Page 1 of 6 The Journal of Engineering

This article has been accepted for publication in a future issue of this journal, but has not been fully edited.
Content may change prior to final publication in an issue of the journal. To cite the paper please use the doi provided on the Digital Library page.

A Comparison of CT Secondary Open Circuit Protection System

Technologies
J. Robertson
M&I Materials Ltd, Hibernia Way,
Trafford Park, Manchester M32 0ZD
[email protected]

Keywords: CT Secondary Open Circuit Protection, Silicon Current Transformer

Carbide Varistor, Gas Discharge Tube, Metrosil.
CT O/C
Abstract Protection
Device

Should a current transformer (CT) secondary winding be left

in or inadvertently become open circuited whilst the primary
is energised, then damage may occur to the CT - leaving it
Burden (Relay, etc)
inoperable or seriously effecting it’s measurement accuracy.
If the CTs are being utilised in particularly difficult places to Figure 1: Arrangement of CT Open Circuit Protection
access, such as in a gas insulated substation or in a power Devices in a Secondary Circuit
transformer bushing, this can then lead to an expensive repair
to a relatively inexpensive component. To prevent damage
occurring to open circuited CT’s, a range of products are Two technologies exist on the market to prevent damage to the
available. These devices are based either on a silicon carbide CT’s should they become open circuited The first of these is a
varistor based technology or on a gas discharge tube based silicon carbide varistor based device [9]. The second is a gas
discharge tube based device [10].
technology. The purpose of this paper is to compare the
performance of these devices. Figure 1 shows the arrangement of a CT open circuit
protection device in a secondary circuit. In terms of the
1. Introduction physical location of the device, it is important to locate it as
close to the CT as possible, to prevent it from being removed
As is widely known, subjecting a current transformer’s with the CT burden, if it is being subject to maintenance.
secondary terminals to an open circuit condition, whilst the
primary is energised, can lead to damage to the CT. Two During normal operating and fault conditions, with the CT
common forms of damage exist. The first is through the burden present, the device is designed to act in a passive
breakdown of the CT’s secondary insulation, due to the manner, drawing little leakage current. When the CT is in an
emergence of high voltage transients. Particularly when CTs open circuited condition, the device then acts to limit the
have high knee point voltages, the magnitude of these voltage across the secondary terminals of the CT.
transients can exceed 10kV across the secondary terminals -
The focus of this paper is to technically compare the
even when low currents are passing through the primary. The
performance of silicon carbide varistor based devices with the
second occurs because of changes in the magnetisation
performance of gas discharge tube based devices.
characteristics of the CT due to heating effects in the CT core,
through high eddy currents [1-8]. The two devices being compared are designed to limit the
voltage across an open circuit CT to around 2kV when subject
Should a CT be located in a particularly difficult location to
to a secondary current of 1A. Both devices are designed to be
access, such as in a gas insulated substation or a power
used in conjunction with high knee point CTs (as are often
transformer bushing, this can lead to a very costly repair.
required for high impedance relay systems).
The requirements for CTs not to be exposed to voltages of
Both these devices have been analysed in terms of their
3000Vpk during their operation are now stipulated by many of
performance whilst:
the utilities, which is also linked with the requirements of IEC
60044. a.) Acting as a passive component in the CT circuit when
a burden is present across the secondary of the CT.
To prevent these high voltages evolving across a CT
secondary, the most direct approach is to simply short out the b.) Acting as an open circuit protection device when the
secondary terminals of the CT. However, if the CT is left in an CT is left in an open circuited condition.
open circuited condition during commissioning, or the CT
inadvertently becomes open circuited during operation, then a
smarter approach is required.

1
IET Review Copy Only
 The Journal of Engineering Page 2 of 6
This article has been accepted for publication in a future issue of this journal, but has not been fully edited.
Content may change prior to final publication in an issue of the journal. To cite the paper please use the doi provided on the Digital Library page.

