(*)

Thermal state of overhead une conductors

WG 05 () of SC 22 (Overhead unes)

PART 1

PROBLEMS ASSOCIATED WITH THE THERMAL STATE 0F CONDUCTORS

1. GENERAL point af view. This is discussed ri 1.2.2. Loss of cor,ductor’s mechanical

detail in the third part of this report. strength
The problem of thermal state of Many of the metals commonly used
conductors in overhead unes has been in the manufacture of the conducting
under consideration and debated by 1.2. Effect of the phenomenon parts of conductors (e.g. aluminium,
researchers and power liner users for This sphere of activity s the one aluminum alloy, copper and copper
a number of years and WGO5 has that provides a justification for the alloy) have a micro-structure that is
been concerned with this abject. research, without which it would by sensitive ta temperature in transfor
There is therefore a great deal of purely a theoretical exercise. mation of the crystaliine structure to
material available on the subject, ai The phenomenon under examina that obtained by an annealing process.
beit sometimes approached from enti tion is seen from the point of view of Consequently, there is a loss in me
rely different points of view. The the reliability of the electrical system, chanical strength and a drop in une
Working Group is of the opinion that whether as regards service cantinuity reliability. This effect is accelerated by
its own activity should consist mainly or as regards operating reliability, it higher temperatures. This process is
in careful analysis of previaus re does become important, because re cumulative in the sense that the effect
search, followed by selection of prac liabiiity is strongly influenced by une due ta a medium-high temperature
tical methods and models depending performance under extreme, unusual adds up over a period of time, sa that
on the type of application, in pinpoin conditions, which, however, may oc the conductor’s mechanical strength
ting sectors where the phenomenon is generally depends upon the tempera
cur several times during the life of the
really likeiy to matter or to have a system. ture duration curve during the life of
significant influence, and in the issue the line [1].
There are a number of effects that
of documents of an applicational na
depend on the thermal state of
ture.
conductor and influencing the line 1.2.3. Joint ageing
It is therefore a question of syste design and operation; their impor
maticaily collating material on the The thermal state of the conductor
tance is briefly reviewed below.
subject and the Group’s activity has is one of the factors that affect the
been broken down into a number of development and performance of
1.2.1. Vertical clearance
headings, under which study is cur joints.
It should be borne in mirid that the
rently in progress Obviously, apart from the characte
position of the conductor in relation ta
The causes of the phenomenon. ristics particular to the joint, a know
—

an obstruction crossed by the line is

The effects of the phenomenon. ledge of the conductor temperature is
—

flot fixed, but varies with the thermal

Practical importance. necessary ta appiy the laws of joint
—
state of the conductor or its tempera
Means and methods of application. ageing [2].
—

ture. During the design the position

can anly be defined statistically and
1.1. Causes of the phenomenon the risk of flashover does not depend 1.2.4. Conductor creep
The problem is that of analyzing only on the electrical stress (operating
The research conducted and pu
quantities that determine the thermal voltage or overvoltage), but also on
blished by the Working Group regar
state of the conductor and to investi the probability of the conductor sag
ding the prablem of creep evaluation
gate the extent these vary, which relatively to the obstruction.
has shown that non-elastic conductor
quantities must be considered and Prevention in design may consist in elongation depends largely on the
which can be ignored from a practical fixing a safety clearance which takes temperature reached by the material
into account the statistical aspect and concerned. lndeed, creep is caused by
() Report on the activities of the WG 05 on makes the risk of flashover negligible. the settling down of the micro-struc
the Subject prepared on the request of the
SC 22 Chairman. Recent development adopted by ture in the metal and, since the state
() Mr. J. Bradbury (U.K.) .— Mr. E. Brandt’ some power line users is ta canti of the latter depends upon the tempe
(F.R.G.) Mr. B. Dalle (France) Mr. T.J.
rature, the amount af creep and, the
—

Hafstad (Norway) — Mr. T. lkeya (Japan) — Mr. nuously monitor the conductor tempe
D.T. Jones (U.S.A.) — Mr. A,R. McCulloch (Ca rature, during the operation of the line, refare, the planning of clearances,
nada) Mr. V.T. Morgan )Australia) Mr. P. have ta take into accaunt the thermal
—

ta prevent the clearance from drop

—

Nicolini )ltaly)
— Mr. E. Palazuelos Serto (Spain)
— Mr. P. Paoli (Italyl— Mr. P. Ralston (Canada). ping below pre-set limits. history of the conductor [3].
N’ 121 E LE CIRA 53

