CURRENT RATING OF CABLES INSTALLED IN DEEP OR VENTILATED TUNNELS

Eric DORISON - EDF R&D, Moret-sur-Loing (France) eric.dorison@edf.fr

George ANDERS - Technical University of Lodz, george.anders@attglobal.net

ABSTRACT Generally, the thermal resistance T4 of the surroundings

of a cable is defined by:
The series of IEC 60287 standards provides methods for
calculating the permissible current rating of cables. Wt
T4  [2]
This paper deals with the present work carried out within s  a
the IEC TC 20 WG 19, intending to extend the scope of
these standards, to groups of cables installed in deep or So that the thermal resistance T4 for a cable in air is:
ventilated tunnels.
1
T4   s   s   a [3]
 .De .h. s0,25
INTRODUCTION
The heat dissipation coefficient h is given as a function of
The IEC 60287-1-1 [1] and 2-1 [2] standards address the the cable diameter:
thermal rating of a cable circuit installed in still air with
Z
given temperature with the IEC 60287-2-2 [3] providing h E [4]
an extension to some homogeneous groups of cables. Deg
The same calculations can be used to rate cables in
tunnels, however, the external thermal resistance of the where Z, E, g are constants, whose values depend on the
tunnel itself requires special considerations.. type of installation.
The paper deals with the work carried out by WG19 of As T4 is a function of s, an iterative process has to be
the Technical Committee 20 of the IEC intended to extend conducted, taking into account the temperature drop
the scope of these standards to cables installed in between the cable conductor(s) and its surface:
ventilated deep tunnels.
 c  s  n.Wd .Td  Tint   Wt .Tint [5]
The IEC standards combine the effect of heat transfer by
radiation and convection into one coefficient. In order to c is the conductor(s) temperature, n is the number of
properly model the effect of air movement inside the conductors, Wd represents the dielectric losses per
tunnel, the convective and radiative heat transfers must conductor and Td and Tint are the equivalent thermal
be treated separately. Hence, first, the IEC method for resistances used in expressing the transfer of dielectric
rating cables installed in still air is reviewed and losses and Joule losses within the cable, respectively:
considerations are given to the modeling of heat transfer
by radiation. In particular, the radiative heat transfer for a T1
group of cables is addressed and the extension of the IEC Td   T2  T3
2.n [6]
method to the groups of cables with different designs is
T 
. 1  1  1 .T2  1  1  2 .T3 
1
proposed considering the effect of dielectric losses. Tint 
1  1  2  n 
Next, a rating method for cables in ventilated tunnels is
presented, based on an analytical approach, originally The permissible current I is obtained from Wt as follows:
developed by CIGRE and published in Electra [4,5].
Finally, for deep tunnels, a fictitious equivalent depth is
Wt  n.Wc .1  1   2   Wd  [7]
introduced, to optimize cable rating, taking into account
the soil thermal inertia, without performing transient Wc  R.I ² [8]
analysis.
R being the conductor resistance at the temperature c.
BASIC MODEL FOR CABLES IN AIR For installations involving several circuits, the value of the
heat dissipation coefficient is defined in IEC 60287-2-2
According to the IEC 60287 standards, the rating of air from the value for a single circuit given in IEC 60287-2-1.
installed cables, , is based on a relationship that links the
total heat loss of a cable Wt with the temperature rise of This approach is based on modelling of the heat transfers
its surface s above the ambient a by: by radiation and convection, using Ohm’s thermal law,
linking temperature drop and heat rate through a thermal
Wt   .De .h. s   a q [1] resistance with the following assumptions :
 The construction of all cables in the tunnel is the
where De is the cable diameter, h is a heat dissipation
same. This means that the surface temperatures of
coefficient depending on the installation and q is a
the cables are similar; and, therefore, the radiative
constant, set equal to 1,25.
heat transfers between the cables may be ignored.
 Cable losses
Radiation cable / tunnel walls
n.Wc n.Wd n.1.Wc n.2.Wc

c s Wrad t
T1/2n T1/2n T2 T3 Tst Te

Wk Ww
Tsa Tat

Convection cable /air Convection air / walls

Wconv
a

Figure 1 : Heat transfer modeling – cables in still air

Tst 
1
 .De .hrad
 
hrad  K r . s . .  s  2732   w  2732 . s  273   w  273 Kr 
Fsw
1   s .Fsw   s
Tij 
1
 .Dei .hij
   
hij  Fij . s . .  si  2732   sj  273 2 .  si  273   sj  273  

F12 
1 
2 

.  c 2  r  12   c
0, 5 2
 r  12 
0, 5  r 1 
 r  1. cos 1 
 c 
1  r  1  
  r  1. cos 
 c 
 r
De 2
De1
c  2.
S
De1

