have a low purchase cost
in comparison to other types of monitors used for the purpose
of rating. It has a very simple design with no moving parts.
As a result, it is much more rugged and reliable than
conventional monitors. Due to its low cost, it can be used as a
system-wide uprating solution.
It must he noted that the new monitor, as with weather and
line temperature monitors, allows determination of rating at a
point in space. Care must be taken to place the monitors in
appropriate locations that represent the weather conditions
along the line. Tension and sag monitors can provide a better
indication of the rating over distance, if the electrical load is
sufficient to allow this. Therefore, for long lines, with varying
weather along the line length, fewer tension or sag monitors
might be needed.
II. LINE RATING FUNDAMENTALS
The thermal rating of an overhead line is the maximum
current that the line can handle without overheating. The line
rating is a function of the weather conditions seen along the
line including wind speed, wind direction, air temperature,
and sun. Other secondary influences might also affect the
rating, which are not considered in IEEE 738. These include
such effects as various forms of precipitation and indirect
solar radiation.
Fig. 1 shows that over time a line has a distribution of
ratings. This figure shows a whole year of actual rating data.
Low ratings correspond to times of low wind speed and full
sun. High ratings correspond to times of high wind speed
crosswise to the conductor, no sun, andor precipitation.
The thermal rating of most lines is based on sag. As
electrical current increases through an overhead conductor,
the line temperature increases and therefore the line sags.
Each line has a minimum clearance to ground, which must
never be violated for safety reasons. The thermal rating is the
maximum current, which results in the line exactly sagging
down to the minimum clearance. Any additional current
would result in too much sag and a safety problem. Most
utilities have adopted a set of semi-worst-case weather
conditions that establish the lines static rating. Common
weather assumptions are 40 C (104 F) air temperature, full
sun, and 2 ft/s (1.4 mph) wind speed crosswise to the
conductor. The static rating for the line shown in Fig. 1 is only
920 amps. In reality, though, the actual line rating rarely falls
to the static rating. It is even less likely that the line rating
falls at the same time as the actual load is high. Therefore,
much of the lines capacity is wasted.
Line Rating Distribution
(I year of data. Drake conductor)
1w
90
80
8 70
s a l
E 50
j ;
20
IO
0
0 500 1000 1500 2000 2500 3000
Line Rating (amps)
Fig. 1. Line rating
There are a number of methods to increase the capacity of
a line. Physical approaches such as raising structures,
reconductoring, or retensioning might he feasible. They are
often expensive or even impossible, though, due to the
required line outage. Monitoring approaches try to harness
the capacity that already exists by permitting safe operations
above the static rating of the line. The data can be used for
either off-line rating analysis or in a real-time mode. In the
off-line mode, the data can be used to perform a better
evaluation of the static rating. Seasonal or time-of-day ratings
can be created. Off-line analysis can also be used to
determine how much can be gained by real-time operation. In
the real-time mode, the ratings are reported to the operator in
real-time, and probably involves posting of the ratings on the
existing SCADA display.
m. GENEW RATING CALCULATION METHOD
An equation to calculate conductorrating is developed by
first recognizing that the total input heat (per unit length) to a
conductor must equal the total output heat in the steady state.
The conductor is heated by ohmic losses (IR) and solar input,
and it is cooled by convection and radiation. Eq. 1 and Fig. 2
show this main heat balance equation:
where:
QSOkZ,
12R
heat input due to solar radiation, Wlft.
heat input due to line current
(R is a function of conductor temperature),
Wlft.
heat output due to convection (a function of
wind, air temp, conductor temp), Wlft.
heat output due to radiation (a function of air
temp and conductor temp), Wlft.
Qco.,c,i;n
Q,d~t i m
881
Solar Step 1, The heat balance equation (Eq. 1) is applied to the
monitor's heated rod and is reorganized to solve for Qmmmmn :
Wind Radiation Heating
Cooling Cooling
Gonvection =Qso~or +12R-Qrodinrion
(3)
fig. 2. Heating and moling of an overhead conductor
Now. the eauation can be reworked to solve for current as c *
a function ol' the weather conditions. The Qmmns,n. Q,eo, ,,,,".
Qah, and R terms are all functions of weather conditions and
of conductor remoeraturc. If the weather conditions are
measured and the conductor temperature is set to the
maximum allowable conductor temperature, the calculated
Fig,3A, 7he-tRatemonita
current is the rating current as shown in Eq. 2.
