UNSW - SCHOOL OF ELECTRICAL ENGINEERING AND TELECOMMUNICATIONS

ELEC4611 POWER SYSTEM EQUIPMENT LABORATORY

EXPERIMENT 1
TRANSIENTS IN POWER EQUIPMENT CIRCUITS

Transients in power equipment circuits occur when sudden changes, such as switching,
take place. The nature of these transients is strongly influenced by the inability of
inductor currents and capacitor voltages to adjust instantly to changed circuit conditions
because of stored energy. Mutual coupling effects leading to induced currents, such as
the transient and sub-transient reactance effects in rotating machines also influence
transient behaviour. The non-linear B-H curve of ferromagnetic materials can also
influence transients, the most common example being the magnetizing inrush current
commonly encountered in transformers.

This experiment investigates the phenomena of transient recovery voltage and the
magnetizing inrush current.

PART 1: TRANSIENT RECOVERY VOLTAGE

In the event of a fault in a system protected by a circuit interrupter (a fuse or circuit

breaker), the overall behaviour can be obtained from an equivalent circuit such as:

Cir cu it Br ea k er or Fu s e

L R A B

Fig.1

In general L >> R and thus the power factor will be low in the faulted state. The
capacitance C represents all the distributed capacitance on the source side of the breaker.
The fault is taken as a terminal short circuit, with terminal B tied to earth. When the
interrupter operates and clears at a natural current zero the voltage across C is non-zero
because of arcing in the interrupter and the stored energy (CV2/2) in C must then
discharge through L and R, resulting in an oscillatory voltage appearing on terminal A of
the interrupter, superimposed on the power system 50 Hz voltage. The frequency of this
oscillation depends on L, C and R and the amplitude (including the power frequency
component) may approach twice normal value for very low power factors, shortly after
interruption. This voltage oscillation across the interrupter is the "Transient Recovery
Voltage" (TRV) and is an important consideration in choosing circuit interrupters in
power systems.

ELEC4611 - Experiment 1: Transients in power equipment circuits - Page 1 - 2024

Both the amplitude and the initial rate of rise of the recovery voltage are important - the
former because it may cause insulation breakdown in the system and the latter because it
may be too fast for the circuit breaker to withstand and then there may be restrike in the
breaker and a failed interruption.

The amplitude is constrained to a maximum of two times peak system voltage if the
current is pure sinusoidal and if interruption occurs at a natural current zero (see
Appendix). If, however, the current is not sinusoidal and interruption is premature, then
stored inductive energy can produce much higher voltages at the interruption and this
must be designed for. In circuit breakers, such an instance is 'current chopping' which
may occur during the last portion of the current cycle. This is avoided by proper choice
of breaker. However, an HRC (High Rupturing Capacity) fuse will almost always
produce an artificial current zero (and behaviour akin to current chopping) and thus high
recovery voltages are always present where HRC fuses blow and this must be taken
account of in design of systems.

In this experiment, you will compare the recovery voltages of a circuit breaker and a
number of HRC fuses to compare the amplitudes as a function of the rate of decrease of
current down to zero.

Also, it will be possible to determine approximate comparisons of the value of I2t for the
various cases, where the I2t is the specific energy 'let-through' of the interrupter when
opening in response to a short circuit current.

Experiment

The general circuit used is shown in Fig.2:

To CRO
A B (Volt a ge Tr a ce)

L Fu s e or
Br ea k er
110V C

C.T.
E

Poin t -on -Wa ve To CRO

S wit ch (Cu r r en t Tr a ce)

Fig.2

The circuit-breaker used is a 240 V, 5 A breaker. The fuses are HRC type, with ratings 2
A, 3 A, 5 A, 7 A and 10 A. C is 4 F. L consists of 2 inductors in parallel. For each
inductor, the inductance of the full winding is 22 mH but only half the winding is used,
i.e. the total value of L is 5.5 mH. Use any setting of the Point-on-Wave (P.O.W.) switch
but do not vary it once adjusted.

ELEC4611 - Experiment 1: Transients in power equipment circuits - Page 2 - 2024

CRO setting for Circuit Breaker and HRC Fuses at different ratings. Make sure you
double check the settings every time before you press the P.O.W switch.

CH1 CH2 Time

Circuit Breaker 100 V/div 100 mV/div 5 ms/div
2 A Fuse 500 V/div 100 mV/div 5 ms/div
3 A Fuse 500 V/div 100 mV/div 5 ms/div
5 A Fuse 200 V/div 100 mV/div 20 ms/div
7 A Fuse 200 V/div 100 mV/div 50 ms/div
10 A Fuse 200 V/div 100 mV/div 200 ms/div

A. Circuit Breaker

Set the digital oscilloscope according to the settings shown on the circuit box. Set up
circuit to simulate a terminal short circuit by shorting B to E. Use your phone camera to
capture oscilloscope screen displays for report writing.

1. Use the oscilloscope (after obtaining guidance on its use) to record the breaker voltage
in Ch1 (including recovery voltage) and the breaker current in Ch2.