3150A /
1600A
A

T
1A
T1 = 3150A

S1 Varistor or
T2 = 1600A CT O/C
240V 3150 / T V Gas

(125A)
v LV
1600 : 1 CT Taps Scope R1
Relay
Discharge
Protection
Device
3150A Burden
Tube

I Resistor I
Probe Probe

A A
Scope Scope

Figure 2: Test Circuit

3. Results
2. Experimental Method
3.1 CT Open Circuit Protection Devices Whilst Activing
in a Passive Manner & Subject to Applied Voltages
Experimental analysis of the CT open circuit protection
devices has been performed using the test circuit shown in
Figure 2. The test circuit comprises a low voltage, high 3.1.1 Silicon Carbide Varistor Devices
current, 30kVA transformer - supplying a range of currents
to the primary of a 1A class X, CT with a ratio of 3150:1. As Figure 3 shows typical current and voltage waveforms,
a high knee point CT has been used in these investigations, which exist when the silicon carbide varistor based device is
no saturation effects are evident under the test conditions. used in the circuit of Figure 2 and switch S1 is closed.

The secondary circuit of the CT supplies current to a 1kΩ From Figure 3, it is apparent that the currents, which flow
resistor, marked R1, which can be switched in and out through the CT open circuit protection device, are non-
remotely via contactor, marked S1. The CT open circuit sinusoidal in nature and the magnitude of the peak current
protection device is connected in parallel to R1 and the increases at a greater rate to the applied voltage. The peak
contactor S1. AC current and voltage values follow the characteristics of
the varistor discs when evaluated under DC and impulse test
To simulate conditions where the CT open circuit protection conditions.
device is connected across a CT with the burden present,
switch S1 is left in a closed position. Under these conditions
the open circuit protection device is subject to an applied
sinusoidal voltage waveform, the magnitude of which is
determined from the magnitude of the current flowing
through the secondary of the CT and the resistance of R 1.
The magnitude of the leakage current has been determined
for each device under an array of applied voltages.
To simulate conditions where the CT open circuit protection
device is connected across an open circuited CT, switch S1 is
opened. Under these conditions the CT open circuit
protection device is subject to an applied sinusoidal current
Figure 3: Typical AC Waveforms of the Silicon Carbide
waveform, the magnitude of which is equal to the CT
Based Device when Subject to an Applied Voltage
secondary current. In the operation of both of these devices,
the current flowing through the open circuit protection
Figure 4 shows the true Root Mean Square (RMS) AC
device, whilst it is limiting the voltage in the secondary
leakage current values at a range of applied voltages for an
circuit, then causes it to heat up. On reaching a
array of silicon carbide based devices. These values have
predetermined temperature, a thermostatic switch then shorts
been determined through an integral technique from the
the terminals of the CT secondary. To monitor the switching
current and voltage waveforms. It should be noted that unless
times of the open circuit protection device, a timer T has
true RMS values are used, as opposed to pk/√2 values, this
been used in the circuit with a secondary thermostatic switch.
can give misleading information on whether the
A temperature logger, using a class K thermocouple, has
characteristics conform to the spec limits due to the non-
been used to monitor the temperature cycles.
sinusoidal currents.
The performance of these devices has been monitored over a
total of 100 heat cycles, whilst subject to a secondary current
of 1A.

2
IET Review Copy Only
Page 3 of 6 The Journal of Engineering
This article has been accepted for publication in a future issue of this journal, but has not been fully edited.
Content may change prior to final publication in an issue of the journal. To cite the paper please use the doi provided on the Digital Library page.

