Miodrag STOJANOVIC1, Dragan TASIC1, Aleksa RISTIC1

University of Nis, Faculty of Electronic Engineering (1)

Cyclic Current Ratings of Single-Core XLPE Cables

with Respect to Designed Life Time
Abstract. This paper provides an overview of overload capacity of single-core XLPE cables buried in the ground. Cyclic rating factor obtained by
IEC60853-1 and IEC60853-2 standards is conservative due to assumption that maximum temperature during the cycle is equal to rated temperature.
Using the procedure proposed by IEC60853-2, transient temperature response of the cables exposed to daily cyclic loading is calculated. Then,
overload capacity of cables is obtained with respect to assigned design life of cables using Arrhenius-IPM electrothermal life model.

Streszczenie. W artykule przedstawiono wyniki analizy możliwości przeciążeniowych jednożyłowego kabla grzebanego w ziemi. Na podstawie
procedury zawartej w standardzie IEC60853-2 określono wpływ dobowego obciążenia energetycznego na zmiany temperatury kabla. Następnie
korzystając z modelu elektro-termalnego Arrhenius-IPM i uwzględniając dane znamionowe określono możliwości przeciążeniowe kabli.
(Wyznaczanie okresowego obciążenia prądowego kabla jednożyłowego XLPE z uwzględnieniem znamionowego czasu życia).

Keywords: single-core cables, cyclic current rating, electrothermal life model.

Słowa kluczowe: kable jednożyłowe, okresowe obciążenie prądowe, model elektrotermiczny.

Introduction [6-8]. As possible alternatives for representing the effects of

Power cables are designed under the conservative the electrothermal aging of insulation, three different life
constraints that their maximum temperature is constant and models are considered in [7], each within the probabilistic
equal to rated temperature, which is the consequence of the framework needed for associating time-to-failure with
fact that rated current of HVAC cables implies the operation reliability.
at the rated temperature, i.e. the maximum permissible In this paper, the overload capability of single-core
continuous temperature of cable insulation. Calculation of XLPE cables buried directly in the ground is analyzed.
cable current caring capacity, as well as the calculation of Firstly, the temperature of the cable is calculated for two
the overload capacity of cables under cyclic loading types of daily load cycle diagram of HVAC cables. Using the
occupies the attention of experts for a long time [1-5]. Arrhenius-IPM electrothermal life model, expected life time
Fundamentals of these calculations are set out in paper [1]. of single-core XLPE cables, as well as failure rate at the
References [4, 5] among others give a method for end of design life time are calculated for different values of
calculating the variation of the cable conductor temperature overload factor.
during a stepwise-constant load cycle by calculating the
thermal response of the HVAC cable and the surrounding Thermal response of cables and cyclic rating factor
environment to each step change of load current. Loading of HVAC power cables varies on the daily
Overload capacity of cables is usually quantified by cycles. In order to perform computations for variable
cyclic rating factor. Standard IEC60853-1 [4] provides the loading, daily load curves are divided into a series of steps
method for calculating the cyclic rating factor for cables with constant magnitude and with the same duration as that
whose internal thermal capacitance can be neglected. illustrated in Fig. 1. In this figure it can be distinguished two
Simplified method presented in this standard requires only different daily load cycles, Cycle I and Cycle II. Duration of
knowledge about the shape of the load variation for not one step is ti=24/N, where N is the number of steps. For
more than six hours immediately preceding the time of different successive steps, the computations are done
maximum temperature and an average value of times repeatedly, and the final result is obtained using the
before that. The method can be applied to all sizes and principle of superposition.
types of cables for nominal voltage up to and including
18/30 (36) kV. Standard IEC60853-2 [5] provides the
manual method for calculating cyclic rating factor for cables
whose internal thermal capacitance cannot be neglected. In
general, this method is applied to cables for nominal
voltages greater than 18/30 (36) kV. Cyclic ratings include
temporary overloads under condition that rated temperature
is not exceeded. Considering this, cable exposed to cyclic
loading with cyclic rating factor calculated according to
standard IEC60853, reaches the nominal temperature only
at one point. Cable temperature, during all day, is lower
than its nominal value. The question is, whether and how
much the cables can be additionally loaded without the risk
or causing economic damage. Temperatures higher than
nominal are allowed in emergency regimes, and these
regimes are considered also when determining the life time
of cable. Therefore, the limit on the maximal permissible
temperature during cyclic loading depends on the life time
of cable or maximal permissible temperature in emergency
regime.
The procedure for life estimation of high voltage AC Fig.1. Daily load cycle diagram
cables subjected to voltage and load cycles is proposed in

