The Effect of Inrush Current on Transformer
Protection
Li-Cheng Wu,,Student Member,IEEE, Chih-Wen Liu,Senior Member,IEEE,
Shih-En Chien,Student Member,IEEE,Ching-Shan Chen,Member,IEEE
Department of Electrical Engineering, National Taiwan University, Taipei, Taiwan
Abstract—Transformers
are key components for electrical
energy transfer in power system. Stability and security of
transformer protection are important to system operation; we
found that many mal-trip cases of transformer protection are
caused by inrush current problems. The phenomenon of
transformer inrush current has been discussed in many papers
since 1958 [1-5]. Therefore, this paper will only discuss and
analyze inrush current problems. Finally, this paper will also
present two cases that were analyzed with the use of digital
simulation technique to make COMTRADE files, to provide
over-current protection and differential protection tests and the
analysis of the effect of inrush current on transformer protection.
Index Terms—Over-current protection, Differential Protection,
Inrush Current, COMTRADE files.
I. INTRODUCTION
In power systems, differential protection is applied for
transformer capacity above 10MVA, while over-current
protection is used for transformer with the banks below
10MVA for main protection that includes simple theory and
best protection results. However, the transformer will create
large inrush currents when the transformer operates on no-load
energizing condition. This inrush currents involves a large and
long lasting dc component, which is rich in harmonics, assumes
large peak values at the beginning about 6 to 30 times of the
rated value. This condition causes unbalance of current loop of
differential relay that will occur with mal-trip. In order to
prevent false tripping due to an inrush current, a technique
using the content of the second harmonic component in the
current waveform is commonly used. However, this method
cannot provide total solution for inrush current. Therefore, we
present digital simulation method to analyze and to test to know
the best transformer protection schemes.
II. SIMULATION AND ANALYSIS OF INRUSH CURRENT
The equivalent circuit of transformer model shown as Fig.1
consists of an ideal transformer of ratio N1 : N 2 and parameter
of elements. The model takes into account the winding
resistances ( R p , Rs ), the leakage inductances ( L p , Ls ) and
the excitation characteristics of iron core. The excitation
characteristics of iron core can be expressed by equation (1) [8]
which can show the excitation curve as Fig.2. Due to the
non-linear of transformer iron core, this will result in excitation
and saturation problems of transformer in power systems.
According to different operation point of transformer core as
Fig.3, we can get different excitation current on transformer.
When switched to a no-load transformer, this will result in
transformer’s working in saturation area of excitation curve
(see Fig.3) in which creates high magnitude asymmetrical
current with a high harmonic and a high direct current
components. This may cause mal-operation of over-current
protection or differential protection. Typically, for steady state
operation, the excitation current of transformer is slightly less
than 5% of the rated current (see Fig.3). In practice, the
magnitude and duration of transient inrush current depend on
the following [9]:
z
Circuit breaker switching angle when the transformer is
energized
z
The value and sign of the residual flux linkage in the
transformer core
z
The saturation characteristic of the transformer core
z
Source impedance
Fig.1 Equivalent circuit of a two-winding transformer
ϕ = −s *[Isat *tan−1 (−s *(dϕ dI )* Ie + Ic ) − s *ϕr * Ie + ϕsat ] (1)
Where： s =1 for an ascending trajectory, s =-1 for a
descending trajectory
I sat for saturation current of transformer
dϕ
for slop of excitation curve of transformer
dI
I e for excitation current
I c for coercive current
ϕr for residual flux
ϕ sat for saturation flux of transformer
values of inrush current vs. residual flux (from -1 to 1 pu) when
CB1’s closed angle is 0 degree and 90 degrees respectively.
The simulation results show that the inrush current can be
reduced by controlling CB1’s closing time and residual flux.
For example, according to Fig. 1, we can write equations as
follows:
1.5
ex c itatio n flux (p u)
1
0.5
Vp = Rp I p + Lp
+ N1
dϕ m
dt
(2)
dt
dI
dϕ
Vs = Rs I s + Ls s + N 2 m
dt
dt
0
-0.5
(3)
When the transformer is energized in no-load, the equation
(3) can be expressed by:
-1
-1.5
-0.02
dI p
-0.015
-0.01
-0.005
0
0.005
excitation current (pu)
0.01
0.015
0.02
Fig.2 Excitation curve of transformer
ϕm =
1
Vs dt
N2 ∫
(4)
Here, we want to get CB1 optimal closed time in the main
flux ( ϕ m ) close to zero. Because the number of turns of
∫
primary N1 is larger than the item of R p I p dt and L p I p , we
can modify the equation (4) as follows:
ϕm =
1 ⎡
⎤ ≅ 1 V p dt
−
−
V
R
I
dt
L
I
(
)
p
p
p
p
p
∫
⎦ N ∫
N1 ⎣
1
(5)
Therefore, the value of main flux ( ϕ m ) can be calculated at
any instant using equation (5). In order to reduce inrush
currents, we can use the main flux information combined with
zero-crossing detector to determine the CB1 closing time,
which is at main flux ( ϕ m ) zero. The simulation result is shown
in Fig. 10. The all of the inrush currents are reduced by
controlling CB1’s closing time.