If the voltage across the terminals of the gas discharge tube

device exceeds a voltage of between 2-2.2kVpk, the tubes
then fire. The entire secondary current then passes through
the gas discharge tube device, and forms an effective short
circuit across the terminals of the CT. Typical waveforms are
shown in Figure 7. It can be shown that the tube fires on the
rising edge of the sinusoidal voltage waveform and
extinguishes when the waveform passes through a zero
voltage crossing. The voltage at which the discharge tube
fires during subsequent AC cycles is not consistent and
differs between the positive and negative half cycles.
Figure 4: Relationship Between RMS Current and
Voltage Test Values for an Array of Silicon Carbide
Based Varistor Devices
3.1.2. Gas Discharge Tube Devices
Figure 5 shows typical current and voltage waveforms for a
gas discharge tube device when used in the circuit of Figure
2, with switch S1 closed. From Figure 5, it is apparent that
when subject to an applied sinusoidal waveform, the
currents, which flow through the CT open circuit protection
device, are small and reactive in nature.

Figure 7: Typical Current and Voltage Waveforms of

the Gas Discharge Tube when Tested Above 2.2 kVpk
3.2 CT Open Circuit Protection Devices Whilst
Protecting an Open Circuited CT & Subject to
Applied Currents
3.2.1. Silicon Carbide Varistor Devices

For the silicon carbide based varistor device, the voltage

across the open circuited CT is clamped to a level
Figure 5: Typical Leakage Current and Voltage determined from the current in the CT secondary and the V-I
Waveforms in the Gas Discharge Tube Device characteristics of the varistor.

Table 1 shows data on a range of silicon carbide varistor

Figure 6 shows the peak AC leakage current values at an
based devices tested during a single heat cycle whilst being
array of applied voltages. The current is approximately
exposed to a 1A open circuit event. It can be seen from this
proportional to the voltage.
table that the voltages across the CT are limited to less than
2.25kVpk. and the switching time is around 40 secs.

Figure 6: Peak AC Leakage Current and Voltage Values

for an Array of Gas Discharge Tube Devices.

Table 1: Performance of Silicon Carbide Based Varistor

Devices Under 1A CT Open Circuit Conditions

3
IET Review Copy Only
 The Journal of Engineering Page 4 of 6
This article has been accepted for publication in a future issue of this journal, but has not been fully edited.
Content may change prior to final publication in an issue of the journal. To cite the paper please use the doi provided on the Digital Library page.

Figure 8 shows typical current and voltage waveforms for

the silicon carbide based varistor device during CT open
circuit conditions. It is apparent that when subject to an
applied sinusoidal current waveform, the voltages which
appear across the CT open circuit protection device are non-
sinusoidal in nature.

Figure 10: Variation in the Clamping Voltage During 100

Heat Cycles for the Silicon Carbide Varistor Based
Device

Figure 8: Typical Current and Voltage Waveforms in a

Silicon Carbide Varistor Based Device Under Open
Circuit Conditions
Figure 9 shows typical heat cycles for a silicon carbide
varistor based device during a number of 1A open circuit
events. As supported by Table 1 for a 1A current, the heating
phase is in the region of 37 secs, whilst the cooling phase is
in the region of 30 minutes, assuming an ambient temp of 17
ºC. The peak temperature was between 120 ºC and 141 ºC
Figure 11: Variation in Switching Temp During the 100
during the heating phase and cooled to 60 ºC before the
Heat Cycles for the Silicon Carbide Based Device
thermostatic switches reopened. The temperature rise during
the first heat cycle is much larger than subsequent heat
3.2.2 Gas Discharge Tube Devices
cycles, since it is being heated from room temperature.
For the gas discharge tube based device, the clamping
voltage during the open circuit event is determined from the
breakdown strength of the spark gap in the gas discharge
tube.

Table 2 shows data on a range of gas discharge tube devices

obtained during a single 1A open circuit event.

Figure 9: Heat Cycles of the Silicon Carbide Varistor

Based Device

The performance of three silicon carbide varistor based

devices was monitored over a total of 100 heat cycles, whilst
subject to 1A in the secondary circuit.