152 PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 89 NR 5/2013

 The variation of the cable conductor temperature during where:
a stepwise-constant load cycle can be determined by
calculating the thermal response of the HVAC cable and the  R (t )
(6)   (t )(1  k1   (t )k1 )
surrounding environment to each step change of load  R ( )
current. The two partial temperature transients are solved
W1 (T4  T4 )
separately in a sequence and then combined, thereby (7) k1 
finding an analytical solution for the whole transient as W J (T A  TB )  W1 (T4  T4 )
follows.
 te   Dc2   2 
Transient temperature rise of the conductor above the
 (t )   Ei -   Ei - h 
ambient temperature, due to the ith step of stepwise- 4 (T4  T4 )    t 
constant load cycle is:   16 t    
(8)
(1)  i (t )   c, i (t )   i (t )   e, i (t )   d2   2  
 2 Ei - 1   Ei - d1  
where c,i(t) is transient temperature rise of the conductor   16 t    t 
    
above cable surface, e,i(t) is cable surface temperature rise
above ambient temperature and i(t) is attainment factor. Thermal resistance T4 for cable buried directly in the
Transient temperature rise of the conductor above cable ground is:
surface is calculated as:
 2 
(2)  c, i (t )  W J , i [Ta (1  e  at )  Tb (1  e bt )] (9) T4 
 te  2h
ln
 2h 
 

 1 
2  Dc 
 Dc  
where WJ,i is power loss per unit length in a conductor due  
to the ith current step, and Ta, Tb, a, b are thermal
resistances and corresponding coefficients of the first loop while thermal resistance ΔT4 in the case of cable line
of CIGRE transient two loop network [5]. consisted of three single-core cables is:
For the case of three single-core cables buried directly in
 te  d1 
the ground in a flat formation temperature rise of cable (10) T4  ln 
surface above the ambient is:   d1 
 teW1, i   Dc2   2  Obviously, the method for cyclic rating factor calculation
 e, i (t )   Ei    Ei  h  does not take into account the variation of conductor
4   t 
  16t    temperature and assumes that Joule losses are directly
(3) proportional to square of current. The results obtained on
  d2   2   this way are therefore also conservative.
 2  Ei  1   Ei  d '1  
  16t   16t 
     Electrothermal life model
Electrical stress and thermal stress are dominant factors
where: te is thermal resistivity of soil,  is diffusivity, Dc is when considering the aging of HVAC cables. The most
outer diameter of cable, h is laying depth, d1 is center-to- popular electrothermal life model is the combination of two
single-stress life models, the Arrhenius model for thermal
center distance between cables, d1  (2h) 2  d12 , life and the Inverse power model for electrical life [7, 8].
W1,i  WJ ,i (1  1 ) , 1 is sheath loss factor, and Ei is Expected life time of the cable whose temperature is T and
electric field in insulation is E can be calculated:
exponential integral [9].
 ( n 0  b cT )
The attainment factor in equation (1) is calculated as:  E 
 c, i (t ) (11) L  L0 e  B cT  