Table 1 : Radiative thermal resistances

 The temperature of the tunnel walls t is close to the Aw represent the cable and wall surface areas, s and w
temperature of the air a in the tunnel. are the reflectivity of the cable outside surface; and the
 The air in the tunnel is still, so that the convective wall inner surface, respectively.
transfers can be described through the well known and
When several cables are present, the mutual radiant area
experimentally checked correlations.
between them must be subtracted from the area radiating
In reality, the heat transfer from the cable surface is to the tunnel inner surface. The effective radiating area to
composed of several components as illustrated in Fig. 1. the tunnel walls is obtained as follows, using Hotel’s
Referring to this figure, Tst represents the radiative heat method of stretched bands [6].
transfer between the cable surface and the tunnel walls,
Tsa and Tat represent the convective heat transfer from the Elastic bands are imagined stretched around two arbitrary
cable surface to the air in the tunnel and from the air to concave surfaces (see Fig. 2).
the tunnel wall, respectively.
D
The thermal resistance T4 is a combination of the thermal C 2 Length of the internal band :
resistances expressing heat transfers by radiation and
C'
Iint  AD  BC' C
convection from the cable surface. The thermal resistance
representing heat transfer between the air and the tunnel where C' is the point of
wall is assumed to be negligible. intersection of BC with
surface CD.
RADIATIVE HEAT TRANSFERS Length of the external band:
1
The expression for the thermal resistance Tst is given in A B
Table 1, where  is the Stefan-Boltzman constant; s is
Iext  AC  BD .
the cable surface emissivity, Kr is the radiation shape
factor, which depends on the view factor Fsw defined as a Figure 2 : Calculation of the mutual radiation area [6]
fraction of the radiation that leaves the considered cable
and is intercepted by the tunnel walls. The mutual radiation area Am per unit axial length is equal
to half the difference in lengths of internal and external
K r  1   s /  s  1 / Fsw  As w / Aw w 1 [9] bands, that is,
Iint  Iext AD  BC' C AC  BD
where w is the emissivity of the wall inner surface; As and Am    [10]
2 2 2
Application of equation [10] can become quite From the relationship [2] and the definition of T4 :
cumbersome. In practice, as the tunnel size is large
compared to the cable surface, and as s and Fsw are T4 n h0  s00, 25  s 0 Wt 0 T40 1  k0
 .  .   k0
close to one, the following expression given in [7] may be T40 hn  n0, 25  sn Wtn T4 n T4 n
used:
T40
K r  1   s .1  Fsw  [11]
It is found that T4n/T40 is a solution of
Fsw and corresponding Kr (for s =0.9) are given in Table 2 0,25
for installations described in the IEC 60287-2-1 standard.  
T4n h0 1  k 
 . 0 k 
0
Installation Fsw Kr T40 hn  T4 n 
 
Single cable 1,000 1,000  40
T 
2 cables touching 0,818 0,833
The derating factor is derived from the relationship given
2 cables - spacing 2.De 0,919 0,927 below, which reduces to the IEC formula if the dielectric
0,636 0,660 losses are small compared to the total losses.
3 cables touching – middle cable
3 cables spacing 2.De – middle cable 0,838 0,852 Wtn n.Wd

Trefoil 0,652 0,676 I tn2 Wtn  n.Wd Wt 0 Wt 0 W
 .  tn
I t20 Wt 0  n.Wd n W
. Wt 0
1 d

Table 2 : View factor and radiation shape factor Wt 0

Table 1 gives the formula for the thermal resistance Tij Cables with different designs
expressing the radiative transfers between two cables If there are cables with different designs, the radiative
[8,9], which is of interest when the rating cables with heat transfers between the cables have to be taken into
different designs or in ventilated tunnels are sought. account [10]. This is achieved by computing first the
thermal resistances involving the view-factor between
GROUPS OF CABLES cables.
Cables with the same design Basic equations are as follows:
The IEC standard 60287-2-2 extends the calculation
method in the IEC 60287-2-1 to some homogeneous  ci*   si  Wi .Tint_ i
groups of cables. The formulae do not consider the effect
 

of the dielectric losses. The following developments
extend the standard method by considering these losses.  si   a  Wi  Wij .Text _ i [13]
 
