Professional Documents
Culture Documents
Contribution for Panel on Modern Maintenance Techniques for enhancing the reliability of Insulation of
Power Transmission Systems
by
ASLE SCHEI
TransiNor As
(Norway)
On behalf of SC 33
SUMMARY
arrester and for the operation reliability of the power
The paper gives an overview of different diagnostic system. Various diagnostic methods and indicators for
methods for surge arresters to a great extent based on revealing possible deterioration or failure of the
the work done by Working Group 10 of IEC TC 37, insulating properties have been utilized since the
presented in Amendment 1 to the Arrester Application introduction of surge arresters. The diagnostic methods
Guide 99-5 [0]. The different methods are described range from fault indicators and disconnectors for
and classified with respect to service experience and indication of complete arrester failures, to instruments
technical properties. that are able to measure slight changes in the resistive
The most reliable diagnostic method for on-site leak- leakage current or the power loss of metal-oxide
age current measurements on gapless metal oxide surge arresters.
arresters, is found to be based on the third harmonic
analysis of the current measured in the earth lead of the For the old type of surge arresters with SiC-blocks and
arrester. However, because harmonics in the system series spark gaps, reliable diagnosis are not possible
voltage will contribute to the harmonic currents in the during normal operation by measurements on the
earth lead, it is necessary to compensate for the amount arrester terminals. The only realistic possibility for
of third harmonics caused by the voltage. This check of service condition for these types of arresters is
compensation is done by a field probe, located at the to bring them to a high voltage laboratory, or to use a
base of the arrester. Without compensation 1 % third portable high voltage test set, for grading current
harmonics in the voltage may give a measuring error of measurements and spark over test. This means the
± 100 % in the measured current. More than ten years arrester has to be disconnected from the network. Even
of experience has confirmed the efficiency of this then the SiC-blocks are not fully checked which means
compensation method. that complete diagnosis is not possible. The situation
is, however, completely different for the gapless metal
Keywords oxide arresters. This because the continuous leakage
Surge arresters. Metal-Oxide Arresters. Diagnostic current that flows through the arrester blocks can be
methods. measured in the arrester earth lead, and the resistive
component of this current has shown to be a good
indicator of the arrester operating condition.
1. INTRODUCTION
The scope of this document is to present an overview
Apart from brief occasions when a surge arrester is
of common diagnostic methods and to give detailed
functioning as an overvoltage-limiting device, it is
information about leakage current measurements on
expected to behave as an insulator. The insulating
gapless metal-oxide arresters, that is the bases for the
properties are essential for the length of life of the
most relevant diagnostic methods .
2. OVERVIEW OF DIFFERENT Direct measurements of the metal-oxide resistor
DIAGNOSTIC METHODS temperature give an accurate indication of the
condition of the arrester, but require that the arrester be
2.1 Fault indicators equipped with special transducers at the time of
Fault indicators give a clear visual indication of a failed manufacturing. Therefore, this method is normally
arrester, without disconnecting the arrester from the used only in special arrester applications.
line. The working principle is usually based on the
amplitude and duration of the arrester current, or on the 2.6 Leakage current measurements of metal-
temperature of the non-linear resistors blocks. oxide arresters
Any deterioration of the insulating properties of a
2.2 Disconnectors metal-oxide arrester will cause an increase in the
Disconnectors, often used on medium-voltage arresters, resistive leakage current or power loss at given values
give a visual indication of a failed arrester by of voltage and temperature. The majority of diagnostic
disconnecting it from the system. The typical working methods for determining the condition of gapless
principle is an explosive device triggered by the fault metal-oxide arresters are based on measurements of the
current; however, the disconnector is not intended to leakage current.
extinguish the fault current. The advantage of the
device is that the line remains in operation after The measuring procedures can be divided into two
disconnection of the arrester. The major disadvantage groups: on-line measurements, when the arrester is
is the lack of overvoltage protection until the failed connected to the system and energized with the service
arrester has been discovered and replaced. voltage during normal operation, and off-line
measurements, when the arrester is disconnected from
2.3 Surge counters the system and energized with a separate voltage
Surge counters operate at impulse currents above a source on site or in a laboratory.
certain amplitude, or above certain combinations of
current amplitude and duration. If the interval between Measurements off-line can be made with voltage
discharges is very short (less than 50 ms), surge sources that are specially suited for the purpose, e.g.
counters may not count every current impulse. Some mobile a.c. or d.c. test generators. Good accuracy may
counters require power follow current and may not be obtained by using the off-line methods, provided
count the short impulse currents through metal-oxide that a sufficiently high test voltage is used. The major
arresters. disadvantages are the cost of the equipment and the
need for disconnecting the arrester from the system.
