Thermal Analysis of Underground Power Cables-

A Monitoring Procedure
Larisa Mariut*, Elena Helerea*, Gavril Lungoci†, Sorin Abagiu†
* Transilvania Univerisity of Brasov/Department of Electrical Engineering and Applied Physics, Brasov, Romania
†
Transilvania SUD Electrica Distribution Networks SA, Brasov, Romania
elena-larisa.mariut@unitbv.ro, helerea@unitbv.ro, Gavril.Lungoci@electricats.ro, s.abagiu@ts.electricaserv.ro

Abstract— The implementation of underground power
distribution lines grown significantly over the last decades
and this has increased to need to obtain a high level of
operational reliability. Current monitoring strategies for
underground power lines are developed based on different
diagnosis tools and many of them are not cheap at all. As
an example, traditional methods of thermal endurance
characterization for XLPE insulated power cables are all
based on analytical tools-infrared carbonyl absorbance,
melting point, differential scanning calorimetry and oth-
ers. These diagnosis techniques are either too time con-
suming or some can only give vague data that cannot be
interpreted. The current paper proposes a new method for
underground power cable monitoring based on cable his-
torical operation temperature. Based on the current study,
the aim of this innovative monitoring procedure is to esti-
mate the degree of ageing for an underground power cable
considering thermal stresses. Different parameters related
to thermal aging are determined such as daily thermal
aging and the average daily rate of aging. The input data
considered for the current researches are chosen arbitrari-
ly, but it can be switched with real data from exploitation.
For thermal lifetime estimation, the thermal endurance
profile for XLPE was applied. The current procedure can
be applied either paper insulated or XLPE insulated pow-
er cables.
I. INTRODUCTION
The insulation system of power cables is subjected
during manufacturing/ exploitation to various thermal
stresses because of environmental conditions or variable
loading [1]. For paper and XLPE insulated power ca-
bles, the thermal aging generates a structural and chem-
ical irreversible change and this aging mechanism af-
fects the lifetime of the entire power cable [1]. The
problem of thermal endurance characterization for dif-
ferent types of power cables is still under debate [2-6].
Thermal degradation can be assessed by applying dif-
ferent analytical diagnosis tools, like: melting point,
crystallinity, gel content, infrared carbonyl absorbance,
relative hardness, differential scanning calorimetry,
electrical strength, tensile strength and elongation at
break. The major disadvantage of these techniques is
that some are too time consuming, some need dedicated
testing procedures, and some can only give vague data
that cannot be interpreted and so on.
Based on these considerations is necessary to develop
highly accurate tests/ procedures to provide: (1) a con-
crete condition monitoring of the power cable, (2) an
efficient asset management procedure for power cable
degradation estimation / lifetime estimation [7-16].
One of the reasons for power cables lifetime reducing
is due to resistive conductor heating process that is car-
rying the load current. Hence, it can be realized the con-
nection between power cables thermal life analysis and
ampacity [17]. The ampacity is strongly dependent on
the operating temperature, but also, the temperature of
the cable conductor is proportional to the amount of
loading current passing through it. The insulation deg-
radation process under thermal stresses began to be
studied in 1930 by Montsinger. His study was related to
the exponential relation between the temperature and
failure time -life is halved by a temperature increase of
8 to 100C [18], [19].
The current paper proposes a novel condition monitor-
ing procedure for power cables based on thermal analy-
sis of the power cable. The analysis was performed con-
sidering that the power cable operating temperature is
strongly influenced by his daily loading diagram.
During the thermal analysis, various parameters/ input
data were considered. In this sense, the input data are
arbitrarily chosen but they can be replaced with data
obtained from exploitation. Based on this assumption, it
should be emphasized that the proposed monitoring
procedure has a theoretic substrate. The aim of this
study is to propose a new method for power cables con-
dition monitoring considering the cable historical oper-
ating loading / temperature curves. The paper has two
main objectives: the first one was to describe the moni-
toring procedure, and the second one was to apply the
procedure in an approximate manner. The second objec-
tive was achieved starting from a daily temperature dia-
gram. The diagram was proposed according to current
standards [22] and technical recomandations from [20-
24] and considering the maximum temperatures within
cable conductor (Table I) in the case of an 10-15 %
overload state [25], [26]. It was considered that during
24 h of operation, the cable was subjected to a variable
load. In this sense, the proposed thermal diagram of the
power cable has 4 different time intervals.
The expected results of the current study refer to: (1) a
new monitoring procedure for power cables during ex-
ploitation; (2) a modification of the current procedure
for power cables performances recording during exploi-
tation, achieved by introducing daily thermal curves of
the cables as a monitoring tool. In this way can be de-
veloped multiple life analysis regarding future perfor-
mances of the power cables in a very simple and cheep B. Work procedure
manner. If the load curve of the power cable is available, for a
TABLE I better remaining lifetime estimation, the thermal history
CONDUCTOR TEMPERATURE LIMITS FOR STANDARD CABLE TYPES of the cable is essential to be determined.
Power cable type Conductor The thermal analysis was performed according to an
temperature [0C] analytical procedure using C Sharp Express Edition
Paper insulated cables 80-90
Rubber insulated cables 80-110
2010 (Fig. 1).
PE/PVC insulated cables 90-95 The first step needed for thermal analysis modeling,
XLPE insulated cables 110-115 consist of calculation of conductor temperature operat-
Oil pressure paper insulated cables 105-110 ing range. In this sense, is obviously that the conductor
power losses estimations are needed. According to the
The values from the above table indicate a major ad- established work conditions, the conductor power losses
vantage for the materials, which can operate at high were determined considering the Joule heating through
temperatures. However, these values are influenced dur- the cable conductor, for each time interval of the exist-
ing exploitation by different factors like environment ing load curve.
temperature / stresses, soil resistivity / temperature /
humidity / cable installation depth. ac 2
Wi  R20  Ii (2)
Continuous operation at cable surface temperatures
above 500C will cause movement of moisture away
ac
from the cables and, with many types of cables, drying Where, R20 is AC resistance (/cm) of the cable
out of the backfill may occur and the cable could exceed conductor and Ii represents the current (A) passing
the permissible temperature. through the conductor at i time interval, according to the
II. METHODOLOGY
daily load curve.
The range of the conductor operation temperature
During different operation conditions / regimes, depends on: thermal resistance of the cable (3) and
whether it is stationary, whether transient, the conduc- thermal resistance of the soil (4). In the current study it
tors of the power cables are crossed by different loading was used a cable without armour and considering this,
currents that cause excessive heating [6]. his thermal resistance reffers only to the thermal
In the case of underground power cables, the insula- resistance between conductor and sheath.
tion resistance decreases exponentially with temperature
and the other parts of the cable are affected by different   2 
physical / chemical structural changes that lead to power Rtc  T ln1  i 
 (3)
2  dc 
cable lifetime reduction and could negatively influence
their safety in operation [6].
Where, T is the insulation thermal resistivity
A. Input data
In the next section are presented the necessary data
 C  cm / W  ,
0
 i represents the thickness of the
needed for the thermal analysis modeling. In this sense, insulation between conductor and sheath (cm) and d c is
one can use the daily/ monthly/ annual load curves of the conductor diameter (cm).
the power cable that should be available and they serves
as input data. Also, the technical parameters of the cable  2 
H  H 
and of the environment where the cable is Rts  0.366  s  lg      1  (4)
R 
undergrounded must be known (cable length (l (cm)),  c  Rc  
 
