CIRED 20th International Conference on Electricity Distribution
Paper 0393
THE TESTING OF VACUUM GENERATOR CIRCUIT BREAKERS
René SMEETS
Henk TE PASKE
Sander KUIVENHOVEN
KEMA T&D Testing – Netherlands
rene.smeets@kema.com
ABSTRACT
Testing of generator breakers is usually in accordance with
the standard IEEE Std. C37.013, to be included in the IEC
framework soon.
These tests are preferably carried out in direct test-circuits,
that supply current and voltage from one source. Above a
certain voltage level, in KEMA's case 17.5 kV, the
requirement of very high rate-of-rise of recovery voltages
needs a synthetic test circuit, normally exclusively applied
for high-voltage circuit breakers.
In such a circuit, a charged capacitor is used as the source
that supplies the transient recovery voltage. The
disadvantage of this method is that due to the gradual
discharge of the capacitors, after interruption the (power
frequency) voltage across the breaker cannot be kept
constant. Power frequency voltage of at least 300 ms dur-
ation after interruption is required in the general purpose
circuit breaker standard IEC 62271-100, in order to verify
absence of restrike(s) that may lead to failure to interrupt.
A new type of hybrid test-circuit is described, combining
the advantages of a direct circuit (adequate duration of
power frequency voltage) and a synthetic circuit (adequate
rate of rise of recovery voltage).
Testresults with a circuit for 24 kV 50 kA are summarized.
In this circuit, both transient recovery voltage and power
frequency recovery voltage are applied in full conformity
with the service situation.
I INTRODUCTION
Generator circuit breakers are normally located between
generator and step-up transformer in power plants. This
location (the vicinity of transformer and generator) calls for
a number a special requirements that generator breakers
have to fulfill differing greatly from "general purpose"
circuit breakers. Apart from the very high nominal current
carrying capacity, the faults that have to be dealt with by
generator breakers have a special nature. This is notably
clear in the voltage that appears across the breaker
immediately after interruption of the fault current. This
voltage is generally known as Transient Recovery Voltage
(TRV). The capability of a breaker to withstand this TRV is
a prime requirement to pass a test. The key parameters of
CIRED2009 Session 1 Paper No 0393
Prague, 8-11 June 2009
Jürgen JÄGER*
Nils ANGER**
Siemens – Germany
*
jay-jay43@web.de
**
nils.anger@siemens.com
the TRV are its rate-of-rise (RRRV) and peak value. Test-
circuits have to be designed to reproduce the specified
TRV. After decay of the transient (recovery) voltage, the
normal power frequency voltage remains across the breaker.
This contribution emphasizes the importance of not only
reproducing (in testing) the transient recovery voltage, but
also the power frequency recovery voltage (RV) that
remains across the breaker after the TRV has disappeared.
Reproduction in testing of both TRV and RV is clearly
acknowledged in the (IEC) standard for general purpose
circuit breakers [1], but it is absent and thus not required in
the (IEEE) standard for generator circuit breakers [2].
This is not in conformity with the service situation, where
the TRV is always followed by the RV.
II GENERATOR CIRCUIT BREAKER FAULT
SITUATIONS
The following faults are known to have a far more severe
impact on the breaker than faults in a standard distribution
network:
1. The system source (SS) fault (fault between generator
breaker and generator) produces very steep transformer-
dominated transient recovery voltages (TRV). The rate of
rise of this TRV can reach extreme values due to the low
capacitance of the transformer and the short connections.
2. Generator source (GS) faults (faults between generator
breaker and transformer) can result in missing current
zeroes that prolong the arcing time. Missing current
zeroes arise because of the combination of various
reactances and time constants of the generator, causing a
DC component that can be temporarily higher than the
AC component of the fault current.
3. In addition, also the out-of-phase (OoP) requirements are
much more severe than the out-of-phase requirements of
general purpose circuit breakers (higher current and
steeper TRV).
III STANDARDISATION STATUS
The IEC standard 62271-100 [1] on circuit breakers
explicitly excludes generator circuit breakers. Therefore, the
IEEE Std. C37-013 [2, 3] is generally accepted for this
CIRED 20th International Conference on Electricity Distribution Prague, 8-11 June 2009