Depending on the operating principle and sensitivity of
the counter, it may give an indication about Measurements carried out on-line under normal service
overvoltages appearing in the system, or it may provide voltage is the most common method. For practical and
information on the number of discharges corresponding safety reasons, the leakage current is normally accessed
to significant arrester energy stresses. The counter only at the earthed end of the arrester. To allow
provides no specific information about the condition of measurements of the leakage current flowing in the
the arrester. earth connection, the arrester must be equipped with
insulated base and earth terminal.
2.4 Monitoring spark gaps
Monitoring spark gaps are used to indicate the number On-line leakage current measurements are usually
and estimate the amplitude and duration of discharge made on a temporary basis using portable or
currents through the arrester. Special experience is permanently installed instruments, but may also be
necessary to properly interpret the marks on the gap. used for continuous monitoring of important arrester
Some spark gaps can be examined with the arrester in installations. Portable instruments are usually
service, while other types require that the arrester is de- connected to the earth terminal of the arrester by means
energized. Spark gaps give no direct information about of a clip-on or permanently installed current
the actual condition of the arrester, but may help to transformer. Long-term measurements of the leakage
make decisions about continued operation. current may be necessary for closer investigations,
especially if significant changes in the condition of an
2.5 Temperature measurements arrester are revealed by temporary measurements.
Remote measurement of the arrester temperature can Remote measurements may be implemented in
be carried out by means of thermal imaging methods. computerized systems for supervision of substation
The measurements are only indicative with regard to equipment.
the condition of the arrester, since the temperature drop
between the resistors and the housing surface may be
substantial. Nevertheless, comparative measurements
made on adjacent arresters or arrester units may in
extreme cases indicate excessive heating.
3.0 PROPERTIES OF THE LEAKAGE CUR- capacitive peak leakage current of about 0,2 to 3 mA
RENT OF METAL-OXIDE RESISTORS under normal service conditions.
1,0
0,8
U / Ur
Leakage current
i = 0,2...3 mA 0,6
Time DC, +20°C
0,4 DC, +40 °C
AC resistive, +20 °C
Figure 3.1 – Typical leakage current of a non-linear 0,2 AC resistive, +40 °C
metal-oxide resistor in laboratory conditions AC capacitive
0,0
0,01 0,10 1,00 10,00 100,00
4 Current - mA
Leakage current - mA
H ig h
2 Figure 3.3 – Typical voltage-current characteristics
for non-linear metal-oxide resistors
0
Low
-2
The resistive component under a.c. voltage is defined
-4 as the current level at the instant of voltage maximum
T im e
(dU/dt = 0), as indicated in figure 3.1. The resistive
leakage current of a non-linear metal-oxide resistor is
Figure 3.2 – Typical leakage currents of arresters in in the order of 5 to 20 % of the capacitive current under
service conditions normal operating conditions, corresponding to about 10
to 600 µA peak resistive current at a temperature of
+20°C.
3.1 Capacitive leakage current
The capacitive leakage current measured at the earth In the leakage current region, the resistive current
terminal of an arrester is caused by the permittivity of depends on the voltage and temperature. Typical values
the non-linear metal-oxide resistors, the stray of voltage and temperature dependencies under a.c.
capacitances and the grading capacitors, if applied. The voltage are indicated in figures 3.4 and 3.5, normalized
specific capacitance of a resistor element is typically to Uc and at +20°C, respectively.
60 to 150 pF·kV/cm² (rated voltage), resulting in a
6 certain methods, is the capacitive current induced in the
earth lead of the arrester by the adjacent phases.
5
resistive current
3.3 Harmonics in the leakage current
4 third harmonic current The non-linear voltage-current characteristic of a
power loss metal-oxide arrester gives rise to harmonics in the
3 leakage current when the arrester is energized with a
sinusoidal voltage. The harmonic content depends on
2
the magnitude of the resistive current and the degree of
non-linearity, which is a function of voltage and
1
temperature. As an example, the third harmonic content
of the resistive current is typically 10-40 %. The
0
0,5 0,6 0,7 0,8 0,9 1,0 1,1 1,2 1,3 harmonic content can, therefore, be used as an
U / Uc indicator of the resistive current. Typical values of the
Figure 3.4 – Typical normalized voltage dependence variations with voltage and temperature of the third
at +20°C order harmonic component are shown in figures 3.4
and 3.5.