1.2.5. JouIe Iosses ding of the thermol state of the actual practical significance of the
conductor and associated environ- problem.
During the initial design of a une or
system, if an attempt is made to mental conditions may help in the
Part 2 reports on the results of this
optimize the conductor cross-section, prediction of ice load and consequen
survey.
tly a possible improvement in une
one of the chief components of the
relia bility.
cost objective function, which must be
minimized is Joule losses cost.
1.3. Practical importance 1.4. Means and methods of applica
In calculating the Joule losses, tion
generally, a single conductor tempera In reviewing the effects due to
ture is assumed, but a greater know conductor temperature, we may note Once the state of the art has been
ledge of the probable statistical rela that this parameter is one of great ascertained from the respondents, the
tionship between the conductor tem importance in the design and opera work of the Group will continue with
perature and the load carried out by tion of lines, therefore it cannot be the preparation of a practical guide
the conductor, a better evaluation of neglected. The Working Group has that, starting with the experiments and
the Joules losses may be obtained. therefore decided that it would be achievements of a large number of
worth having direct information on the researchers, will provide information,
1.2.6. Ice formation opinions of power line users, manu on the basis of which calculations or
Some quantitative evaluations have facturers and contractors. To this end, methods may be applied for a correct
already been made in relationship a survey on thermal problems was set evaluation of conductor temperatures
between the formation of ice over up among the members of SC 22, and under any condition.
loads, environmental conditions, and a questionnaire circulated with the
A brief outline of this work is
conductor temperature. It is possible, object of collecting information that
would make it possible to evaluate the presented in part 3.
in the future, that a better understan

PART 2

ANALYSIS 0F REPLIES TO THE QUESTIONNAIRE

ON THERMAL STATE

2. GENERAL 2.1. Analysis of the replies of section

TABLE I
A
The 1984 questionnaire was broken figures 1 to 4 summarize the replies
down into three sections: to the first four questions of the
. Nunber of
Geographical area questionnaire.
Section A rus pond en ts

The respondent was asked how the This was arranged such that the
Africa 1
problem of thermal rating was treated, respondents were asked to give a
for what conditions and in what terms YES/NO reply to a number of specific
rating was specified, what effects America 26 questions.
were taken into account, and lastly,
what means were used to evaluate Asia and Australia 12

those effects. 75
Europe 27
Section B 50

This section was the more practical 25

part of the survey, inasmuch as res
pondents were asked about the para Calcul
FRDM A
meters relating to the thermal rating CALCULATION
especially detailed information on SIANDAPI
for the various types of conductor D[TEMINlsTg FRAEA8II.ISDI
equations that the user normally em HETHID METHDD
used on mes and in respect of various
ploys to determine the parameters A partir Méthode Méthode
conditions that may be considered
involved in thermal rating. 66 replies dune déter- proba
(normal or exceptional operating norme ministe biliste
were received to the questionnaire,
conditions, short-circuit, etc..).
from a total of 16 countries.
Fig. 2. Approches différentes utilisées
Section C
—

These countries are listed by geo pour guider le calcul de la capacité ther
This section asked respondents for graphical area in Table I. This report mique.
any information considered useful but summarizes only the replies to part A Fig. 2. — Different approaches used to
not included in previous questions, and part B. govern thermal rating calculation.
N 121 E LE CIRA 55

The analysis given in this section 10

When analyzing these figures
was obtained by simply expressing the consideration must be made that the
75
number of positive replies as a per solar radiation and the maximum tem
centage of the number of respondents 50 perature depend, to some extent, to
repiying to the specific questions. 25
the geographical area of the respon
Many questions required multiple re o
50

plies, this explains why the sum of the DOLAI MAGNETIC 40

JOULE
percentages of a given question may EFF[tT RADIATION EFFECT
30
by higher than 100%. Effet Radia- Effet
Joule lion magné- 20
In some cases some respondents
solaire tique
did not reply to individual questions 10

and in this case when dealing with this 100 O

question the respondent was omitted. b)
75 SLMM[R PEUR SOLAR RADIATION
lt wiii be noted that ail the respon Radiation solaire de pointe en été
50
dents referred to the thermal rating of 100 100
conductors in normai conditions, a 25
high percentage of respondents aiso 0 75 75
referred to the temporary and emer CONVECTION RADIATION EVAP000TION
gency conditions, and even over 25 °h 53 50
of respondents also referred to fault Convec- Radia- Evapo.
lion lion ration
conditions (Fig. 1).
Fig. 6. — Pourcentage d’utilisatiQn des 2: 2:
Many different appoaches to the différentes sources de : e) gains de cha
problem are used, the most common leur, b) pertes de chaleur, dans les cal 0 0.2 0.4 0.6 Q8 1 0 0.2 0./, 0.6 0.8 1
being the deterministic calculation culs. ta NEW LONOUCÏDB
— —NEW CONDUCTOR
(Fig. 2). Fig. 6.— Percentage 0f use, in calculation, Conducteur neuf Es Conducteur neuf
Ea — —

of different sources of: a) heat gains.