Paper 0393

60 OOP fault
sense, direct testing (especially three-phase) is the best
SS fault envelope for general simulation of network reality. Distribution breakers are
50
purpose CB T100 normally tested by this method and in three-phase test
GS fault
40 circuits voltage stresses are like in fig. 2a.
TRV (kV)

30 50
transient recovery voltage
20 40
envelope for general temporary pow freq voltage for first-pole-to-clear
10 purpose CB OOP 30
permanent power frequency recovery voltage
0 20
0 20 40 60 80 100 120
time (us) 10

0
Fig. 1: Standardized values of TRV of generator (for 200-
400 MVA transformers, red ) - and general purpose circuit -10
breaker (blue). Rated voltage is 24 kV. -20
reality & direct three-phase testing

purpose, and will be included in the IEC framework in 2009 0 50 100 150 200 250 300
by a joint IEEE / IEC working group. 50

40 transient recovery voltage

III 1 Transient recovery voltage 30 DC revovery voltage
In fig. 1, (IEEE) standardized values of TRV's of 24 kV
generator breakers for step-up transformer rated 200-400 20

MVA are compared with (IEC [1]) standardized 24 kV 10

breakers for distribution systems. In the generator breaker
0
case, IEEE prescribes the (sinusoidal) TRV waveshapes
directly: in fig. 1 the TRVs associated with system source -10

(SS), generator source (GS) and out-of-phase (OOP) fault. synthetic testing with unipolar recovery
-20
In the general purpose breaker case, IEC prescribes only the 0 50 100 150 200 250 300
required TRV envelopes: in fig. 1 the terminal fault (100% 50
rated short-circuit breaking current) and out-of-phase faults transient recovery voltage
40
(25% rated short-circuit breaking current).
From the figure it is immediately clear that the rate-of-rise 30

depends greatly on the application: either generator- or oscillatory recovery voltage

20
distribution application.
10

III 2 Power frequency recovery voltage 0

The pole of the breaker most severely stressed is the first-
pole-to-clear. The power frequency recovery voltage across
this pole has temporarily a value of 1.5 times the phase-to-
ground voltage (in medium voltage networks). The elevated
voltage lasts until all phases have interrupted (after a quarter
cycle) the current. After this temporarily elevated power
frequency voltage, the rated power frequency voltage (at
phase-to-ground value) sets in and remains.
In fig. 2a, all components of voltage across a breaker (in
this case a 24 kV generator breaker) are outlined.
III 2.1 Direct tests
For general purpose circuit breakers, IEC 62271-100 clause
6.104.7 [1] requires the presence of at least 95% of the
(phase-to-ground) power frequency voltage until 300 ms
after interruption. This is to verify the dielectrical withstand
of the breaker also beyond the TRV region, where the
switchgear has not yet reached a mechanically (and possible
dielectrically) stable situation.
In direct test circuits, where current and voltage are supplied
from the same source, this is a natural situation. In this
-10
-20 synthetic testing with oscillatory recovery
0 50 100 150 200 250 300
Fig. 2: Recovery voltage across breaker:
a: upper: in network and in three-phase direct testing
b: middle: in synthetic testing with unipolar recovery voltage
c: lower: in synthetic testing with damped oscillatory recovery
III 2.2 Synthetic tests
In synthetic test circuits, normally applied for high-voltage
circuit breaker testing, current is normally supplied by a
generator (or power grid, or capacitor bank) and recovery
voltage is supplied by a pre-charged capacitor bank.
Whereas such a capacitor bank is very well suited to supply
the adequate TRV, it cannot maintain an adequate recovery
voltage for a sufficiently long time.
Basically, there are two possibilities:
1) Maintain the remaining capacitor bank voltage as a
unipolar (DC) voltage across the breaker. This is
unrealistic, and stresses the test-object in a way very