The voltage distribution along an arrester may be Another source of harmonics, beside negligible ones,
uneven, primarily due to the influence of stray that may considerably influence the measurement of
capacitances to earth and to adjacent equipment. The harmonics in the leakage current, is the harmonic
voltage across the non-linear metal-oxide resistors at content in the system voltage. The capacitive harmonic
the earthed end of the arrester may, therefore, deviate currents produced by the voltage harmonics may be of
in both magnitude and phase from the average value the same order of magnitude as the harmonic currents
along the arrester. This phenomenon affects the created by the non-linear resistance of the arrester. An
measurement of the resistive leakage current in two example of harmonics in the leakage current caused by
ways: Firstly, the resistive current measured in the system voltage harmonics is seen in figure 3.2 .
earth connection depends on the magnitude of the
voltage across the non-linear metal-oxide resistors at 3.4 Power loss
the earthed end. Therefore, the measured resistive The power loss may be used for diagnostic indication
current may differ from the average resistive current of arresters in the same way as the resistive leakage
along the arrester. Secondly, the phase shift of the current. Typical values of power losses are 5-300
voltage across the non-linear metal-oxide resistors at mW/kV (rated voltage) at Uc and +20°C. The
the earthed end influences the result of resistive current temperature and voltage dependencies are practically
measurement for methods that are using the voltage the same as for the resistive current, as seen in figures
across the complete arrester as a reference for the phase 3.4 and 3.5.
angle.
3.5 Surface leakage current
As with any other outdoor insulator, external surface
12 leakage current may temporarily occur on the arrester
housing in rain or in conditions of high humidity
10 combined with surface pollution. In addition, internal
resistive current surface leakage current may appear due to moisture
8 third harmonic current
penetration. During measurements, the surface currents
power loss
may interfere with the leakage current of the resistors.
6
However, the sensitivity to external and internal
surface currents may be different for the various
4
measurement methods. The content of third harmonic
2
in the surface leakage current is normally small
compared to the main frequency, which is beneficial to
0 the methods based on third harmonic analyses. The
-40 -20 0 20 40 60 80 100 120 140 influence of the external surface leakage current can
Temperature - °C
also be avoided, either by performing the
Figure 3.5 – Typical normalized temperature measurements in dry conditions, or by any other
dependence at Uc suitable method, e.g. bypassing the surface leakage
current to ground.
The sensitivity of the r.m.s, mean, and peak values of In practice, the accuracy is limited mainly by the
the total leakage current to variations in the resistive phase-shift of the reference signal and by the
current is illustrated in figure 4.1. The low sensitivity deviations in magnitude and phase of the voltage
to changes in the resistive current level makes the across the non-linear metal-oxide resistors at the
measurement of total leakage current suitable as a earthed end of the arrester. The presence of harmonics
diagnostic indicator only in the rare cases when the in the voltage may further reduce the accuracy of the
resistive current is in the same range as the capacitive method.
current.
A restriction on the method during measurements in
1,6
service is the need for a reference signal. Temporary
connection to the secondary side of a potential
transformer or to the capacitive tap of a bushing is
necessary and may be complicated to obtain. The
Relative increase in total leakage current
1,5
capacitive currents induced in the earth connection of
peak
the arrester by adjacent phases may reduce the
1,4 mean accuracy during measurements in service.
r.m.s.
5.2 Method- Compensating the capacitive
1,3
component by using a voltage signal
By using a voltage signal to compensate the leakage
1,2 current for its capacitive component, the sensitivity in
the measurement of the resistive part may be further
increased. The basic principle is a HV bridge where the
1,1 capacitive-resistive arm is adjusted to balance the
capacitive component of the leakage current so that
only the non-linear resistive part contributes to the
1,0
output voltage, which can be studied with the help of
1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0
an oscilloscope.
Relativeincreaseinresistiveleakagecurrent
According to IEC TC 37 / WG 10
of voltage or currents
phase shift in measurement
surface currents
Separate d.c. voltage source n.a. n.a. high high high limited 1
Service voltage or separate a.c. voltage source
Measurement of total leakage current low low mean low low extensive
Measurement of resistive leakage current
- using voltage reference 5.1 mean high high mean high limited 2
- using capacitor compensation 5.2 mean high high mean high limited 3
- using synthetic compensation 5.3 mean high high mean low n.i.a. 4
- using capacitive current cancellation 5.4 high high high low low limited 5,6
Harmonic analysis of leakage current
- using third harmonic 5.5 high low low mean low extensive 7
- using third harmonic with compensation 5.6 low low low high mean extensive 8,9
- using first order harmonic low high high mean high limited 10
Measurement of power loss 5.7 low high high mean high n.i.a. 11