Qconvecrion+Qrmjiotion - Ysohr
Irminp = R (2)
IV. NEW MONITOR PHYSICAL DESCRIPTION
The ThermalRate monitor determines the line rating by
measuring how the weather conditions heat and cool the
conductor. The monitor consists of two aluminum rods which
function as simple conductor replicas (Fig. 3A and 3B). The
rods are chosen to be the same material and diameter as the
line conductor. The rods are each ahout 2 feet long to account
for temperature fringe effects on the ends, and they are
Fig. 3B. monitor
step 2. It can be proven that the equivalent cooiing will
result if the solar input, Qmjm, is set to zero and the air
separated by about 2 feet so that they'do not affect each temperatures within the Qmnmhn and Qwlhn terms are
other's temperature. An internal near the longitudinal
replaced by the measured unheated rod temperature. This
center measures the temperature of each rod, The monitor is
eliminates the QSOI v and makes it to
the ambient temperature. The proof of this equivalence is at
located near the line and pointed in the same direction as the
the end of this paper.
line in order to experience the same weather conditions as the
line itself.
Step 3. In Eq. 3,' Qmnmtion (watts per unit length) for the
One of the rods has a resistive heater (running at a nearly heated rod can be easily determined. The input power, IzR, is
measured. Qdhfwn is a simple function of air temperature
constant wattage) within it, which increases its temperature
above that of the other rod, The weather conditions influence
(now Illheated rod temperature), conductor temperatnre (now
the temperatures of the
conductor replicas. By comparing heated rod temperature). monitor emissivity. and monitor
the temperatnres of two rods, the line capacity can be
diameter.
calculated and eventually supplied io the power system Step 4. Knowing Qmm=tion. the ''effective'' wind velocity
can be calculated from the IEEE 738 convection equations.
operator.
This effective wind velocity is a wind speed and wind
direction pair that results in the correct conductor cooling. For
example, as far as the conductor rating is concerned, the wind
speed at a particular angle is equivalent to a different wind
speed at some other angle. For example, a 2 ft/s wind speed
perpendicular to a Drake conductor is equivalent (meaning
10.6 ft/s wind speed parallel to the conductor. The effective
wind speed is calculated perpendicular the
conductor axis, and its calculation is also useful for additional
weather analysis. If it is desired the absolute
wind speed and direction, the monitor must consist of 3 pairs
v, CALCULATION DESCRIPTION OF METHOD
In by measuring the temperatures Of both the
heated and unheated conductor replicas and the wattage into
wind) can be calcu'ated from the heat balance equation.
Then, the rating can he determined using this wind in a
second Of the heat ':lance equation. The
calculation method for steady-state line rating is as follows:
the heated replica, the total cooling effect (mainly due to that it yields the same cooling effect and therefore rating) a
882
of rods at 120-degree angles to each other, and the
appropriate algorithm must he used. This approach is useful
for calculating the rating of multiple lines extending out of a
substation at different directions, or for lines that frequently
change direction.
It is important to note that the monitor does not need to
have exactly the same emissivity or diameter as the line itself.
The effect of differences is to alter the unheated rod
temperature, and if there are known differences, this value can
be compensated.
Step 5. The line rating is now calculated using the heat
balance equation applied to the line conductor (Eq. 2 is
repeated here). The previously determined effective wind
velocity is used within the Qmnmtioo term.
lTalin8 =J V (4)
where:
QmI, is set to 0, and the air temperature terms within
Qmnvecuon and Qdkuon are set equal to the measured
unheated rod temperature.
The conductor temperature terms within Qcom, ~u.
and Qdkuo. set to the user-input maximum
allowable temperature.
The line emissivity and diameter are used, not the
monitor parameters.
Resistance, R, is a function of conductor temperature and
is calculated at the maximum allowable
conductor temperature.
The previously determined effective wind velocity is
, used within the Qmnvecdon term.
In summary, the three monitor measurements measure the
effect of all weather conditions to allow accurate calculation
of the line thermal rating.