Figure 10 shows the variation in the clamping voltage over

the 100 heat cycles; likewise, Figure 11 shows the variation
in switch closing times of the thermostatic switches over 100 Table 2: Performance of the Gas Discharge Tube Device
heat cycles. The high peak temperatures in heat cycles 1, 10, Under CT Open Circuit Conditions
20, etc. are again due to the heat cycle commencing at
ambient temperature, instead of 60 ºC, as previously
discussed for Figure 9.

4
IET Review Copy Only
Page 5 of 6 The Journal of Engineering
This article has been accepted for publication in a future issue of this journal, but has not been fully edited.
Content may change prior to final publication in an issue of the journal. To cite the paper please use the doi provided on the Digital Library page.

Figure 12 shows typical current and voltage waveforms, The performance of a number of gas discharge tube devices
which exist during a 1A open circuit event. was monitored over a number of heat cycles, whilst subject
to 1A in the secondary circuit.

As the clamping voltage varied significantly within a single

heat cycle, due to the flickering of the gas discharge tube,
data has not been included on how the clamping voltages of
the devices vary over the number of heat cycles. Figure 15
shows the variation in switching times of the gas discharge
tube devices with the number of heat cycles.

Figure 12: Typical Current and Voltage Waveforms in a

Gas Discharge Tube Based Device under Open Circuit
Conditions

Despite the waveforms showing a 60Vpk clamping voltage, it

was found that these waveforms were not stable and
fluctuated with the flickering of the gas discharge tube, as
illustrated in Figure 13. It was also observed that the
flickering of the gas discharge tube generated a high degree Figure 15: Variation in Switching Temp of the Gas
of radio frequency noise, which interfered with the Discharge Over a Number of Heat Cycles.
instrumentation during the tests. This could create issues
with EMC compliance of the devices and could potentially Despite attempting to subject the gas discharge tubes to 100
interfere with signals to protection relays and / or switchgear. heat cycles; testing was aborted after a number of heat
cycles, due to the tubes firing at a voltage below 1.5kV.
Table 3 summarises the total number of heat cycles each
device was subject to.
Gas Discharge Number of
Tube Device Heat Cycles
A 42
B 42
C 21
Figure 13: Arcing in a Gas Discharge Tube Device D 24
E 67
Figure 14 shows typical heat cycles for the gas discharge
tube based device, during a 1A open circuit event. As is Table 3: Number of Heat Cycles Performed Before
supported by Table 2, for a 1A current, the heating phase is Aborting Testing of the Gas Discharge Tube Devices
in the region of 6 minutes, whilst the cooling phase is in the
Further tests were performed on two gas discharge tube
region of 20 minutes, assuming an ambient temp of 17 ºC.
devices to establish how the firing voltage changes over the
The peak temperature obtained during the heat cycles was
number of heat cycles. Figure 16 illustrates how the
between 100 ºC and 120 ºC during the heating phase, after
flashover voltage appears to decrease further with the
which the thermostatic switch closed and cooled to 40 ºC
number of heat cycles performed.
before reopening.

Figure 14: A Typical Heat Cycle for a Gas Discharge Figure 16: Variation in Gas Discharge Tube AC
Tube Based Device Flashover Voltage Across 100 Heat Cycles

5
IET Review Copy Only
 The Journal of Engineering Page 6 of 6
This article has been accepted for publication in a future issue of this journal, but has not been fully edited.
Content may change prior to final publication in an issue of the journal. To cite the paper please use the doi provided on the Digital Library page.