cable type (e.g. ACHPI), conductor cross section (s
(cm2)), conductor resistivity at 200 AC (  c , (   cm )),
conductor diameter (dc, (cm)), soil temperature (around
Where,  s is the soil thermal resistivity 0 C  cm / W , 
200C)). H represents the installation depth of the cable (cm) and
The following work conditions were considered: Rc is the outer radius of the cable (cm). In this study it
 Considering the cable length, the temperature 
was considered that  s =50 0C  cm / W . If the ratio 
should have a uniform distribution; H / Rc  10 , then the thermal resistance of the soil can
 The entire amount of heat received by radiation
from around entities of the conductor should be be determined using:
considered;
4H
 Because the cable conductor is crossed by Rts  0.366  s  lg (5)
Dc
normal alternative currents, the skin and proximity
effects are neglected. Considering these, the following
equation for the cable resistance () was applied: Where, Dc represents the outer diamater of the cable
(cm).
ac l
R20  c  (1)
s
 START

Daily load
curve is known.
It has n time
intervals.

All the input data,

according with Table
II, are also known.

Ec. (1) Write R

i=1

Y
i<=n Ec. (2) Write Wi i=i+1

Ec. (3) Write Rtc

Calculation of the ratio

H / Rc

N Y
H
Write Rts Ec. (4)  10 Ec. (5) Write Rts
Rc

i=1

Y N Y Y
i  Rtc Wi  exp1  ti   Rts Wi  exp 2  ti  i<=n i 2  0 i<=n i  Rtc  Wi  (1  exp 1  ti )  Rts  Wi  (1  exp  2  ti )

N
Write i N i Write

i=i+1 i=i+1
i=1

N Y
STOP i<=n Ec. (7) Ec. (8) i=i+1

Write Write
relative relative
thermal aging
aging speed

Write average thermal aging Ec. (9) Ec. (10) Write average aging speed

Fig. 1. Thermal analysis procedure flowchart.