VI. EXAMPLE RATING CALCULATION
As an example, consider a monitor that has been set up to
rate a line with Flamingo conductor. The monitor is oriented
in the same direction as the line. Assume the following
parameters:
Monitor diameter: 1 .O inch
Line conductor:
Monitor emissivity:
Line emissivity:
Maximum allowable temperature: 100 C
Flamingo (1 .O inch diameter),
at sea level
0.8 (set by factory)
0.8 (determined by measurement
or engineering judgement)
Also assume that the monitor measurements at a particular
time are the following:
Heated rod input power:
Unheated rod temperature: 30.2 C
Heated rod temperature: 59.1 C
First, the output radiation term is calculated using the two
rod temperatures, monitor diameter, and monitor emissivity,
eradintion =0. I38De
18.94 Wlft
, ' I (5)
[ [ Theared;g +273 ,' - [ %heated rod +273
100
= 3 . 3 3 w / p
where:
D monitor diameter, inches
e monitor emissivity, unitless
Thmd heated rod temperature, C
Thead unheated rod temperature, C
Now, QmnMtion is simply calculated according to QComeuon
=12R - Qdk,jon =18.94 - 3.33 =15.61 Wlft.
IEEE 738 has three convection equations, and magnitude
of the wind speed determines which one is applicable. The
first forced convection equation is appropriate in this
example, and is listed below. Refer to IEEE 738 for a
detailed description of the variables and of the other
convection equations.
The effective wind velocity, V,, is calculated to he 3.71
ftJs perpendicular to the conductor.
Now the Flamingo conductor rating can be calculated using
the effective wind velocity, the unheated rod temperature
substituted for air temperature (solar input set to zero), the
actual line parameters (emissivity, diameter), and the 100 C
maximum allowable conductor temperature. Eq. 4 is used,
and the rating is 1205 amps. It is important to note that the
monitor does not need to have exactly the same emissivity or
diameter as the actual conductor. This method of calculation
will allow compensation for differences.
VII. MEASURED DATA AND RESULTS
Fig. 4 shows an example of the monitor temperature
measurements over '24 hours. In this example, the monitor is
oriented parallel to a line running hue North, located at 45
883
ENective Wind Speed (perpendicular l o conductor)
Themal Rate Monitor vs Ultrasonic Wind Monitor
North latitude, and represents a 1 inch conductor with
emissivity of 0.65. During this September day, solar input
causes both rod temveratures to rise above air temperature. -4
The figure shows that the unheated rod temperature is nearly
the same as air temperature at night, and rises 3 to 5 C above
air temperature during the day. Scattered clouds in the
afternoon cause periods of decreased solar input. In this
example, the heated rod power input was nearly constant at
19.2 W/ft (not displayed on the figure). In addition to air
temperature and solar input, wind greatly influences the
heated rod temperature. The effective wind speed
perpendicular to the conductor was calculated from the two
rod temperatures and heated rod power and is shown in the
fimre. The correlation between effective wind and heated rod
3 2 0 . 0
il
0.n
-
7. 7. 7.
temperature can beclearly seen.
c z $ : ; i + o . - f
Fig. 5. Effective wind speed of new itor vs. ultrasonic wind sensor
Measured Rod Temperatures and
Calculated Effective Wind, Sept 26
.l_*
ENective Wind Speed Boosted During Rain
ThermalRate Monitor YS Ul trasoni c Wind Monitor
I-
.. . . ~ U ~ ~ ~ ~ N ~ ~ ~ ~ ~ ..
0 - - - ;;go
0 o - - - - - - - -
X g ~ ~ ~ ~ g ~ e a o o n * a r
Fig. 4. Measured md temperawes and calculated effective wind
Y - - -
Fig. 5 compares the effective wind speed determined by the
ThermalRate monitor and with an ultrasonic wind sensor.
The wind speed and direction measured by the ultrasonic
wind sensor were converted to an equivalent wind speed
perpendicular to the monitor. The Thermalkite monitor is
shown to accurately determine the wind, even at low wind
speeds where a rotating anemometer would stall. In addition,
the figure shows that the ThermalRate monitor responds to an
average of the wind speed and direction (as would the actual
conductor), which is useful in limiting fluctuations in wind
and calculated rating. The ability to calculate wind accurately
is an interesting verification of the accuracy of the IEEE 738
model.
The design of the monitor intrinsically considers weather
effects in the rating calculation that normal weather
monitoring does not consider. These include direct solar
input, indirect solar radiation from the ground or clouds, a
vertical wind speed component, clear sky radiation, and the
various types of precipitation. Fig. 6 shows an example of
how rain increases the conductor cooling effect, which can be
represented by an equivalent (effective) wind speed.
Ultimately the increased cooling results in increased line
rating.