Let  c* and  c* are defined as follows :  j i 
 si   sj W ij .Tij
   c  n.Wd .Td  Tint 
*
c      a
*
c
*
c
where Text_I is the thermal resistance between the cable
Using the relationships [5] and [2],  c* is expressed as: surface and the tunnel walls, representing a combination
of the radiative and the convective thermal resistances,
 c*  Wt .Tint  T4 
respectively Tst andTas with the former one defined in
Table 1.
For given air temperature a and the maximum Tas is determined as:
permissible conductor temperature c , as Td and Tint
depend only on the parameters internal to the cable,  c* 1
Tas 
is the same whether the cable is alone or is a part of a  .De .hconv [14]
group. Consequently, defining Wt0 and Wtn as the losses hconv  h. s   a 0, 25  hrad _ 0
for a cable alone or a part of a group, respectively, we
have: where h is the global heat dissipation coefficient given in
Wt 0 Tint  T4 n IEC 60287 and hrad_0 is the radiative heat transfer
 [12] coefficient for an isolated cable.
Wtn Tint  T40
Substituting Wi and Wij in the second equation in [13],
Iintroducing the ratio k0 of the cable surface temperature from the first and third one respectively, leads to a set of
rise to the conductor temperature rise due to Joule losses equations linking the cables surface temperatures:
in the metal components only, i.e., neglecting dielectric
losses, we can write equation [12] as:.    sj a  ci*
 1
 T
1 1
 si .     
 1  k 0   k 0 . 4 n
Wt 0 T T40 T
 ext _ i T Tij  Text _ i Tint_ i
k0  int_ i j i  j i ij
Wtn T40 Tint  T40
Once the surface temperatures are determined, the
permissible heat rate of the cable i is derived using the 240 mm² Cu 50 mm² Al
first equation in [13], and, thus, the ampacity is obtained Cables 3-core SA type 63 kV 20 kV
from relationships [7] and [8]. Outer diameter (mm) 94.3 53.6
As the radiative thermal resistances between the cables Tint (K.m/W) 0.189 0.240
depend on the cables surface temperatures, there is a Each cable alone
need for an iterative process. Text (K.m/W) 0.293 0.475
Following table gives sample results for 2 different 3-core Surface temperature (°C) 66.5 69.9
cables, installed horizontally with a spacing equal to 1,75 IEC Ampacity (A) 637.6 186.2
times the mean diameter.
S = 1.75.(De1+De2)/2
The results in Table 2 confirm the assumptions in the IEC Text (K.m/W) 0.293 0.492
60287-2-2 that, for the clearances between cables, as
Surface temperature (°C) 68.1 68.1
given in Table 1 of this standard, the thermal proximity
effects are negligible. Ampacity (A) 633.9 182.2
S = 100000 m
This approach is quite correct if the clearance between
the cables is large enough so that the proximity does not Text (K.m/W) 0.286 0.460
lead to significant disturbance in the convection transfer. Surface temperature (°C) 66.2 69.4
If not, it is suggested to determine the convective thermal Ampacity (A) 638.3 186.4
resistance using for the global dissipation coefficient h in
[14], the IEC value for a cable clipped to a wall.
Table 2. Sample results of two different cable ratings
installed in a tunnel.

STILL AIR MODELLING

 s   e  Ts . N .W k
STAR – DELTA TRANSFORMATION

 e   t  Tt . N .W k  W a 
s t
T3 Te
 t   0  Te . N .W k  W a 
Ts Tt
(N) e
 a   e  Ta .W a T3
N.Wk