Many respondents are governed in b) heat fosses. 100J 100
their caiculation by some forms of
standard (National Code, internai 75 75
conditions, figure 6 summarizes the
standard, conductor manufacturer
replies obtained figure 6a gives the
standard, etc..). In addition the ques 50 50
heat gains and figure 6b gives the heat
tionnaire reveaied (Fig. 3) that a high
losses. As expected the 100 % of res
proportion of the respondents suppor 2:_fl25
pondents consider Joule effect and
ted the Group activity for improving
convection and a lower percentage
the knowledge of the thermal state
consider solar radiation and conductor
under normal, temporary, emergency O 0.2 06 05 0.8 1 0 0.2 0.4 0.6 0.6 1
radiation. Ea—WNEAIHER[D LUNAUCIOR ts—WHEAONER[D CONDUCIOR
conditions (approximately 75% of the
respondents repiying in the affirma The questionnaire also asked ques Ea — Conducteur vieilli Es — Conducteur vieilli

tive). Even when considering thermal tions on the reference parameters and 50
state under fault conditions, just under coefficients which are used in the heat
1.0
50 % replies feit it necessary to consi balance equation, as follows
der this matter further. a) Peak value of solar radiation, 30

The reasons the respondents feit a b) Coefficients of surface absorp 20

detaiied knowledge is required on tion and emission, 10

thermal state are given in figure 4. The c) Wind speed,
main reasons were in ensuring clea d) Ambient temperature. o 0.2 0.4 QA 0.8 1 1.2 1.4 15 mIs
rances are maintained and for obtai TRANSVERSE WIND SPEOS

ning a better evaluation of the loss of Figure 7 shows in form of histo Vitesse de vent transversal
grams the results of the replies obtai
strength due to annealing and creep.
0f these replies the WG was surprised ned. Briefly, it may be noted that the
by the reply to the annealing which average values are:
was previously considered a less im — Summer solar radia
20
portant factor. tion 1000 W/m
2
With specific reference to the pro — Emission coefficient: 10

new conductor 0.4

blem of clearances, the situation here 0 10 20 30 40 50 60 70’C
wheathered conductor 0.6
is shown in figure 5 which indicates MAXIMUM SOMMER AMOIEXI IEMPERAIURE
Absorptivity coeffi
the relationship between the various Température ambiante maximale en éte
—

cient : new conductor 0.4

states of the system and une clearan Fig. 7. Distribution statistique des diffé
—

wheathered conductor 0.6 rents paramètres utilisés dans les équa

ces.
— Wind speed 0.6 m/s tions d’équilibre thermique.
Another group of replies concerns Maximum ambient Statistical distribution of diffe
Fig. 7.
—

the thermal balance equation used to temperature (sum refit parameters used in the heat balance
calculate temperature under normal mer) 35°C equations.
 N° 121 ELECTRA 57

dent. The WG was particularly surpri conductor temperature (ordinate), ing the replies relating ta section C of
sed when considering the low devia with reference ta normal operation the questionnaire.
tian in the maximum ambient tempe conditions and ta ACSR conductors of
As a first conclusion, however, t
rature. These values coincide whith lines above 220 kV.
may be said that the problem of the
those normally used in calculations
The central narrow band, includes thermal state of conductors has arou
employing the deterministic approach
about the 50% of the replies, the sed considerable interest, and, above
ta the probiem.
larger one the 90 %. ail, that a great deal of work is called
Figure 8 cortsiders the respondents for to coliate aIl the information avail
The replies concern lines located in
who collect meteorological data. This able, which it is hoped may be possi
different climatological areas; this fact
section of the questionnaire conclu ble in the later documents.
may explain the notable scattering of
ded by considerating the measure
the data because the quantities assu
ments of the conductor temperature
med for temperature calculation are,
under operating conditions. 130
normally, dependent on the location of
The first part, figure 8a, deals with the line. 120
the current practice of respondents 110
C-,
who monitor meteorological data, 2.3. Final considerations
while figure 8b deals with respondents
who monitor live conductors. The Working Group is now examin
90

C W
80
2.2. Analysis of the replies of section o o.
B D W 70
Fig. 9. Résumé des réponses concer
—

As regards Section B, which mainly nant la température admissible des 60

asks about boundary conditions, ex conducteurs alu-acier des lignes de ten
D
pressed as the temperature and cur sion supérieure à 220 kV. Chaque point o
D.
EZ
représente la densité du courant et la OC 0
rent admissible in conductors, more température admissible applicable sous
i—W

than 300 replies relating to as many des conditions normales de fonctionne 30

conductors have been summarized. ment.
20
By way of an initial processing Fig. 9. Summary of the replies cancer
—