Rg. 6. Rain bmsfs the effective wind sped
WI. FINAL COMMENTS
If the transmission system operator knows the true thermal
limits of operation, the operator can seize opportunities to
ship more power and avoid unnecessary load shedding when
contingencies arise. As a result, more power can be
transferred with little additional investment. This is especially
important in the new business climate, which has dramatically
changed the existing usage patterns of transmission. This new
climate includes considering the demands of transmission
open access, uncertain future circuit loading due to load
growth and dispatch changes, and new patterns of wheeling
power to achieve economic energy transfers not envisioned in
the deregulated environment. New lines are costly and permits
can be very difficult to obtain. Monitoring allows existing
lines to carry additional power while maintaining high
reliability and safety. The new monitor and method of
calculation described in this paper have advantages over the
existing technology including accuracy, ease of installation,
and low cost.
884
4. Now, in the case where lZR heat input really does exist,
the original heat balance equation is again listed as
foil0ws:
IX. SOLAR HEATING PROOF
This is a mathematical proof that measuring solar heat
input, Qah, and air temperature, T,, is equivalent (yields the
same cooling effect and rating) to the following:
Q~~~~=I 2 R =K~*(T, - T,)+K ~ *(T: -T:)
(11)
(I ) Setting the Qmh term to zero, and
(2) Measuring the solar temperature, T,, and substituting it
for air temperature, T,, within the QFowccmn and Qdm,to. terms.
The solar temperature is defined as the temperature of the
Substituting the equation for Qmh (Eq. IO) into the original
heat balance Eq. 11:
conductor with no electrical load (only solar input).
1. A heat balance eauation for the conductor is written
[Kl * (T, - T, )+ Kz * (T: - d )I +I ' R =
(12)
yr, -T, )+ K, *(e -T:
!mowing that the heat input must equal the heat output
(watts per unit length):
Simplifying:
In this final heat balance equation, T, is shown to be
replaced with T,, and Q,,, is zero. This proves what weset
out to do. Therefore, the unheated rod temperature (solar
temperature, TJ can be measured to yield equivalent results to
measuring the air temperature (T,) and solar input (Qa1,).
where:
I'R
Q, , , .
Qradiation
heat input due to solar radiation, Wlft.
heat input due to line current, W/ft.
heat output due to convection, Wlft.
heat output due to radiation, Wlft.
Q ~I ,
X. REFERENCES
[I ] EEE Std 738.1993, "I EEE Standard for Calculating the Current-
2. According to IEEE 738, the convection term, QcaWwn. is
proportional to the conductor temperature rise above air
temperature. It can be written in the form KI*(T,-TJ . T,
is the conductor temperature. and Ta is the air [ZI D. C. hv, " Ov e h d Line wing Monitor," U. s. Patent 6 441 603,
Te ~ r a t u r e RelationshipofBareOverheadCondunors."
temperature. K1 simply represents the part of the Aug. 27. ZOM.
convection equation proportional to (Tc-TJ . Similarly,
the radiation term, Qdm,mn , can be dtten in the form
Kz*(T$-T:). Substituting into Eq. 1, the heat balance
XI. BIOGRAPHIES
equation becomes:
Q~~~~~=I ~ R =K , *(T, -T, ) + K , *(T: - T: )
( 8)
3. Performing the heat balance on the unheated rod, which
has no conductor current (IZR=O):
Dan h wr y has a BSEE degree from Clarkson
University. He has worked for Power Technologies,
Inc. since 1993 in the area of thermal uprating.of
overhead lines and calm outdcor power equipment. Mr.
Lawry has been a "her of the I EEE since 1993.
(9) Jose Dseonti has waked for Power Technologies, Inc. since ZWI. He has a
MSFE Honors Degree from Federal Schml of
Engineering of hajuba (Brazil) and is a Camell
The solar temperature, T,, is now defined as the University HHH Fellow in t he areas of E l d c Power
temperature of the conductor with no electrical current System and Electric Power Quality. Since 1978, most of
(12R=O). When there is no current in the conductor, the
his work h s been concentrated in the areas of
"ission line eleehieal design and el&om@etic
compatibility. In ZWO, he was nominated a CERE
conductor temperature, T,, will rise above air temperature due
to the solar heat input, Qmlw
Since weare performing the heat balance on the unheated
rod, we can substitute T, for T,. An equation for Qs o~, has
then been derived as follows:
Q,,~, =K~ *(T, -m+ K, YT, ~ - ~, 4)
Distinguished Member. Mr. Daconti has been an E35 " h e r since1993.
Qsoror =KI *( T, -To )+K, *f d -d )
(10)
Note that Qml, is Zero only if T, is set equal to T, (KI and
KZ terms are never zero).
885