(4)  i (t )   E0 
W J , i (Ta  Tb )
where E0 is value of electric field below which electrical
Finally, the temperature of conductor is obtained by aging is negligible, cT=1/T0-1/T is conventional thermal
adding the temperature rise for each step of the load stress, T0 is reference temperature, n0 is voltage endurance
diagram, calculated by (1), to temperature of the ground coefficient, L0 is life at T=T0 and E=E0, B=W/k, W is
and temperature rise due to dielectric losses. Electrical activation energy of the main thermal degradation reaction,
resistance of conductor and corresponding Joule losses are k is Boltzmann constant, and b is parameter that rules
calculated with respect to temperature of conductor reached synergism between electrical and thermal stress. The
in each step of daily diagram. parameters of Arrhenius-IPM model for XLPE insulation are
The cyclic rating factor is defined as the factor by which given in Table 1 [7]. During one step of daily stepwise-
the rated current may be multiplied to obtain the permissible constant load cycle temperature of insulation varies.
peak value of current during a daily load cycle such that the Therefore, different values of thermal life of cable insulation
conductor attains, but does not exceed the rated are obtained for each moment during the day. Loss-of-life
temperature. For calculating cyclic rated factor, according to fraction relevant to the ith step of stepwise-constant load
standard IEC60853-2, only load cycle over a period of six diagram (Fig. 1) is defined as:
hours before the time of maximum temperature is needed,
while earlier values are replaced with constant one, t i
dt
proportional to loss-load factor. Therefore, the cyclic rating (12) LFi  
factor is given by: 0 Li (T )
1
(5) M 
1/ 2
 5   R (i  1)  R (i )    R (6)  
  Yi  
   ()   ()    1   ()  
 i 0  R R   R 

PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 89 NR 5/2013 153

Table 1. Parameters of Arrhenius-IPM model Table 3. Daily load cycle
Parameter Value I/Imax
Step. No. Time [h]
b [K] 4420 Cycle I Cycle II
B [K] 12430 1 00-04 0.26 0.26
n0 [non-dimensional] 15 2 04-08 0.70 0.91
E0 [kV/mm] 5 3 08-12 1 1
T0 [K] 293 4 12-16 0.83 0.83
t 2 5 16-20 0.96 0.98
6 20-24 0.52 0.52
According to Miner’s cumulative damage theory, the
sum of all loss-of-life fractions should yield 1 at failure. So, Daily variation of conductor temperature for the Cycle I
number of cycles to failure (number of days to failure in of single-core XLPE cable with 1000 mm2 cross-section is
case of daily cycles) is: shown in Fig. 2 for different values of overload factor
1 (maximum current relative to rated current). Numerical
N  values of stepwise-constant cycle loadings shown in Fig. 1
(13) K    LFi 
are given in Table 3. For overload factor (Imax/IR) value of 1,
 i 1 
difference between maximum and minimum temperatures is
Design life of cable corresponds to certain design failure only 16.8ºC (varies between 47.1ºC and 63.9ºC), while for
probability. The cumulative probability distribution function overload factor value of 1.4 this difference is 39.4ºC (varies
that is commonly used for associating time to failure between 82ºC and 121.4ºC). Also, from the Fig. 2 it can be
probability in case of polymeric insulation for power cables noted that for overload factor value of 1.2, the maximum
is the Weibull’s one. Failure probability at mission time tp is: temperature of conductor is close to 90ºC. Having in mind
[ (t p / L63% )  t ] that XLPE cable may be overloaded up to 105C in
(14) P (t p , E , T )  1  e emergency regime, from Fig. 2 can be concluded that the
overload factor for cycle I loading must be lower than 1.3.
where t is share parameter of cumulative probability
distribution function and L63% is failure-time with 63.2%
probability. The relevant failure rate at time tp can be
estimated through the following hazard function:
 t 1
t  t p 
(15) h(t p , E , T )   
L63% L 
 63% 
Based on the equation (15) failure rate can be
calculated for insulation of cable loaded by defined daily
stepwise-constant cycle at the end of design life time.
Test example
The procedure described in the previous sections is
applied to HVAC XLPE insulated single-core cables [10]
with aluminium conductors and copper wire screen.
Maximum voltage of cables is 123 kV. The data about
cables are shown in the Table 2. In Table 2, Sc is conductor
cross section, dc is diameter of conductor, Di is diameter
above the insulation, i is insulation thickness, Dc is outer Fig. 2. Temperature variation during daily cyclic loading I
diameter of cable, and IR is rated current for considered
The last row of Table 2 shows the results of cyclic rating
cable formation and used bonding method of metal screens.
factors (Cycle I and Cycle II) calculated by (5). As can be
It is assumed that three single-core cables are laid in a flat
seen, these values vary in very narrow range between
formation and metal screens of cables are cross-bounded.
1.174 and 1.196 for the Cycle I, and between 1.146 and
Table 2. Cable data 1.162 for the Cycle II. For 1000 mm2 cross-section cyclic
2
Sc [mm ] 630 800 1000 1200 1400 rating factors have values of 1.186 (Cycle I) and 1.154
dc [mm] 29.8 33.7 37.9 42.8 46.4 (Cycle II). For obtained daily variations of conductor
Di [mm] 58.6 62.5 67.3 73.8 77.4 temperature, the expected life time of cable during cyclic
i [mm] 13 13 13 13 13 loading (shown on Fig. 1) for different values of overload
2
Dc [mm ] 72.3 76.8 82 89.5 93.3 factor is calculated using relations (11)-(13).
IR [A] 740 845 950 1025 1100 Fig. 3 shows expected life time of single-core XLPE
Cycle I 1.174 1.18 1.186 1.192 1.196 cables for the case of cycle loading I, with different cross-
M
Cycle II 1.146 1.15 1.154 1.159 1.162 section areas in the case of daily cyclic loading as function
of overload factor. Having in mind that design life of cable is
Laying depth of cables is 1 m, ground temperature 20°C
30 years, it is obvious that for daily cyclic diagram I cables
and distance between cables is d1=Dk+70 mm. Design life of
can be overloaded more than 28% without shortening the
cables is 30 years, while rated temperature of conductor is
life time. As previously noticed, according to the Fig. 2,
90°C. Thermal resistivity and thermal capacity of cross-
6 3 overload factor of the cable for given daily cycle diagram
linked polyethylene are 3.5 Km/W and 2.4·10 J/(m K),
and maximal temperature of 90ºC is slightly above 1.2.
respectively; thermal resistivity of soil 1 Km/W, and thermal
Fig. 3 shows that cable life time from the aspect of
capacities of aluminium and copper are 2.5·106 J/(m3K) and
6 3 electrothermal aging, for given overload factor of 1.2, is
3.45·10 J/(m K), respectively. It is assumed that the
nearly 80 years. For cyclic rating factor of 1.186 calculated
constant electric field is equal to design values. Currents
according to relevant standard, life time from the aspect of
that correspond to each step of daily load cycle are given in
electrothermal aging is approximately 90 years.
Table 3.