CIRED2009 Session X Paper No 0393 Page 2 / 4

CIRED 20th International Conference on Electricity Distribution
different from reality. This situation is shown in fig. 2b.
2) Realization of a damped oscillatory voltage by applying
a large inductor across the capacitor bank. Strong
damping can not be avoided in the circuit, even when
inductors with a high quality factor are applied.
In both cases, for general purpose circuit breakers, IEC
62271-101 [4] clause 4.1.3 allows recovery voltage to drop
not lower than 50% of the rated phase-to-ground voltage
after 100 ms. Although this is also not realistic, it is a
generally accepted compromise.
IV GENERATOR BREAKER TESTING
IV 1 Direct testing
Since the transient recovery voltage for generator breakers
is much steeper than for general purpose breakers, testing of
these breakers is not straightforward. Conventional direct
test-circuits are normally designed for distribution breakers
and may have problems with these steep TRVs. This is
mainly due to the relatively large capacitance of the supply
transformers, that limits the upper frequency of transient
recovery voltages. For KEMA, generator breaker TRVs can
be produced in direct three-phase test circuits roughly up to
17.5 kV rated voltage. This was demonstrated with a
vacuum generator breaker rated 17.5 kV successfully tested
up to 80 kA [5]. In fig. 3, the result (of one pole of such a
breaker) is demonstrated. In this case, test current with an
RMS value of 80.1 kA with an asymmetry of 53% reached a
peak value of 230 kA.
IV 2 Synthetic three-phase testing
In this case KEMA’s synthetic installation is added to the
high-current circuit. Hereby, the appropriate values of TRV
can be produced using the current injection method.
KEMA’s synthetic installation is a double LC circuit
(energy 2*1.7 MJ), that enables three-phase synthetic
testing. One LC circuit applies the adequate TRV on the
first pole-to-clear- and the other circuit applies TRV across
the last clearing poles.
IV 3 Single-phase synthetic testing
This method is used for testing the larger SF6 and airblast
units typically 25 kV with 100 – 120 kA (for generators in
the order of 1000 MVA). The main problem is in the
magnitude of current involved (up to 600 kApk). This
requires very tight control of the extreme forces on the
conductors supplying the main current. Specially designed
heavy buswork is in use, yet allowing a certain degree of
flexibility. This is connected to the main 400 kA busbar
system of the laboratory which is directly connected to the
current source (four generators in parallel).
In these tests one synthetic unit supplies TRV to the first
clearing pole, the other unit is used for arc prolongation. A
method for arc prolongation is developed that can produce
far higher current than the arc prolongation circuits
normally used in high-voltage synthetic testing. Such a
CIRED2009 Session X Paper No 0393
Prague, 8-11 June 2009
Paper 0393
current: 50 kA/div
1 0 0
cs
voltage: 10 kV/div
5 0
time: 20 ms/div
0
- 5 0
- 1 0 0
TRV
- 1 5 0
- 2 0 0
current
- 8 0 - 6 0 - 4 0 - 2 0 0 2 0 4 0 6 0 8 0
Fig.3: Interruption of 80.1 kARMS current with 53% asymmetry
by a vacuum generator circuit breaker. cs: contact separation.
The figure shows one phase of a three-phase current.
circuit is necessary to produce a realistic arc duration in
synthetic tests.
Basically, three-phase- or single phase currents are applied:
1) Three-phase current tests correctly represent the dynamic
interaction between the phases as well as between
extinction chambers and mechanism. Experience at
KEMA with this circuit is up to 120 kA at 25.3 kV [6].
2) Single phase current tests. In case the circuit breaker
poles share a single common drive, verification of the
influence of the three-phase current and arc on the contact
motion is necessary. This is done by comparison of the
mechanical contact travel characteristics recorded during
direct three-phase make and break tests at full current
with these characteristics during the single phase test.
This philosophy is taken from IEC 62271-100 and is not
described in ANSI/ IEEE. Experience at KEMA with this
circuit is up to 210 kA at 25.3 kV [7].
IV 4 Hybrid direct - synthetic testing
In order to fulfil Siemens' requirements of producing both
generator breaker TRV as well as a constant power
frequency recovery voltage, a new type of (single phase)
circuit is developed and tested that combines the advantages
of direct and synthetic circuits. It has the capabilities to:
- produce the required fault current;
- produce TRV's of steepness adequate for generator
breakers as in IEEE C37.013;
- produce realistic, constant power frequency recovery
voltages (not specified in IEEE, but specified in IEC);
- independent control of current, TRV and RV voltages in
development tests;
- realistic simulation of events occuring at restrike. Upon
restrike, upto the full fault current is provided by the
current circuit.
At the same time, a number of challenges had to be met, the
most severe one dealing with arc prolongation. Arc pro-
longation is necessary in synthetic testing, because, unlike in
direct test circuits, the adequate TRV can only be applied at
Page 3 / 4
 CIRED 20th International Conference on Electricity Distribution Prague, 8-11 June 2009