Table 1 The parameters of the simulation system
Fig.3 The operation point of excitation curve determines the magnitude
of the excitation and inrush current
Fig.4 shows the simplified single-line diagram of
transformer protection scheme used in TPC’s (Taiwan Power
Company) substation. We will simulate a transformer running
in no-load situation. Table 1 shows the parameters of
simulation system. Typically, the simulation results of the
inrush current are shown in Fig.5. At the same time, the field
test result is presented in Fig.6. Those currents of all figures are
used to CT (Current Transformer) secondary values whose
ratio is 1200:5. The inrush current is about 5 times larger than
the rated current of the transformer. The large inrush current
will hit transformer protection, which causes mal-trip for
different protection or over-current protection. We will discuss
these issues in the next section. Here, inrush currents are
formed based on the following three major factors: the
transformer energized angle, residual flux of iron core and
structure. Fig. 7 shows the relations between peak values of
three-phase inrush currents and CB1 closed angles when the
residual flux of transformer is zero. Fig. 8 and Fig. 9 show Peak
Fig. 6 Inrush current field test result
CB Switching degrees vs. peak values of Inrush current when residual flux is zero.
10
9
8
7
A
6
5
4
3
Fig.4 Simplified single-line diagram of transformer protection
2
Ia
Ib
Ic
1
0
High Voltage Side
15
Phase A
Phase B
Phase C
0
50
100
150
200
degrees
250
300
350
Fig. 7 Peak values of inrush current vs. CB1 closed angles
10
Residual flux vs. peak values of Inrush current
20
5
Ia
Ib
Ic
18
14
12
A
A
16
0
-5
10
8
6
-10
4
2
-15
0
1
2
3
4
5
Cycles
6
7
8
Fig. 5 Inrush current simulation result
9
10
0
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
Residual (pu)
0.4
0.6
0.8
1
Fig. 8 Peak values of inrush current vs. residual flux(from -1 to 1 pu) when
CB1’s closed angle is zero degree
transformer neutral line is caused by transformer when it is
energized in no-load.
Residual flux vs. peak values of Inrush current
20
Ia
Ib
Ic
18
16
14
A
12
10
8
6
4
2
0
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
Residual (pu)
0.4
0.6
0.8
1
Fig. 9 Peak values of inrush current vs. residual flux(from -1 to 1 pu) when
CB1’s closed angle is 90 degrees
A
10
0
CB1 uncontrolled Phase A
CB1 controlled Phase A
-10
0
2
3
4
5
Cycles
6
7
8
9
10
5
Cycles
6
7
8
9
10
5
Cycles
6
7
8
9
10
CB1 uncontrolled Phase B
CB1 controlled Phase B
10
A
1
0
-10
0
2
3
4
CB1 uncontrolled Phase C
CB1 controlled Phase C
10
A
1
0
Case I: The connection diagram of differential relay of
transformer is as shown in Fig. 4. The basic theory of
differential relay is formed with the use of current balancing of
transformer of the two-sides as a trip signal. When the
unbalance currents of transformer of the two-sides is larger
than the pick-up setting of differential relay, the trip signal will
be sent from differential relay to trip circuit breaker (CB) for
isolated fault. From the basic differential theory, we know that
the inrush currents only go through one of the transformer
winding when the transformer is energized in no-load. At this
time, if the differential relay of transformer doesn’t include the
best blocking function of the inrush currents, than the mal-trip
will occur. The differential relay of transformer use either
harmonic or blocking principles for inrush currents of
transformer. Generally speaking, the second-harmonics content
of inrush currents of transformer usually is 15% larger than the
fundamental component. Therefore, the setting of the
second-harmonics of differential relay is usually set at 15%.
In the Fig.4, the frequent mal-trip of differential relay is
caused by the inrush currents when the transformer is energized
in no-load. Here, the differential relay is electrical-mechanical
type that utilizes the second-harmonic blocking for inrush
current restrain. The setting of the second-harmonic is at 15%.