154 PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 89 NR 5/2013

 1000 1000
2 2
630 m m 630 m m
2 2
800 m m 800 m m
2 2
1000 m m 1000 m m
2 2
1200 m m 1200 m m
100 2 100 2
1400 m m 1400 m m
Life Tim e [years]

Life Tim e [years]

10 10

1 1
1.0 1.1 1.2 1.3 1.4 1.0 1.1 1.2 1.3 1.4
Im ax/IR Im ax/IR

Fig. 3. Life time of cables for cyclic loading I Fig. 6. Life time of cables for cyclic loading II

Fig. 4 shows results of failure rate calculation at the end 1

of design life time. As can be seen from this figure, for 630 m m
2

overload factor of 1.186 or 1.2, values of failure rates at the 800 m m

2

end of design life time are very low. Having in mind all the 0.1 1000 m m
2

facts stated above, it can be concluded that considered 1200 m m

2

power cable (with 1000 mm2 cross-section) has the

Failure rate [1/yearkm ]
2
1400 m m
potential for additional current load increase of 8.3% 0.01
compared to the value obtained by the standard. This
additional current load increase is similar to ones which
correspond to other power cables with different cross- 1E-3
sections.
1
2 1E-4
630 m m
2
800 m m
2
0.1 1000 m m
2
1E-5
1200 m m 1.0 1.1 1.2 1.3 1.4
Failure rate [1/yearkm ]

2
1400 m m Im ax/IR
0.01

Fig. 7. Cables failure rate at the end of design life for cycle II
1E-3
Fig. 5 illustrates the temperature variation during daily
cycling loading II, while Fig. 6 and fig. 7 illustrate life time of
cables and cables failure rate at the end of design life for
1E-4
cycle II, respectively. The same analysis can be carried out
for the case of cycle loading II and also, the same
conclusion can be derived. The only difference between two
1E-5
1.0 1.1 1.2 1.3 1.4 cycles is that for the cycle loading II, additional current load
Im ax/IR
increase is approximately 6% and for the cycle loading I it
accounts approximately 8%, compared to values obtained
Fig. 4. Cables failure rate at the end of design life for cycle I by the standard.