 According to the load curve, for each time interval of
the diagram it must be determined the conductor The load curve was developed according to current
temperature variation considering a variable loading. standards [20-24] and databases from a Romanian Elec-
tricity Compay for an A2YSY 20 kV single phase pow-
 i  Rtc  Wi  exp1  ti   Rts  Wi  exp 2  ti  (6) er cable. According to Fig. 2, one can observe 4 (n) time
intervals (t1=6 h, t2=4 h, t3=8 h, t4=6 h) with 4 different
Where, α1 represents the time constant of the cable currents passing through the conductor (I1=5 A, I2=15
(α1=0.6 1/h) and α2 represents the time constant of the A, I3=10 A, I4=15 A). Therefore, the present application
soil (α2=0.031 1/h). So for each time interval one must has a theoretical substrate.
determine the conductor temperature according to (6). As future work, we want to develop an expert system
The second step necessary for the thermal analysis based on the presented procedure in order to test under-
modeling is to determine the relative thermal aging of ground power cables and to improve current condition
the cable during 24 h of operation and the average aging monitoring strategies for medium voltage power cables.
speed of the cable. Firstly, for each  i it must be A. Input data
determined the relative thermal aging (7) and the
relative aging speed(8). The parameters need for thermal analysis modeling,
are presented in Table II.
vimb ( i )  Dt ( i ) 1 (7) Considering the input data from Table II and based on
the thermal analysis procedure presented in Fig. 1, for
r ( i )  vimb ( i )  t i (8) the considered A2YSY power cable, were determined:
i) the thermal resistance of the cable, the thermal
resistance of the soil and the conductor
Secondly, based on data previously obtained, the
temperature distribution for each time interval
average thermal aging (9) of the cable during 24 h of
(Table III, Table IV);
operation and the average aging speed of the cable (10)
ii) the relative thermal aging, the relative aging
that corresponds to a daily thermal aging can be
speed, the average thermal aging and the
determined.
average aging speed (Table V, Table VI).
n
rzi   ri (9) TABLE II
i 1
INPUT DATA FOR THERMAL ANALYSIS
Parameter Value Units
Where n is the total number of the time levels. s 1.5 cm2
rday l 2  10 5 cm
vi (imb )  (10)
24 H 70 cm
c 2.8264  10 6   cm
dc 1.37 cm
III. APPLICATION Dc 3.38 cm
º
s 20 C
In order to meet the second objective of the current
paper, within the next section a thermal analysis was T 350 
C  cm/W
performed. The analysis was developed considering a i 1 cm
daily load curve (Fig.2) for a single phase medium volt- 0.6 1/h
1
age power cable.
2 0.031 1/h
s 50 
Current C  cm/W
I [A]
16 Rc 1.69 cm
14
12 B. Results
10
8 In the next section, are presented the determined pa-
6
4
rameters of the cable thermal analysis.
2
0
2 4 6 8 10 12 14 16 18 20 22 24
Current 5 5 5 15 15 10 10 10 10 15 15 15
TABLE III
Time [h] THERMAL PARAMETERS OF THE CABLE
Thermal Thermal
resistance of resistance of
the cable the soil
Rtc  50.16 Rts  35.10
Fig. 2. Daily load diagram-20 kV A2YSY power cable.
 Based on the determined conductor temperature val- Considering the parameters presented in Table VI, it
ues at different time intervals during 24 h of operation, was estimated the lifetime of the power cable consider-
(Table III), the current power cables has a temperature ing the Arrhenius model. In this sense, for the current
diagram according to Fig. 3. cable that has an average thermal aging around
6.527 10 6 h, during 24 h of operation, the estimated
TABLE IV lifetime is about 1.62 years. The Arrhenius parameters
THERMAL PARAMETERS OF THE CABLE CONDUCTOR were determined at 2 known temperature levels of the
Power Conductor cable insulation (according to Fig. 3).
losses temperature According to Table V, it was shown that:
1. The relative thermal aging has a increasing value;
W1  14 1  19
2. The lifetime estimation of the power cables decreases
W2  127  2  58 as the temperature increases;
W3  57  3  23 3. The relative aging speed increases as the temperature