10
result, the diagrams in figure 9 sum ning the admissible temperature of A CSR
conductors of unes above 220 kV: each
marizes the replies concerning the
point represents the current density and
current densities (abscissa) as a func the relevant admissible temperature in CURFCENT DENSITY 2
A/mm
tian of the admissible maximum normal operation. Densité de courant A/mm’

PART 3

METHODS AND MODELS

3. GENERAL following equation

This part presents a summary of heat stored = heat supplied heat lost
the work that WG is developing with in conductor to conductor by conductor
the aim of issuing a practical guide ta
the application of thermal rating which in mathematical terms, s the t = time (s)
concepts bath in design and in ope- equation
ration. The solution O = 0(t) expresses
CdO (P + M + F’ — —

the variability over a period of time of

Pr — Pjdt (1) conductor temperature.
3.1. Basic equations used for evalua
ting the heat transfer of an where
C = conductor thermal capacity The above terms depend, in gene
overhead conductor rai, on the temperature itself, on
(J/m °C)
Physical observation shows that meteorological quantities such as
O = conductor temperature (°C)
the environmental quantities (wind, wind, ambient temperature, etc., and
= heat gain due ta Joule effect
ambient temperature, solar radiation on the electrical ioad. t is evident that
(W/m)
etc.) vary over a period of time, and calculation of O (t) is mathematically
= heat gain due to magnetic
often the load transmitted. For these effect (W/m) complex and would be possible only
reasons the thermal state of the if the “time series” of the aforemen
= heat gain due ta solar radia
conductor varies continuously and so, tioned quantities were available.
tion (W/m)
consequently, does its temperature For the sake of simplicity t is
P, = heat loss by convection (W/m)
[4]. convenient ta consider three different
= heat loss by radiation (W/m)
ln general, the thermal behaviour P = heat loss by evaporation situations in respect of the boundary
of the conductor is represented by the (W/m) conditions of equation (1):
 N° 121 ELECTRA 59

a) Steady state case, the total heat supplied is due is expressed by the following simpli
only ta the Joule effect, and there is fied relationship if, for the sake of
If the meteorological parameters
na heat exchange with the environ- simplicity, the reflectance effect is
are assumed ta be constant and the
ment: this heat must therefore have ignored and it is assumed that = O
electrical load does flot change, the
been stored in the conductor. The
heat supplied to the conductor is P = e D [I sin + it/2 ‘d j (6)
equation expressing this fact is the
constant in time and it is balanced by
following The intensity of the direct beam l
an equal constant in respect of heat
dissipation. This condition is defined varies with the air-mass traversed, and
hence with the solar altitude H , and
5
as steady-state and, in practice, s
assumed to be true even when the
cd9
= Ç P(t) R(9) dt
(4) with the turbidity of the atmosphere.
parameters involved vary little around where 1(t) is the current flowing during The turbidity is due ta the attenua
an average constant value. the short circuit, R (9)is the conductor tian of the solar beam as the resuit of
In this case CdO = 0, and therefore resistance and t is the duration of the scattering by air molecules, dust parti
equation (1) is simplified as follows short circuit [7]. The factors included des and water vapour, and absorption
in equations (1), (2), (3), (4) wilI be by ozone.
P + S + M = 00
P0 Pr + ‘‘v (2) examined in the items that follow. In still air, solar heating of a
that is, the well-known “balance equa 3.1.1. JouIe heating conductor s significant, even on
tion”. cloudy days. In windy conditions, the
The heat generated by the Joule relative importance of solar heating
This is the characteristic situation effect is expressed by:
during normal operation or, during decreases as the wind velocity increa
temporary operation, after the tran = k /2
0 (1 + co 7)
R (5) ses. Although the coincidence of high
sient due ta the load change has been solar heating with low convective
where cooling is important, hardly any field
normalized.
I = effective (r.m.s.) current flow tests have been carried out ta carre
b) Transient state ing through the conductor (A) late these twa parameters on a proba
When the electrical load varies 0
R = electrical resistance (d.c.) per bilistic basis. The solar absorptivity of
suddenly with constant, or practically unit length at 0 oC (/m) weathered conductors is greater than
constant, meteorological parameters, 7, = average conductor tempera 0.5 and may approach unity.
the terms P., P, in equation (1) ture °C
depend only on the temperature 9, and = temperature coefficient of re 3.1.3. Magnetic heating
5 and Pv may be assu
the terms P,,, P sistance between O OC and 7,
(oc)-’ With alternating current, the cyclic
med ta be constant.
k = coefficient, taking into account magnetic flux causes heating by eddy
In this case, the following expres currents, hysteresis and magnetic vis
skin effect (with d.c. k = 1)
sion can be used; cosity. The eddy-current heating P. is
The resistance temperature coeffi usually negligible with non-ferrous
9(t) =
cient, con, for steel cored conductors canductors at power frequency, but