Conclusion
In this paper, a transient temperature response
calculation of single-core XLPE cables buried in the ground,
at two daily load cycles, is conducted. For the assumed
daily load cycles, cyclic rating factors and daily temperature
variations at different values of overload factor are
determined. On the basis of obtained results, it is shown
that rated temperature is attained at values of overload
factors slightly higher than cyclic rating factor values. Using
the Arrhenius-IPM electrothermal model, expected cable life
time for different values of overload factors, respecting the
daily cable temperature variations, is determined. On the
basis of obtained results, it is concluded that overload
capability of single-core XLPE cables, considering the
design life time, is higher than one obtained by relevant IEC
standard. It is shown that cables can be additionally loaded
Fig. 5. Temperature variation during daily cyclic loading II up to 8%, depending on the cycle.

PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 89 NR 5/2013 155

 REFERENCES Expectancy of High Voltage AC Cables, IEEE Trans. Power
[1] Neher J.H., McGrath M.H., The Calculation of the Temperature Delivery, 22 (2007), No.4, 2000-2009
Rise and Load Capability of Cable Systems, AIEE [8] Mazzanti G., The Combination of Electro-thermal Stress, Load
Transactions, Power Apparatus and Systems, 76 (1957), No. 3, Cycling and Thermal Transients and its Effects on the Life of
752-772 High Voltage AC Cables, IEEE Trans. On Dielectrics and
[2] Anders G.J.,El-Kady M.A., Transient Rating of Buried Power Electrical Insulation, 16 (2009), No. 4, 1168-1179
Cables, Part 1: Historical Perspective and Mathematical Model, [9] Abramowitz M., Stegun I., Handbook of Mathematical
IEEE Trans. Power Delivery, 7 (1992), No.4, 1724-1734 Functions, New York, Dover Publications, INC., 1972
[3] Anders G.J., Rating of Electric Power Cables in Unfavorable [10] XLPE Land Cable Systems, ABB, User’s Guide, 5 (2010)
Thermal Environment, New York, Wiley-IEEE Press, 2005
[4] Calculation of the Cyclic and Emergency Current Ratings of
Cables, Part 1: Cyclic Rating Factor for Cables up to and Authors: MSc. Miodrag Stojanovic, University of Nis, Faculty of
Including 18/30 (36) kV, IEC Std. 60853-1, 1985 Electronic Engineering, Aleksandra Medvedeva 14, 18000 Nis,
[5] Calculation of the Cyclic and Emergency Current Ratings of Serbia, E-mail: miodrag.stojanovic@elfak.ni.ac.rs; prof. dr Dragan
Cables, Part 2: Cyclic Rating Factor of Cables Greater than Tasic, University of Nis, Faculty of Electronic Engineering,
18/30 (36) kV and Emergency Ratings for Cables of All Aleksandra Medvedeva 14, 18000 Nis, Serbia, E-mail:
Voltages, IEC Std. 60853-2, 1989 dragan.tasic@elfak.ni.ac.rs; dipl. ing. Aleksa Ristic, University of
[6] Mazzanti G., Montanari G. C., A Comparisom between XLPE Nis, Faculty of Electronic Engineering, Aleksandra Medvedeva 14,
and EPR as Insulated Materials for HV Cables, , IEEE Trans. 18000 Nis, Serbia, E-mail: aleksa.ristic@elfak.ni.ac.rs;
On Power Delivery, 12 (1997), No. 1, 15-28
[7] Mazzanti G., Analysis of the Combined Effects of Load Cycling,
Thermal Transients and Electrothermal Stress on Life

156 PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 89 NR 5/2013