W 4  127  4  62 increases.
During 24 h of operation the considered power cable
that develops a daily temperature diagram according to
Fig. 3, will have an average aging speed of about
TABLE V
RELATIVE AGING PARAMETERS OF THE CABLE 2.72 10 7 [h-1].
Relative Relative Based on the current results the lifetime estimation of
thermal aging aging speed the cable can be performed and the future daily curves
vimb (1 )  1010 r 1   6 106 will be developed in order to maintain a constant speed
aging. In addition, knowing the speed of the aging pro-
vimb ( 2 )  0.143  105 r  2   0.572  105
cess considering the thermal stress of the cable, the
v imb ( 3 )  10 7 r 3   8 107 maintenance engineers will establish the future correc-
tive / preventive actions.
vimb ( 4 )  10 9 r  4   6 10 9
By applying the current monitoring procedure for
power cables, one can develop condition-monitoring
All the equations applied in order to perform the ther- actions assuring a high degree of safety in operation.
mal analysis were used as described in [27].
TABLE VI
IV. CONCLUSIONS
AVERAGE AGING PARAMETERS OF THE CABLE The novelty of this paper consists in developing of
Average Average prognosis about lifetime distribution for a power cable
thermal aging aging speed based on records about his historical loading. In order to
achieve such a forecast, it is necessary to obtain a histo-
6.527 10 6 2.72  107 gram / diagram about daily thermal cycles of the cable.
The main issue relates to the poor management of such
kind of records by electricity suppliers.
The performed analysis was developed considering
70
that the temperature distribution is strongly affected by
Temperature 
60
the cable loading.
[0C]
50
A complete thermal analysis of a power cables was
done and based on the analysis results a monitoring pro-
40
cedure for power cables was proposed.
30
The thermal analysis was developed based on the dai-
20
ly thermal diagram of the cable that was achieved ac-
10
cording to cable daily load curve.
0
2 4 6 8 10 12 14 16 18 20 22 24 Different parameters related to thermal aging were de-
Temperature 19 19 19 58 58 23 23 23 23 62 62 62 termined such: as daily thermal aging and the average
Time [h]
daily rate of aging. The input data considered for the
current researches are chosen arbitrarily, but it can be
switched with real data from exploitation. For thermal
lifetime estimation, the thermal endurance profile for
XLPE was applied.
Fig. 3. Conductor temperature diagram.
The obtained results of the study refer to:
(1) A new condition monitoring procedure for power
cables;
According to Fig. 3, the cable has 2 temperature peaks (2) Modification of procedure to record the power ca-
during 24 h of operation: at 58 0C (2nd time interval) bles performances during exploitation. This was
and at 62 0C (4th time interval). achieved by introducing daily thermal curves of the ca-
bles as a monitoring tool.
 Future work will be focused on the development of an
expert system that will be applied as
- testing tool (to examine the underground power
cables regarding their thermal aging);
- condition monitoring tool (to improve current
condition monitoring strategies for medium volt-
age power cables);
- prognosis tool (to estimate the remaining lifetime
of an underground power cable based on his daily
loading).
ACKNOWLEDGMENT
This paper is supported by the Sectorial Operational
Programme Human Resources Development (SOP
HRD), financed from the European Social Fund and by
the Romanian Government under the contract number
POSDRU/88/1.5/S/59321.
REFERENCES
[1] M. Cacciari, G. C. Montanari, “Electrical Life Threshold Models
for Solid Insulating Materials Subjected to Electrical and Multi-
ple Stresses”, IEEE Transactions on Electrical Insulation, vol.
27, October 1992, pp. 987-999.
[2] G. Mazzanti, G. C. Monatanari, “A comparison between XLPE
and EPR as Insulating Materials for HV Cables”, IEEE Transac-
tions on Power Delivery, vol. 12, January 1997, pp. 15-28.
[3] S. A. Boggs, “Mechanisms for Degradation of TR-XLPE Im-
pulse Strength During Service Aging”, IEEE Transactions on