Ta Heat removed
 a by the air
W a  C av .
z Wa
a

Figure 3 : Model for heat transfer in a ventilated tunnel

1 V .De
VENTILATED TUNNELS Tsa  with Re 
 .k t .K p . Re 0.65 
In the case of ventilated tunnels, the same models as for
still air apply, but the convective heat transfer coefficient is V is the bulk air velocity,  is the kinematic viscosity and kt
computed taking into account the effect of the air is the thermal conductivity of the air. Kp is an
movement. In addition, the heat removed by the flowing experimentally determined constant with the values given
air has to be taken into account in the energy balance in Table 3 taken from reference [11].
equations.
Figure 3 presents a heat transfer model and gives the Cable Arrangement Kp
corresponding set of equations, obtained from the initial Single Cable 0.130
diagram, using a star-delta transformation. The tunnel 3 cables spaced horizontally (spacing>2Diameter) 0.115
axis is assumed to be perpendicular to the page surface 3 cables touching horizontally 0.086
(z coordinate). 3 cables spaced vertically (spacing>2Diameter) 0.115
3 cables touching vertically (spacing<=2Diameter) 0.086
Convective transfer between the cables and 3 cables touching in trefoil 0.070
air
The following formula for the thermal resistance Tsa
Table 3 : Experimental constant for cables / air convection
applies if the Reynolds number Re is larger than 2000.
Convective transfer between air and walls. 
1
 T1  2
The Dittus-Boelter correlation, in agreement with the  max   0   0   Wd .  n.T2  T3  T4t  
Kitagawa works, is used in the formula of the thermal I  2 
resistance Tat.  R.T1  n.1  1 .T2  n.1  1   2 
. T3  T4t  
 
This applies if the air flow is turbulent, i.e. if the Reynolds  
number is larger than 2500. If the Reynolds number is
smaller than 2500, this thermal resistance may be with:
considered negligible .
Tt  Te
 0   a 0    0 .
L
1 .e 
Tat  Ta  Tt  Te
 .k t .0,023. Re 0,8 . Pr 0,4
   Tt  Te 
T4t  N .Ts  Tt  Te .1 
V . Dt L
with Pr  C pair . Re  .e  

 Ta  Tt  Te
kt
 
Cpair is the specific heat of the air per unit volume.

  Ta  Tt  Te .C av
This formula does not take into account the wall
roughness.

Conductive heat transfer in the tunnel

surroundings Temperature distribution
The thermal resistance Te is given for circular shallow
The air temperature a(L) at tunnel outlet is estimated as :
tunnels as:

Te 

2

.Ln u  u ²  1  u
2.L
Dt
[15] 
 a L    a 0    0  Tt  Te .N .Wk   a 0 .1  e

L



 
where  is the soil thermal resistivity, L is the depth of The cable surface temperature and the tunnel wall
tunnel axis and Dt is the tunnel diameter. temperature at the tunnel outlet are derived from the air
temperature as :
.For a square tunnel, the expression derived by
Goldenberg for a buried square trough and reported by  s L    a L   Ta .Wa L   Ts .N .Wk
Symm (1969) can be used. This expression is
  L
Te  ln  3.388   t L    a L   Ta .Wa L   Tt .N .Wk  Wa L 
2  a
where a is the height and width of a square cable tunnel,
where Wa(L) is the heat removed by the air at tunnel
Heat removed by the air outlet, given by :

Tt  Te .N .Wk   a L    0 

The heat removed by the air is linked to the air
temperature variations according to:
 a
Wa  C av . C av  C pair .V . At
z
where At is the tunnel cross-section.
Permissible current
CIGRE works show that, for typical installations, air
properties may be considered constant along the tunnel
route and computed using air temperature at the tunnel
outlet. With this assumption, solving the set of heat
balance equations is straightforward and the temperatures
of the cables’ surface, air and tunnel wall are easily
derived as a function of the cables losses.
As regards the heat generated by the cables, it is
assumed to be constant along the cable route, computed
for the core maximum permissible temperature, leading to
an estimate of the current carrying capability on the safe
side.
The permissible current rating is obtained from the
following formula which is similar to the classical formula
for cable rating:
W a L  
Ta  Tt  Te
DEEP TUNNELS
The losses of the cables installed in a tunnel are
dissipated in its surroundings, leading to a temperature
rise of the tunnel wall.
Due to the soil thermal inertia, a long duration is
necessary to reach the steady-state value. Considering a
circular tunnel (2 m in diameter) and assuming a constant
heat rate, Figure 4 gives the tunnel wall temperature rise
evolution with time, reported to the steady-state value of
the current, for 3 laying depths (10, 20, 40 m).
The temperature rises quickly at first but then the growth
is very slow. Applying standard steady state calculation
algorithm would yield ampacities that are too small.
A more appropriate approach would be to use the
transient analysis algorithm and iteratively find out what
value of the current would give desired temperature at the
end of the study period.
 Tunnel w all tem perature rise ratio vs tim e Ratio Equivalent / Actual depth vs Duration
100
1