ç’ R ( I±() + PM + Ps “ depends an the ratio of cross-section may be significant with steel-cored

dt
—

area of the non-ferrous wires to that conductors, because a longitudinal

(3) of the steel wires, and the coefficient magnetic flux is produced in the steel
K, for conductors in common use, wires by the current in the non-ferrous
which can be solved by numerical varies from 1.0 ta 1.05 [8]. wires spiralling around the steel core.
integration [5]. This equation repre
sents the thermal state of the conduc It is worth noting that skin effect is The heat gain, P0 due ta eddy
tor during temporary and emergency usually negligible for steady-state ra currents in the steel wires, s a func
operation if there is no marked varia tings. tian of the magnetic induction in the
tion in atmospheric conditions. steel wires due ta the magnetizing
force, and s also a function of the
c) Unsteady state 3. 1.2. Solar heating
physical characteristics of the conduc
This condition must be considered The solar heat gain P depends on tor [10].
when variation in the meteorological [9] : the diameter D of the conductor
The hysteresis heat gain, PH, of the
parameters, especially wind intensity and its inclination ta the horizontal;
steel wires due to the alternating
and direction cannot be kept within a the absorptivity c5 of the surface of
magnetic field depends on the magne-
Iimited range. AIl the terms of equa the conductor; the intensity / of the
tic induction, the conductor characte
tian (1) depend, in general, both on direct solar beam on a surface normal
of the ristics, and the chemical composition
temperature 9 and time t, sa that only ta the beam; the intensity ,

of the steel.
the solution of the general differential diffuse sky radiation in relation ta a
equation (1) has ta be considered [6]. horizontal surface; the solar altitude The ferramagnetic heat gain is PM
H; the angle ri of the solar beam with = P. + PH, because the heating due
d) Adiabatic state respect ta the axis of the canductor; ta magnetic viscasity is negligible. t
During a short circuit, the current and the albedo (reflectance) F of the is also possible ta include the ferra
flawing in that very short space of surface of the ground beneath the magnetic heating term PM with the
time produces an adiabatic thermal conductor. The total solar heat recei Joule heating term P. in equation (5),
response of the conductor. In this ved per unit length of the conductor , taking
by redefining the coefficient k
2
 N° 121 E LE CIRA 61

into account both skin and magnetic In stili air, convection cooling de The models are based on different
effects. creases as the orientation of the philosophies
Magnetic heating in steel-cored conductor to the horizontal increases. — One modelling method is based on
conductors produces a small increase In wind, convection cooling increases fixed conventional terms of refe
in the effective resistance with increa with wind velocity and turbulence in rence and is the welI-known and
sing current when there is an even tensity; stranding increases the heat very widely used deterministic
number of layers of non ferrous wires transfer by approximately 10 per cent method.
with alternately reversed directions of for 100 < (Re)Df 2650, and by 50
per cent at (Re)Df = l0 for crossflow. — Two modelling methods take into
lay. With an odd number of layers, the account the statistical nature of
increase in resistance is greatest for a Forced convection cooling decreases
with the angle of attack of the wind in the relevant parameters appearing
single layer; for three layers, the resis in the thermal equation of the
tance increases by 5-10% at full rated relation to the conductor axis. The
orientation of the forced flow in rela conductor. They are known as the
load current. statistical parametric method and
tion to the natural flow is important.
Typical ranges of variation of k for the real-time parametric method.
conductors having various layers of Two other modelling methods are
3.1.5. Radiative cooling —

non-ferrous wires, at full-load current based on direct observation of an

are Because the radiation loss is
usually only a small fraction of the exposed conductor, either of a line
k = 1,15 for 1 layer of non-ferrous total heat loss, particularly with forced or of a model, monitored conti
wires, convection, it s sufficiently accurate nuously. These are called the sta
k = 1,01 1,04 for 2 layers,
—