Power Delivery, vol. 17, April 2002, pp. 308-312.
[4] S. Pélisson, R. Harp, R. Bristol, et. all. “A review of possible
methods for defining tree retardant Crosslinked Polyethylene”,
IEEE Electrical Insulation Magazine, vol. 24, Septem-
ber/October 2008, pp. 22-29.
[5] G.C. Montanari, G. Mazzanti, L. Simoni, “Progress in
Electrothermal Life Modeling of Electrical Insulation during the
Last Decades”, IEEE Transactions on Dielectrics and Electrical
Insulation, vol. 9, October 2002, p. 730-745.
[6] T. Tanaka, “Aging of Polymeric and Composite Insulating Mate-
rials-Aspects of Interfacial Performance in Aging”, IEEE Trans-
actions on Dielectrics and Electrical Insulation, vol. 9, October
2002, pp. 704-716.
[7] D. Wald, “Evolution of Medium Voltage Cable Technology in
Europe”, IEEE Electrical Insulation Magazine, vol.24, Septem-
ber/December 2008, pp.31-35.
[8] Normativ pentru proiectarea si executarea retelelor de cabluri
electrice, NTE 007/08/00, S.C. Electrica S.A.
[9] D. W. Pryer, P.F. Gale, Cables, BICC Cables, Liverpool,
Blackwell Publishing Company, pp. 2-44, 1997.
[10] Rolf Isermann, Fault-diagnosis applications Model based condi-
tioning monitoring: actuators, drives, machinery, plants, and
fault tolerant systems, 2010, Springer Heidelberg Dordrecht
London New York-pp. 11-22, pp. 49-80.
[11] N. Murtagh, T. Bamba, K. Iwawa, “An evaluation tool for eco-
design of electrical products” in First International Symposium
on , Environmentally Conscious Design and Inverse Manufactur-
ing, February 1999, pp. 766-770.
[12] R. Bartnikas, K. Srivastava, “Power and communication cables
theory and applications”, IEEE Series on Power Engineering,
New York, NY, 1999.
[13] F. Faria da Silva, C. L. Bak, W. T. Wiechowski, “Study of High
Voltage AC Underground Cable Systems”, in PhD Seminar on
Detailed Modelling and Validation of Electrical Components
and Systems, Fredericia, Denmark, 2010, pp. 10-15.
[14] F. Aras, V. Alekperov, N, Can, H. Kirkici, “Aging of 154 kV
Underground Power Cable Insulation under Combined Thermal
and Electrical Stresses”, IEEE Electrical Insulation Magazine,
vol. 23, September/October 2007, pp. 25-33.
[15] C. Kim, Z. Jin, P. Jiang, Z. Zhu, G. Wang, “Investigation of
dielectric behavior of thermally aged XLPE cable in the high-
frequency range”, Polymer Testing, vol. 25, March 2006, pp.
553-561.
[16] M.H.Shwehdi, M.A.Morsy, A. Abugurain, “Thermal Aging
Tests on XLPE and PVC Cable Insulation Materials of Saudi
Arabia”, Annual Report Conference on Electrical Insulation and
Dielectric Phenomena , 2003, PP. 176-180.
[17] Jean Carlos Hernández Mejía, “Characterization of real power
cable defects by diagnostic measurements”, PhD dissertation,
Georgia Institute of Technology, 2008.
[18] M. R. Jaafari, Mousavi, “Underground distribution cable incipi-
ent fault diagnosis systems”, PhD dissertation, Texas A&M
University, 2005.
[19] T. Bengtsson, F. Dijkuizen, et. all., “Repetitive fast voltage
stresses-causes and effects” IEEE Electrical Insulation Maga-
zine, vol. 25, July/August 2009, pp.26-38.
[20] L. M. Dumitran, Sisteme de izolaţie electrică, Bucureşti : Editura
Printech, 2008.
[21] XLPE Land Cable Systems Users Guide-US, ABB, Sweden,
2010
[22] Normativ pentru proiectarea si executarea retelelor de cabluri
electrice, NTE 007/08/00, S.C. Electrica S.A.
[23] M.L.Dyer, Electric Power Generation, Transmission, and Distri-
bution, Chapter 12 Insulated Power Cables Used in Underground
Applications, Edited by Leonard L . Grigsby CRC Press 2007.
[24] T. Horváth, E. Németh, B. Máthé, V. Stanciu, Încercarea
izolaţiei electrice, Bucureşti: Editura Tehnică 1982.
[25] N. Hampton, R. Hartlein, H. Lennartsson, et. all., “Long life
XLPE Insulated Power Cable”, in Conf. Jicable, 7th Internation-
al Conference on Insulated Power Cables, Paris-Versailles, 2007.
[26] S. R. Szaniszlo, “1050/1400 Rated TR-XLPE Voltage Power
Cables”, Raport tehnic, August 2003.
[27] G. Georgescu, V. Varvara, M. Istrate, "Sisteme de distribuţie a
energiei electrice ", Volumul 2, Partea I, Editura "Tehnică,
Ştiinţifică şi Didactică CERMI", Iaşi, 2008.