2. From the frequency of oscillation, determine the effective inductance L using C=4 F.

3. Measure the peak value of the TRV relative to the power frequency value and
compare with the theoretical value (i.e. 2Vm).

4. Determine the I 2t value of the breaker operation. A simple triangular approximation

or similar is adequate to obtain I 2t =  i 2 dt by area integration of i 2 .
B. HRC Fuses

Set oscilloscope follows settings mentioned on circuit box and repeat the experiment
using HRC fuses of the various ratings (2 A, 3 A, 5 A, 7 A and 10 A). After blowing the
fuse, throw it into the waste bin.

1. Use the oscilloscope to record the fuse voltage (including recovery voltage) and fuse
current. Obtain also the prospective current waveform.

2. Determine the frequency and hence L again; use the formula f = 1 2 LC . ( )

3. Take particular note of the current just before interruption - estimate the dI/dt and
compare it to the value observed with the breaker.

4. Measure the peak value of the TRV relative to the power frequency value for each
fuse and compare with the theoretical value.
In such a situation when interruption is not at a natural current zero, inductive energy
must also be dissipated after interruption. In this case, the peak value of recovery
voltage is not 2Vm but is:
L
V p = Vm2 + i02
C
where Vm is the peak ac power frequency voltage, io is the cut-off current, L is the
current limiting inductance and C is the leakage capacitance (4 F in the experiment).

5. Determine the I 2t value for each fuse. Compare them with that for the breaker.
ELEC4611 - Experiment 1: Transients in power equipment circuits - Page 3 - 2024
 Trigger setting

Time div Time setting

Trigger level setting

Cursor setting
Channel setting

Voltage div Measurement

Fig.3. Current and voltage waveforms of the 5 A fuse

Data table for Part 1

Vp Vm f L I 2t dI/dt
Circuit breaker not required
2 A Fuse
3 A Fuse
5 A Fuse
7 A Fuse
10 A Fuse

Discuss the following:

1. Explain the difference in the TRV of the fuse and circuit breaker.

2. Comment on the respective values of I 2t . How does this affect the comparative
ability of fuses and circuit breakers to limit thermal effects of high fault currents?

3. Comment on any limitations of the measurement techniques (i.e. the transformers

used for V and I measurement)?

4. What would be the effect of a situation where there was some inductance and
capacitance between terminal B and the fault as shown in Fig. 4 ?

Circ uit Breaker

L1 A B L2

C C Fault
1 2

Fig.3

5. Comment on the comparative values of Vp for the various ratings of fuses [hint:
consider di/dt].

ELEC4611 - Experiment 1: Transients in power equipment circuits - Page 4 - 2024

PART 2: MAGNETIZING INRUSH CURRENT

Unloaded transformers draw only magnetising current in the steady state. This current
establishes the flux in the core and supplies the required power to meet hysteresis and
eddy current losses. However, the magnetic flux must build up from the residual
magnetism state when the transformer is turned on - it is not possible to change the flux
distribution instantaneously. Residual magnetism may cause an offset of magnetic flux
density in the core which, in the worst case, may produce two times the normal flux
density one half cycle after switch-on. This will drive the core into saturation, requiring
a very large magnetising current to establish the high flux level and the large
magnetising current will persist until the flux returns in its normal symmetrical steady
state cycle.

In large (> 10 MVA) transformers the initial inrush current of an unloaded unit may
reach 20-40 times the rated load current of the transformer and may take a second or
more to return to the normal magnetising current. Even the humble laboratory variac
suffers and for a typical 2 kVA variac, with magnetising current of about 0.5 A and
rated current on load of 8 A, the initial inrush current may reach 60 A (possibly blowing
the fuse). Thus, in addition to the problem of high electrodynamic forces on the
windings, inrush currents can cause unnecessary operation of protection and steps may
be required to stop spurious tripping.

Fig.5.

1. Connect the circuit up as in Fig. 5. The inductance used is a steel-core inductance

without air gaps, so that a flux offset will cause saturation.

2. Set the variac to full voltage.

3. Obtain the measurement results from oscilloscope (current waveform), and

determine the steady state magnetising current, the peak inrush and the decay time
constant.

4. It is possible to see the effect of remnant flux orientation on the subsequent

magnetic inrush if you adjust the sweep rate so that the last cycle of magnetising
current can be seen.

ELEC4611 - Experiment 1: Transients in power equipment circuits - Page 5 - 2024

 Trigger setting

Time div Time setting

Trigger level setting

Cursor setting
Channel setting

Voltage div Measurement

Fig.6. Oscilloscope setting for Part 2

REFERENCES

1. A. Greenwood, Electrical Transients in Power Systems, 2nd ed., Wiley, 1991.

2. R.T. Lythall (ed.), The J&P Switchgear Book: an Outline of Modern Switchgear
Practice for the Non-specialist Use, 7th ed., Newnes-Butterworths, 1972.
3. C.H. Flurscheim (ed.), Power Circuit Breaker Theory and Design, Peregrinus, 1982.
4. E. Jacks, High Rupturing Capacity Fuses: Design and Application for Safety in
Electrical Systems, Wiley, 1975.
5. F. Kussy and J. Warren, Design Fundamentals for Low-voltage Distribution and
Control, M. Dekker, 1987.
6. Bharat Heavy Electricals Limited, Handbook of Switchgears, McGraw-Hill, 2007.