to write tistical con ductor- temperature me

k = 1,05 1,10 for 3 layers.
— thod and the real-time conductor
P, = 5 + 273)
itDa[(T tempera ture method.
Because magnetic induction de-
pends on the current flowing through —
(7; + 273)9 (8)
the aluminium (or alloy) wires, the
magnetic heating becomes important 3.2. 1. The deterministic method
where B is the Stefan-Boltzmann
in temporary or emergency conditions. constant (5.67 x 10.8 W/m
2 K ); T
4 , 7
5 In the deterministic model repre
are the temperatures of the conductor senting the thermal state of the
3.1.4. Con vective cooling conductor, the maximum permissible
surface and the ambient respectively,
and E 5 the effective emissivity of the temperature is fixed and is determined
During zero-wind-speed or still-air
conductor. In still air, the radiation by the need to limit mechanical dete
conditions, the density of the air near
heat loss is usually less than one third rioration, the loss of clearance, and
the heated conductor is less than that
of the total heat Ioss. With wind, this creep of the conductors.
of the rest of the air; this fact produ
ces buoyancy forces, thus causing an fraction falls to a much lower value; This temperature, and chosen value
upward flow of the air around the hence, the value chosen for the total for the atmospheric parameters and
conductor surface. This process is emissivity of the conductor is not the emissivity and absorptivity of the
called natural or free convection. critical. conductor surface, are inserted in the
appropriate heat equation to obtain
On the other hand, forced convec 3.1.6. Evaporative cooling
the maximum permissible current. The
tion occurs when the wind blows on The presence of water vapour in wind is usually assumed to be non-
the heated conductor. The heat trans the air flowing over the conductor has turbulent and normal to the axis of the
fer in this condition is a function of a negligible effect on heat transfer. conductor.
many variables, the primary ones
being the speed, its direction and the When water droplets are entrained The determinist method is clear-cut
intensity of turbulence. in the flow, they have little effect as and only requires computing facilities,
long as the conductor remains dry; once the values of the various para
Others factors affecting the phe but, when the surface of the conduc meters have been chosen; hence, it s
nomenon are conductor roughness, tor is wetted, the heat transfer increa the method usually used for system
and the arrangement of phases (single ses considerably due to evaporation design, and it may be used for control
or bundled conductors) [11]. from the surface, particularly the front ling operation.
The general expression of heat loss surface where the droplets impinge.
This model is applied to control of
by convection is
any type of operation, because it does
P,,,, = it (7; —
7;) (Nu)D (7) 3.2. Analysis of methods and models not take into consideration variation,
where (Nu)Df is the Nusselt number of solutions in the parameters involved in the tem
based on the diameter D, and is equal Knowledge of the thermal beha perature rise.
000 D/X, ,,
to h . 00 being the convective
h viour of an overhead conductor can be However, it is fairly inflexible, and
heat transfer coefficient and the acquired by applying different me results in rather conservative ratings,
thermal conductivity of the air. The thod, and modelling of the phenome because of the Iow probability of
subscript f signifies that the value of non, depending on the purpose (de coincidence of the value usually cho
the relevant quantity is based on film sign or operation), and on the condi sen for the wind velocity and the air
temperature T, equal to 1/2 (7; + 7;), tion to which the analysis is referred temperature; for example, U =
T and 7; being the surface and am (normal, temporary, emergency, fault 0.6 m/s and 7; = 35 °C. Some impro
bient temperatures, respectively. condition). vement s achieved by time zoning
 N°121 E LE CT RA 63