Fig. 11 shows the situation when the traces of inrush current is
entering the protection zone of differential relay during the
transformer is energized in no-load. Fig. 12 is the
second-harmonic contents of the inrush currents after it is
transferred by Fourier transfer. We can see the
second-harmonic contents of the inrush currents of phase A, B
and C at 10~12%, 20~23%, and 19~35% respectively. It is very
obvious that the second-harmonic contents of phase A is less
than 15%, so the mal-trip of differential relay of phase A
usually occurs in transformer when it is energized. In order to
improve this situation, we adjusted the setting of the
second-harmonic from 15% to 10% and since then the
differential relay has never mal-tripped during the time when
the transformer is energized in no-load.
-10
2
10
0
1
2
3
4
Characteristic curve of differential relay
IA
IB
IC
Fig. 10 Inrush currents compare CB1 uncontrolled with CB1
controlled
1
Idiff (A)
10
III. PRACTICAL EXAMPLES AND RESULTS OF THE EFFECT OF
INRUSH CURRENT
0
10
In this section, we will use two cases to explain and
analyze how and why the protective relay mal-trip occurs in
inrush currents. Case I for the mal-trip of differential protection
relay occurs during the time when transformer is energized in
no-load. Case II for the mal-trip of over-current relay of
0
1
10
10
Ires (A)
Fig. 11 the traces of inrush currents
2
10
Waveform of TR-SW1 neutral-line
15
30
IA
IB
IC
A
0
0
-10
-5
-20
-10
0
1
2
3
cycles
4
5
-15
6
0
5
10
15
cycles
RMS values of Waveform of TR-SW1 neutral-line
10
40
IA
IB
IC
30
Field test
Simulation result
8
6
A
20
4
10
0
2
0
1
2
3
cycles
4
5
0
6
Fig. 12 Second harmonic contents of inrush currents
Fig. 13 shows one-line diagram of generator ‘s cooling
pump systems at a power plant in Taiwan. Here, the cooling
system is very important for generators. If the cooling system
shut down, then all the generators will be tripped by
over-heating. The loss power of the load of transformer
(TR-SW2) results from the mal-operation of TR2_CB1 by
operators and the factor that the TR2_CB2 remains closed. At
the same time, the transformer (TR-SW2) only produces inrush
current when the Tie CB is closed. The inrush currents results
in the mal-trip of over-current relays (50/51Z, the relay setting
is 4A for instant trip) of two transformers’ (TR-SW1 and
TR-SW2) neutral-line. The current waveforms are shown in
Fig. 14 and Fig.15. In order to solve the mal-trip of 50/51Z, we
designed an inter-lock logic of the Tie CB as Fig. 16 for
security and reliability of a power plant. This logic can keep Tie
CB working in normal condition. In general condition, when
we close the Tie CB, the currents waveforms of the
transformers’ (TR-SW1 and TR-SW2) neutral line are shown
in Fig.17 and Fig 18. The results are satisfying because the
magnitude currents are less than the setting (4A) of the relay.
Therefore, the systems can work in the best condition.
0
5
10
15
cycles
Fig. 14 Current of neutral line of TR-SW1 when TR2_CB1 is opened
Waveform of TR-SW2 neutral-line
15
Field test
Simulation result
10
5
A
2sd Harmonic content (% of fundamental
5
10
-30
Field test
Simulation result
10
0
-5
-10
-15
0
5
10
15
cycles
RMS values of Waveform of TR-SW2 neutral-line
10
Field test
Simulation result
8
6
A
Inrush current (A)
20
4
2
0
0
5
10
15
cycles
Fig. 15 Current of neutral line of TR-SW2 when TR2_CB1 is opened
Fig. 16 Improvement logic of case study of inrush currents
Fig. 13 Single-line diagram of case II
Waveform of TR-SW1 neutral
1.5
TR-SW1-50N
1
A
0.5
0
-0.5
-1
0
0.1
0.2
0.3
0.4
0.5
0.6
times (sec)
0.7
0.8
0.9
1
RMS values of waveform of TR-SW1 neutral
0.5
TR-SW1-50N
0.4
A
0.3
0.2
0.1
0
0
0.1
0.2
0.3
0.4
0.5
0.6
times (sec)
0.7
0.8
0.9
1
Fig. 17 Current of neutral line of TR-SW1 when TR2_CB1 is closed
Waveform of TR-SW2 neutral
1.5
TR-SW2-50N
1
A
0.5
0
-0.5
-1
0
0.1
0.2
0.3
0.4
0.5
0.6
times (sec)
0.7
0.8
0.9
transient data exchange (COMTRADE, IEEE C37.111) for a
standard format for the exchange of data in 1991. Here, we will
use digital simulation to produce COMTRADE file for
transient tests of differential protection.