4. Discussion 5. Conclusion

In an unaged state, when acting as a passive device and where This paper reports on several investigations to compare the
voltages of less than 2 - 2.2kVpk are present across the performance of two types of devices designed to protect
terminals of the CT, the gas discharge tube based devices current transformers from damage if subject to an open circuit
draw a very low leakage current. As such, considerations condition. Such types of protection systems are used with
towards power dissipation in the selection of the device do high knee point CTs, often in conjunction with high
not have to be made. impedance relay systems. They are also commonly used
where the CT is difficult and expensive to replace, such as in
For the silicon carbide based device, appropriate calculations gas insulated substations (GIS) and bushing CTs. The two
of leakage current and power dissipation need to be made, as types of device compared consisted of a silicon carbide
well as how the device will share current with any other varistor based technology and a gas discharge tube based
voltage limiting devices used in a high impedance relay technology. The devices have been analysed in terms of their
system. performance when acting as a passive device with a burden
connected across a CT as well as their performance when
It has been shown that if subject to a voltage of greater than acting to limit the voltage across an open circuited CT.
2.2kVpk, the gas discharge tube based device forms an
effective short circuit across the CT. If the CT was then being It can be concluded that the silicon carbide based device is the
used in a high impedance relay system, such voltages may be superior choice for the reliable protection of current
present during fault conditions. Short circuiting of the CT transformers during open circuit conditions. This is due to:
may then influence / prevent the operation of the protection
systems used in conjunction with the CT. • The silicon carbide varistor based device can be reliably
used with protection systems, where the voltage across
The gas discharge tube based devices can quickly become the CT may exceed 2kVpk, without effecting the
aged when connected across an open circuited CT. The operation of the system
ageing effects reduce the firing voltage, which could then o This may be the scenario where a class X CT is
further influence the operation of protection devices being used in a high impedance differential relay system
used with the CT. during a fault event

The silicon carbide varistor based devices are not aged by • The silicon carbide varistor based device does not age
by being subject to a series of open circuit heat cycles
being subject to a sequence of open circuit events,
maintaining consistent operation following 100 heat cycles. • The silicon carbide varistor based device can be used in
As such, the device will resume in acting in a passive manner environments with ambient temperature of greater than
once the burden is replaced across the secondary terminals of 40C
the CT, both under normal operating and fault conditions.
• The silicon carbide varistor based device does not emit
radio frequency noise, which could influence the
The thermostat reopening temperature of the gas discharge operation of electronic systems such as switchgear
tube based device is 40 ºC. This limits the operation of the
device to where the ambient temperature is less than 40ºC. If References
used at ambient temperatures of greater than 40ºC, the device [1] [1] Lythall F., Lythall P., Reginald t. ‘J&P Switchgear Book’ (Sept
would not be able to switch back to being a passive device 1972 ISBN-13:978-0408000697)
once an open circuit heat cycle had commenced. [2] Seely H.T.‘Curret Transformers on Open Circuit’ (IEE Electrical
Review April 1969)
[3] Kaufmann, R.H., Camilli, G. ‘Overvoltage Protection of Current-
Finally, it has also been observed that significant amounts of Transformer Secondary Windings and Associated Circuits’ American
RF electromagnetic noise are generated whilst the gas Institute of Electrical Engineers, Transactions of, Volume:62 , Issue: 7,
discharge tube device is acting to clamp the open circuit Pages 467-472 July 1943
voltage across the CT. This may influence the response of [4] Costello, D. ‘Open-circuited CT misoperation and investigation’
Protective Relay Engineers, 2014 67th Annual Conference ‘April 3 2014
electronic systems, such as relays, etc., within substations. Pages 383 – 392
However, further work is required to validate this statement. [5] ‘Current transformers with secondary current rating lower than 5 A’
Power Delivery, IEEE Transactions on (Volume:3 , Issue: 2 Pages 503 –
508 Apr 1988
[6] Levine J. ‘Are you using the right CT’s and PT’s for your Appliction’
GE Internal Press
[7] Bridger B. ‘Open Circuit Protectors for Current Transformers;’ Powell
Internal Press Fall 2009
[8] ‘Suretech CT Protector’ Camilli, G. ; Bewley, L.V. Suretech Internal
Press 15 April 2004
[9]http://static.mimaterials.com/metrosil/documents/English/Metrosil_CTP
Us.pdf
[10]http://www.leutron.de/uploads/tx_leutronpdb/550411_TF2000Tr-Th-
Pk_en_v02.pdf

6
IET Review Copy Only