that s by using different values for /,

7; and U for different parts of the day
(e.g., Is = O at night) and in different
seasons (e.g., 7; = 10°C in winter, and
7 = 35°C in summer).
The potential increase in rating as
a result of wind turbulence is neyer
considered, and the decrease in rating
when the wind is in the same direction
as the une is usually either ignored, or
allowed for to some extent by adop
ting a very low crossflow velocity. This
method s clearly of limited use to
system operators.
Chaleur stockée
Chaleur fournie
3.2.2. The statistical parametric me Chaleur perdue
thod
The statistical parametric method
takes into account the statistical varia
tion of the parameters involved in
conductor heating that is, their pro
bability distributions [12].
Referring to figure 10, which gives Fig. 10.
a time sample of the quantities in
fluencing the temperature of the
conductor, the assumption is made variations in the meteorological phe gical quantities may be correlated
that each of them could be represen nomena, which means that the chosen with the electrical Ioad). Methods of
ted by a series of discrete constant
interval must no be too long either. calculation may or may not take into
calculated account this aspect.
values, at intervals of t, the magni (b) The temperature
tude of the quantity in that interval should be an accurate estimate of the Specifically within the framework
being equal to the mean value of the actual average temperature of the of the above-mentioned assumptions,
quantity itsef. conductor during the interval in ques the use of Monte Carlo calculation
Therefore, in the i’’ interval of time tion. The appropriate choice would be method requires these steps
Ai, the relevant parameters can be an interval /xt of one hour. — Random selection from the statis
characterized by the group of values If the above basic assumption is tical distributions of the set values
(U, 7;, I, I, X
1 X) which are the accepted, the time trends of the va- of the variables U, 7;, I, I, X X,,
mean values of the following quanti nous quantities may be discretized taking account possible correlation
ties during that interval into intervals of one hour, during between them.
— Wind speed, U, which each of those quantities may be — Determination of the hourly ave-
— Ambient temperature, 7, considered constant and equal to its rage conductor temperature by
average hourly value. means of the heat equation.
— Solar radiation, 1
s,
— Current, I. This is equivalent to making calcu — Repetition of this process for a
— 1
X X other parameters pre lations whith the temperature in the sufficient number of times by ob
viously described. steady-state, and therefore the tem taining the simulated distribution
perature at each interval is irrespec of conductor temperature.
In generating this statistical model, tive of the temperature reached during
the problem is that of determining the the previous intervals. At his point, Clearly, this method of calculation
interval of time Aisuch that the actual there is no longer any interest in the approximates to the temperature sta
mean value of the temperature in the time series of the average values, but tistical distribution around its mean
conductor during the interval is equal only in the frequency of their appea value, while the tails of distribution
to the value calculated on the assump rance; that is, their statistical distri must be evaluated by using methods
tion of steady-state conditions, cha bution. that take into account temporary and
racterized by the constant values (U, emergency conditions.
T, ‘‘ I, X X;,). In general, each parameter invol
ved in the thermal state of the conduc This means that the method can
This interval is chosen by conside tor may be correlated with other profitably be applied to me design,
ring the following requirements parameters. This correlation depends because knowledge of the tempera
(a) The average value of each para on the nature of the parameters (me ture distribution is obtained merely by
meter must be ‘stationary”, i.e. it teorological data may be more or less starting from general information on
must flot be affected by fluctuations strictly correlated with each other) meteorological quantities and forecast
around the mean value; therefore, the and on line operation (peak Ioads electrical load and does not require an
chosen interval must not be too short, occur during given periods of the day assumption on the possible or actual
nor must t be affected by longer-term and the year, and relevant meteorolo quantities in operation. On the basis of
N°121 ELECTRA 65

this distribution, the system designer racteristics, as in the case of the line An alternaltive ta performing mea
can assess the risk of exceeding the to which the model s applied. surements on an energized line s to
maximum permissible temperature of As far as operation is concerned, use a thermal model of the conductor
conductors, or estimate the probable bearing in mmd that the method at Iow voltage. The model consists of
loss of tensile strength of the conduc would be unlikely to predict actual a cylinder that is long enough ta
tors from their thermal history. The conductor temperature with any pre enable axial heat flow ta be ignored,
statistical parametric method could, cision, because the data are usually and has the same diameter, surface
on the contrary, be used by system properties and thermal time-constant
only available at too long, (say hourly),
operators with the same limitations as as the me conductor. The current
intervals, its application is reliable only
the deterministic method, the only in normal operation, while, in tempo passed through the model should
difference being that the former pro- rary or emergency operation, cannot produce the same Joule heat as that
vides knowledge of the probability in the me conductar. The temperature
supply any information on the dyna
that a given limiting temperature may of the model should be measured at
mic aspect of the termal behaviour of
or may flot be reached. No information the conductor. its mid-point. An advantage of the
on the actual conductor temperature thermal model several samples can be
can be obtained from this model. placed with different orientations in
3.2.4. Statistical conductor-tempera the same place; or else one model can
ture method be ratated to various azimuths.
3.2.3. ReaI-time parametric method
In this method, a conductor car A third and pramising solution is
In this method, an appropriate time rying a constant current at low voltage represented by the purely analytical
interval t is selected, as in the statisti is exposed outdoors for a period of at thermal models that can predict the
cal parametric method. The average, least one year, and the conductor temperature af an overhead conduc
calculated in the interval A t, of the temperature and the air temperature tor. The system involves cailecting
actual measured values of the metea are monitored. The distribution of the data an the actual metearological
rological and load parameters (or temperature rise of the conductor is a quantities at the lacation of the une
stored data from previous measure function of the distribution of the these data are then entered in a
ments) are inserted in the heat equa meteorological parameters. This pro computer program that, provided the
tion to determine the calculated value cedure can be repeated for other line characteristics and electrical load
of the conductor temperature at that values in respect of current. through dispatching are known, pre
specific time.
The method requires to continuous dicts actual line temperature at pre
The distribution of the conductor monitoring the sample, and its appli selected locations (i.e. particularly
temperature can then be determined, cation s meinly for design, because its critical spans versus clearances...) in
after repeating this process a suffi use in operation is similar to that of many cases, the three alternatives are
cient number of times ta include data previous models. combined and used to contrai a whole
from ail four seasons. The advantage network.
of this method lies in the fact that the 3.2.5. Real-time conductor tempera At present, these systems are
meteorological quantities taken as ture method knawn as Dynamic thermal une rating.
above are automatically correlated
In this method, the temperature of The real-time conductor-tempera
with each other and the temperature one or more of the conductors of an ture method takes into accaunt ail
evaluated approximates ta the real energized high-voitage line is measu meteorologicai and Ioad fluctuations.
average coriductor temperature in the red, either continuously or at short From the system design point af view,
interval A t. The parametric method intervais, by means of a transducer, the methad wiiI provide temperature
requires access to electrical load pat and the signal is transmitted ta ground distribution curves that enable the
terns and meteorological data measu by means of an optical or telemetry probability, or risk, of exceeding a
red at fairly frequent intervals over a Iink. Because the actual temperature given temperature to be assessed.
period of at least one year, and a of the conductor is measured, this
computer ta solve the equation repea Moreover, the real-time conduc
method is the most appropriate one
tedly, to obtain conductor temperature tor-temperature method gives ta the
for the une operator to use; however,
distribution. system control engineer a continuaus
since the conductor has a fixed azimu
indication af the temperature of the
thal direction, the temperature of
The resultant temperature distribu conductor, so that he can handie load
conductors of other azimuths should
tion wiIl be an approximation, even if dispatching, particularly in emergency
also be monitored, in order to ensure
the wind direction and turbulence conditions, confident that there
that the worst combination of meteo
have been measured, because no should be no risk of overheating the
rological parameters is obtained. if the
aceaunt has been taken of random conductor.
me is long, it is also advisable to
factars such as rainfall, and the cup
measure the temperature at other
anemometers used at meteorological
points along the une. This method
stations rarely record velocities lower
enables the thermal history of the
than about 1 m/s. References
conductors to be recorded : this s
The statistical information obtained useful for estimating the loss of [1] V.T. Morgan The Ioss of tensile
by collecting the results of the model strength of the conductor due to strength of hard-drawn conduc
may be used in designing unes having annealing and the progressive creep tors by annealing in service, IEEE
similar structural and operation cha ot the conductor. Trans. PAS Vol. 98 (1979).
N° 121 ELECTRA 67