Paper 0393

80
voltage [20 kV/div] current [20 kA/div]

V SUMMARY AND CONCLUSIONS

60 transient recovery voltage

40 power frequency recovery voltage

At higher rated voltages (> 15 - 17 kV), in order to obtain
the required very steep transient recovery voltage generator
20
breakers can only be tested in synthetic circuits. Standard
0 synthetic circuits do not have the opportunity to supply
-20 power frequency recovery voltage as in reality. To
-40 recovery voltage overcome this deficiency, a hybrid test circuit has been
current
developed, able to provide both transient - as well as power
-60
frequency recovery voltage.
-80
250 300 350 400 450 500 550 600 650 700 Now it is possible to test with voltages in conformity both
time [50 ms/div] with the generator breaker standard regarding transient
50 recovery voltage as well as with the general purpose breaker
transient recovery voltage standard regarding power frequency recovery voltage.
40 This guarantees full equivalence with reality.
TRV [10 kV/div]

30
REFERENCES
20
[1] IEC 62271-100, High-voltage switchgear and controlgear –
10
Part 100: Alternating-current circuit-breakers, 2008
0
[2] IEEE Std. C37.013-1997: "Standard for AC-High Voltage
Generator Circuit Breakers Rated on a Symmetrical Current
-10 Basis", IEEE New York
341.9 341.92 341.94 341.96 341.98 342
[3] IEEE PC37.013a/D6, Nov. 2006, "Standard for AC High
time [20 us/div]
Voltage Generator Circuit Breakers Rated on a Symmetrical
Fig.4: Result of test with vacuum generator circuit breaker (24 kV Current Basis - Draft Amendment 1: Supplement for use with
50 kA) in accordance with IEEE C37.13 and continuous power Generators Rated 10 - 100 MVA", IEEE New York
frequency recovery voltage. Upper: Full interruption, incl. current [4] IEC 62271-101 High-voltage switchgear and controlgear –
and 60 Hz recovery voltage lower: transient recovery voltage Part 101: Synthetic testing, 2006
[5] R. Smeets, J. Hooijmans, J. Jäger, N. Anger, Test experiences
with a new generator circuit breaker based on vacuum
a single current zero. This implies that a suitable method technology, CIRED Conference 2007, paper 0383.
must be applied to test with the same arc duration as [6] R.P.P. Smeets, W.A. van der Linden, 1998, "The Testing of
expected in reality. Normally a so-called Re-Ignition Circuit SF6 Generator Circuit-Breakers", IEEE Trans. Pow. Del.,
(RIC) is applied, forcing a reignition in the breaker on a vol.13, no.4, pp. 1188 - 1193
microsecond scale. Traditional RIC’s that are commonly [7] R.P.P. Smeets, H.D. Barts, L. Zehnder, 2006, "Extreme
used at high-voltage SF6 circuit breaker testing rely on a Stresses of Generator Circuit Breakers", CIGRE Conference,
paper A3-306
charged capacitor that releases its energy by means of a
steep current pulse just before the moment the breaker
would interrupt the test current. The zero crossing of the
current is now forced by this current, that is much steeper
(several hundred A/µs) compared to the decline of the
power frequency short-circuit current (few tens A/µs).
Vacuum circuit breakers, however, have such an excellent
interrupting capability that they can interrupt even the steep
current pulse of the RIC. An adjustable and modified RIC
has been used to gain experience in this field. In com-
bination with a high-frequency and high-resolution
measurement the critical parameters that result in a reliable
reignition could be identified.

A result of such a test is shown in fig. 4. The test-object is

shown in fig. 5. An interesting feature for development
testing is the possibility to control current, TRV and RV
completely independent from each other. In one of such
single phase tests, power frequency recovery voltage up to
26.1 kV did not result in a restrike after interruption of 51.5
kA.
Fig. 5: Siemens vacuum generator breaker 3AH3727-8 at KEMA

CIRED2009 Session X Paper No 0393 Page 4 / 4