90 0.9
Temperature rise ratio (%)

0.8
80 0.7

0.6

Ratio
70 0.5
0.4
10 m
60 '10 m'
20 m 0.3
'20 m'
40 m 0.2
'40 m'
50
0.1
0
40 0 10 20 30 40 50 60
0 10 20 30 40
Tim e (years) Duration (years)

Figure 4 :Tunnel wall temperature rise Figure 5 : Equivalent depth for deep tunnels
The formula for the steady-state external thermal
resistance is given above as relationship [9]. REFERENCES
The transient thermal resistance of the tunnel [1] IEC 60287-1-1, " Electrical cables – calculation of the
surroundings was computed with a standard equation current rating – Part 1-1 Current rating equations
using the exponential integrals given in [12]. (100% load factor and calculation of loses – General”

   D2   [2] IEC 60287-2-1, " Electrical cables – calculation of the

T4 t   . Ei  t
4.  16.t.
  
where  is the soil thermal diffusivity;  is the soil thermal
resistivity, Dt is the tunnel diameter and L its burial depth,
to tunnel centerline.
One can define a fictitious equivalent depth [13] of the
cable circuit that with the application of the steady state
algorithm would give the same value of the current as the
one obtained from the transient analysis (see Figure 5).
D 1   D² 
LT  . cosh  . Ei    Ei  
2 2   16.t. 
CONCLUSION
As the number of cable tunnels has been increasing over
the last decade, IEC TC20 WG 19 started a new work
item aiming at extending the scope of the present cable
rating methods, within IEC 60287 standard series.
This paper summarizes some of these works.
As distribution and transmission systems may be installed
in a tunnel, to take full benefit of the costly equipment, the
rating of groups of cables, with different designs, is
addressed.
The main issues addressed in the standard to be issued
shortly are presented, dealing with the rating of cables in
ventilated tunnels .
Finally, as it was recognized that, for deep tunnels,
applying standard steady state calculation algorithm
would yield ampacities that are too small, a fictitious
equivalent depth is introduced, to optimize cable rating,
taking into account the soil thermal inertia, without a
transient analysis
  Ei   L ²  current rating – Part 2-1 Thermal resistance –
  t. 
Calculation of thermal resistance”
[3] IEC 60287-2-2, " Electrical cables – calculation of the
current rating – Part 2-2 Thermal resistance – A
method for calculating reduction factors for groups of
cables in free air, protected from solar radiation”
[4]. CIGRE, (1992a) "Calculation of Temperatures in
Ventilated Cable Tunnels - Part 1", Electra, No.143
[5] CIGRE, (1992b) "Calculation of Temperatures in
Ventilated Cable Tunnels - Part 2", Electra, No.144
 L ²  
[6] G.J. Anders, “Rating of Electric Power Cables –
  Ampacity Calculations for Transmission, Distribution
 t.   and Industrial Applications" (1997) IEEE Press, New
nd
York. 2 printing jointly IEEE Press and McGraw-Hill,
(1998) New York.
[7] P.Slaninka, 1969, " External thermal resistance of air-
installed power cables", Proc. IEE Vol. 116 n°9,
September 1969
[8] F.P. Incropera – D.P. De Witt, WWWW, McGraw-Hill
Handbook of heat transfer [Third Edition]
[9] W.M. Rohsenow – J.P. Hartnett – Y.I. Cho –
“Handbook of heat transfer” – third edition – McGraw-
Hill Handbooks – page 7.83
[10] J.A. Pilgrim, D.J. Swaffield, P.L. Lewin et alii, “Rating
independent cable circuits in ventilated tunnels”, IEE
Trans. On power delivery, Vol. 25 n°4, October 2010
[11] Weedy – El Zayyat “Heat transfer from cables in
tunnels and shafts”, IEEE Con Paper C72 506-4
[12] IEC 60853-2, " Calculation of the cyclic and
emergency current rating of cables – Part 2: Cyclic
rating of cables greater than 18/30 (36) kV and
emergency ratings for cables of all voltages”
[13] E .Dorison, G.J. Anders, F. Lesur, “Ampacity
calculations for deeply installed cables”, IEEE Trans.
on Power Delivery, TPWD, Vol. 25, No. 2, April 2010