APPENDICES

Appendix 1: Transient Recovery Voltage

Fig.7

di
L dt + Vc = Vm cos(t) after interruption
A E A

dVc
I = C dt A E

d2Vc Vc Vm
Thus: dt + LC = LC cos(t)
A E A A E A A E A

ELEC4611 - Experiment 1: Transients in power equipment circuits - Page 6 - 2024

Using o2 = 1/LC and transforming:
s
s2 vc(s) - sVc(0) - Vc'(0) + o2 vc(s) = o2 Vm 2 2
s +
A E

whence:
s s 1
vc(s) = o2 Vm 2 2 2 2 + Vc(0) 2 2 + Vc'(0) 2
(s + )(s +o ) s +o s +o2
A E A A E A A E

(Because the fault is a short circuit, the TRV across the breaker is identical to the
capacitor voltage here).

If we neglect the (low) arc voltage at t=0, then Vc(0) = 0 and also, because I(0)=0:
Vc'(0) = I(0) / C = 0
Thus we have:
s
vc(s) = o2 Vm 2 2 2
(s + )(s +o2)
A E

o2 s s
= Vm 2 2 { 2 2 - 2 }
(o - ) s + s +o2
A E A A E A A E A

Hence:
o2
vc(t) = V [cost - cosot]
o2-2 m
A E A

In general, o>>. Therefore:

vc(t) = Vm [cost - cosot]

As the TRV portion of interest is only the first cycle, which is of short duration in terms
of a 50 Hz cycle, we can write:
vc(t) = Vm (1 - cosot)

Thus, Vc(peak) = 2 Vm and the average initial rate of rise of TRV is:
2Vmo
(RRRV)av = volts/sec

A E A

Fig.8

ELEC4611 - Experiment 1: Transients in power equipment circuits - Page 7 - 2024

Appendix 2: Magnetizing Inrush Current

S t=0 L

v=Vsin( t + ) R

Fig.9

When S is closed in Fig.10, the inductor constrains the current to start from zero. If S
closes at a supply voltage angle that corresponds to a non-zero steady-state AC current
then a decaying transient component of current is required to make i = 0 at t = 0. The
onset of a fault in a power system is similar to the behaviour of the above circuit from t
= 0, because the power system fault current is usually much larger than any load current
that may be flowing prior to the fault.

If  = V sin (t + ) then the resulting current is:

V
i(t) = Z (sin (wt +  - ) - e-tR/L sin ( - ))
A E

L
where Z = R2 + L ,  = tan-1 R
A E A
A

and  is the closing angle: i.e.  = Vsin at t = 0

Thus, the current contains a decaying transient term called the d.c. offset, the amplitude
of which depends on (-) the difference between the closing and impedance angles.
Note that for the total current:

(i) i = 0 at t = 0 always
V
(ii) i = Z sin t when  =  i.e. under these conditions there is no d.c. offset.
A E A

V
(iii) i = Z (sin (t ± 900) ± e-Rt/L) when  =  ±  i.e. maximum d.c. offset.
A E E

In case (iii) above, the first asymmetric peak may be almost twice the symmetric
maximum current if L is high and R low (as is usual).

We now want to consider the effect of saturation in the magnetic core on such
transients. For simplicity, we shall consider only the most unfavourable instant of
switching. For pure inductance this will be closing an unloaded coil at t = 0 on v = V
sin t. Neglecting any resistance:
d
N dt = V sin t.
A E A

V
  = - cos t + A where  = = flux amplitude
N
A E

Now () =  at t = 0,
  =  and  =  (1 - cost)

[Remanence can be included by specifying () = ±  then  = ±  +  ( − cost)]

ELEC4611 - Experiment 1: Transients in power equipment circuits - Page 8 - 2024

Thus, the offset flux oscillates between 0 and 2, as shown in Fig.11. Usually the
circuit will be designed so that flux 0 will imply magnetic flux density close to the
knee-point of the magnetization curve, requiring a magnetising current i1 (the normal
value), as shown. When the flux rises to 20, the magnetising current will become very
large (i2) - the magnetizing inrush phenomenon.

There will be a decay in the offset flux because of circuit resistance, thus more
accurately:
 =  (e-tR/L - cost)

This is plotted in Fig.12 and it can be seen that the flux does not quite rise to 2

Apart from the angle of closing, the magnitude of inrush current is also affected by
remanent flux in the core prior to t = 0. This will change the initial condition in the flux
equation and thus the magnitude of the transient, depending on whether the remanence
is in phase or anti-phase with the new flux cycle.

Fig.10: Magnetising current for normal and offset flux

ELEC4611 - Experiment 1: Transients in power equipment circuits - Page 9 - 2024

 Fig.11: Practical magnetising inrush waveforms

Latest update: 11 Feb 2024 by Mr. Hua Chai

ELEC4611 - Experiment 1: Transients in power equipment circuits - Page 10 - 2024