[2] B. Dalle Etude théorique et ex femerits et des effects électrody [10] R.D. Findlay and H. Riaz A non
périmentale de la capacité de namiques. Techniques de l’ingé linear diffusion model for the
vieillissement des raccords de nieur DG73 — 12— 1978. electromagnetic field configura
conducteurs nus des lignes aé tion of single layer ACSR, Proc.
riennes; Technical Memorandum [6] S.D. Foss, S.H. Lin, R.A. Fernan
IEEE Power Engineering Society
dez : Dynamic Thermal Line Ra
EdF HM/72-04566/BD/MC, pre Winter Meeting, New York, NY,
tings Part I Dynamic am
sented by WGO5 SC 22 3 February 1978,
— —

— —
29 January —

pacity rating algorithm. IEEE

CIGRE. Paper A 78 152-1.
Trans. PAS — Vol. PAS-102, N°6,
[3] CIGRE : Permanent elongation of June 1983.
conductors. Predictor equations [11] V.T. Morgan The heat transfer
and evaluation methods. CIGRE [7] V.T. Morgan Ratings of conduc
from bare stranded conductors
Study Committee No. 22, Paper tors for short-duration currents.
by natural and forced convection
22-78 (WGO5) 02, 1978, Electra, Proc. IEE, Vol. 118, No. 3/4,
in air. Int. J. Heat Mass Transfer,
No. 75 (1981), pp. 63-98. March/April 1971.
V. 16 (1973), pp. 2023-2034.
[4] V.T. Morgan The current-car [8] H.B. Dwight : Skin effect and
rying capacity of bare overhead proximity effect in tubular
[12] G. Di Giacomo, P. Nicolini and P.
conductors. Institution of Engi conductors, AIEE Trans. Vol. 41
Paoli Criteria for the statistical
neers, Australia, Electrical Engi (1922) pp. 203-209.
evaluation of the temperature in
neering Transactions, Vol. EE4,
[9] R. King and R.O. Buckius Direct conductors. Thermal rating pro
(1968), pp. 63-72.
solar transmittance for a clear blems, CIGRE Study Committee
[5] P. Johannet and B. Dalle Calcul sky, Solar energy, Vol. 22 (1979), No. 22 Colloquium, Sienna, 1979,
de chutes de tension, des échauf pp. 297-301. Paper 3.18.