2. INTRODUCTION
Metal Oxide Surge Arrester has been in service since 1976.
They protect major electrical equipment from damage by
limiting overvoltage and dissipating the associated energy
Metal Oxide surge Arrester should be highly reliable in
most application because the metal oxide surge arrester are
less fail than silicon carbide arrester as a result of moisture
ingress and contamination.
Experience has shown that most failures of silicon carbide
arrester occur because gap spark over.
2
3. These failures have been reduced using Metal oxide surge
arrester, by moisture ingress or contamination, to a level
low enough to permit repetitive spark over at or near
normal operating voltage.
Gaps are not required in metal oxide surge arrester.
The elimination of gaps results in an increased probability
of failure of metal oxide surge arrester on system
overvoltage lower than normally required to cause the
spark over of gapped arresters.
metal oxide surge arresters might be more vulnerable to
high energy lightning surge than silicon carbide arresters.
3
4. THE ADVANTAGE OF METAL OXIDE SURGE
ARRESTER IN COMPARISON WITH SILICON
CARBIDE GAPPED ARRESTER
Simplicity of design, which improves overall quality
and decreases moisture ingress.
Easier to maintenance especially to clear arc.
Increased energy absorption capability.
4
5. DISADVANTAGE OF METAL OXIDE
SURGE ARRESTER
Voltage is continually resident across the metal oxide and
produces a current of about one milli ampere. While this
low-magnitude current is not detrimental.
Higher currents, resulting from excursions of the normal
power frequency voltage or from temporary over voltages
such as from faults or ferroresonance, produce heating in
the metal oxide.
If the temporary overvoltage are sufficiently large in
magnitude are long in duration, temperatures may increase
sufficiently so that thermal runaway and failure occur
5
6. Operating analysis of metal oxide surge arrester, IEEE
and Pinceti model, using ATP-EMTP is done.
The lightning arrester models base on the Ohio Brass
PVR 221617 21kV and Precise PAZ-P09-1 9 kV 10kA
class 1 by applying standard impulse current wave
(8/20 µsec) is being used.
The models can predict the operation of the metal
oxide surge arrester in the system within 10% errors.
6
7. IEEE MODEL
The model recommended by IEEE W.G 3.4.11 .
In this model the non-linear V-I characteristic is
obtained.
For slow surges the filter impedance is extremely low
and A0 and A1 are practically connected in parallel
7
8. On the contrary, during fast surges, the impedance of
the filter becomes significant, and causes a current
distribution between the two branches.
For precision sake, the current through the branch A0
rises when the front duration decreases.
A0 resistance is greater than A1 resistance for any
given current, the faster the current surge, the higher
the residual voltage.
8
9. The comparison of the calculated peak values with the
measured values shows that the frequency dependent
model gives accurate results for discharge currents
with times to crest between about 0.5 µs and 45 µs.
The main problem of this model is how to identify its
parameters. The W.G.3.4.11 suggests an iterative
procedure where corrections on different elements are
necessary until a satisfactory behavior is obtained.
The starting values can be obtained through formulas
that take into account both the electrical data (residual
voltages), and the physical parameters (overall height,
block diameter, columns number)
9
10. The inductance L1 and the resistance R1 of the model
comprise the filter between the two non-linear
resistances. The formulas for these two parameters are
L1 = 15 d/n μH
R1 = 65 d/n Ω and for ,
L0 = 0.2 d/n μH
R0 = 100 d/n Ω
C = 100 n/d pF
10
11. PENCETI MODEL
The operating principle is quite similar to that of the
IEEE frequency-dependent model.
11
12. PINCETI MODEL
The model here presented derives from the standard
model, with some minor differences. By comparing the
frequency dependent model and pinceti model, it can
noted that :-
The capacitance is eliminated, since its effects on
model behavior is negligible.
The two resistances in parallel with the inductances are
replaced by one resistance R (about 1 MΩ) between
the input terminals, with the only scope to avoid
numerical troubles.
12
13. STATIC CHARACTERSTICS OF THE
NON LINEAR ELEMENT
These curve derives
from the curves
proposed by IEEE
W.G.3.4.11, and are
referred to the peak
value of the residual
voltage measured during
a discharge test with a
10 kA lightning current
impulse.
13
14. To define the inductances, the following equations can be
used :-
L1 = ¼ [(V r1/T2 –V r8/20) ÷V r8/20 ] * Vn µH
L0= 1/12 [ (V r1/T2 –V r8/T2 )÷V r8/20]* Vn µH
14
15. DATA USING IN THE MODEL (PRECISE
PAZ-P09-1-9KV 10KA CLASS 1)
IEEE Model
15
19. MODELING RESULTS OF PRECISE
PAZ – P09-1
Modeling results and errors in each model prepare
With manufacture’s data sheets.
19
20. Percent error of each model prepare with
manufacture’s data sheets.
20
21. IEEE model has errors higher than Pinceti model in the
high current testing more than 10 kA with standard
wave front (8/20 µs.)
When testing with the steep wave front, Pinceti model
has errors higher than IEEE model.
for the switching impulse testing, the percentage errors
are equal.
21
22. MODELING RESULTS OF OHIO BRASS
PVR 221617
Modeling results and errors in each model prepare
with manufacture’s data sheets.
22
23. Percent error of each model prepare with
manufacture’s data sheets.
23
24. IEEE model has errors higher than Pinceti model both
for the standard wave front (8/20 µs.) testing and
switching impulse testing.
When testing with the steep wave front, Pinceti model
has errors higher than IEEE model.
24
25. CONCLUSION
The percentage errors from the modeling result can be
describes as follows :
1. At standard wave front percentage errors of IEEE model
is higher than Pinceti model.
2. At steep wave front percentage errors of IEEE model is
less than Pinceti model.
3. At switching overvoltage condition percentage errors of
IEEE model is nearly the same as Pinceti model.
25
26. Both models can predict the operation of the metal
oxide surge arrester in the system.
There is less than 5% errors for standard wave front
(8/20 µs.) at 5-20 kA .
less than 10% errors for standard wave front (8/20 µs.)
at others.
less than 10% errors for steep wave front and less than
14% for switching impulse testing.
26
27. REFERENCE
IEEE Working Group 3.4.11 Application of Surge Protective Devices Subcommittee
Surge Protective Devices committee, ”Modeling of Metal Oxide Surge Arrester,”
IEEE Transaction on Power Delivery, Vol.7, No.1, pp 302-309, January 1992.
P. Pinceti , M. Giannttoni, “A simplified model for zinc oxide surge arrester,” IEEE
Transaction on Power Delivery, Vol. 14, No.2, pp 393-398, April 1999.
E.C. Sakshaug, J.J. Bruke, J.S. Kresge, ”Metal Oxide Arrester on Distribution
system Fundamental Considerations,” IEEE Transaction on Power Delivery, Vol.4,
No. 4, pp 2076-2089, October 1989.
Ikmo Kim, Toshihisa Funabashi, Haruo Sasaki, Toyohisa Hagiwara, Misao
Kobayashi, “A study of ZnO Arrester Model for Steep Front wave,” IEEE
Transaction on Power Delivery, Vol.11, No. 2, pp 834841 April 1996.
Andrew R. Hileman ,”Insulation Coordination for Power systems,”Marcel
Dekker,Inc., New York Basel ,1999.
27