Fig. 19 shows a transient test structure of differential relay
for transformer protection. We use simulation tools of PSCAD
and MATLAB to simulate a number of different fault types and
to produces COMTRADE files for transient test of differential
relay. The transient tests of open loop are used to input
COMTRADE file to wave amplifier that produce real currents
or voltages to inject into protection devices for performance
evaluations of protective relay. In addition, this method can be
combined by GPS (Global Position System) for end-to-end test.
Fig. 20 shows the inrush current for differential relay test. A
transient test of transformer internal-fault (AG) is shown in Fig.
21. The differential relay is operated because the differential
current trace of phase A falls into the operation area of
differential curve. On the contrary, in Fig. 22, the differential
relay is of no-trip when the external fault occurred in out of the
protection zone of differential relay of transformer. The
transient test can nearly simulate real situation of differential
protection of transformer to clearly show whether the
differential relay should be operated or not in the internal or the
external faults. However, the traditional steady-state test cannot
control dynamic processes, namely, internal faults, external
faults, energized and saturation of transformers. Usually, it is
discussed when the faults are occurred. From the above
discussion, the transient test of protective relay is important for
power systems protection.
1
RMS values of waveform of TR-SW2 neutral
0.5
TR-SW2-50N
0.4
A
0.3
0.2
0.1
0
0
0.1
0.2
0.3
0.4
0.5
0.6
times (sec)
0.7
0.8
0.9
1
Fig. 18 Current of neutral line of TR-SW2 when TR2_CB1 is closed
IV. TRANSIENT TEST ON DIFFERENTIAL PROTECTION
Nowadays, the fault and transient data recording are widely
used in power systems. Their data are being used with various
devices to enhance and automate the analysis, testing,
evaluation, and simulation of power systems and related
protection schemes during fault and disturbance conditions. In
order to get a bridge between different electric devices, IEEE
defines a common language that is a common format for
Fig. 19 the structure of transient test for differential relay
20
A
HV Phase A
LV phase A
0
-20
0
2
4
6
Cycles
8
10
12
4
6
Cycles
8
10
12
4
6
Cycles
8
10
12
4
6
Cycles
8
10
12
20
A
HV Phase B
LV Phase B
0
-20
0
2
20
A
HV Phase C
LV Phase C
0
-20
Fig.20 Transient test of differential protection
0
2
T rip s ignal
1
100
HV Phase A
LV phase A
A
0
0.5
0
-100
0
2
4
6
Cycles
10
8
10
Phase A trip
Phase B trip
Phase C trip
0
2
12
A
HV Phase B
LV Phase B
Characteristic Curve of Differential Relay
0
-10
30
0
2
4
6
Cycles
10
8
10
25
12
20
0
-10
Idiff
A
HV Phase C
LV Phase C
0
2
4
6
Cycles
Trip signal
1
8
10
15
12
Phase A trip
Phase B trip
Phase C trip
0.5
10
5
0
0
2
4
6
Cycles
8
10
12
0
Characteristic Curve of Differential Relay
2
4
6
8
10
Ires
12
14
16
18
20
Phase A
Phase B
Phase C
25
V. CONCLUSIONS
20
15
10
5
0
0
Fig. 22 A transient test of external-fault (ABCG) for transformer
30
Idiff
Phase A
Phase B
Phase C
0
2
4
6
8
10
Ires
12
14
16
18
20
Fig. 21 A transient test of Internal-fault (AG) for transformer
This paper provides detail analysis of simulation and field
measurement for problems of inrush current and transformer
protection. At the same time, we have also identified mal-trip
factors from our case studies in differential relay and
over-current relay, as well as resolutions for improvement. In
addition, in protection test field, we present advanced transient
test method that is applied in the setting of relay, fault analysis
and new algorithms research of relay and so on, which will
result in better performance in transformer protection and thus
salability, reliability and security of power systems operation
can be reached.
REFERENCES
[1] Saied, M.M., “A study on the inrush current phenomena in transformer
substations”, IEEE , Volume: 2 , 30 Sept.-4 Oct. 2001 Pages:1180 - 1187
vol.2
[2] Stringer, N.T., Lawhead, L., Wilkerson, T., Biggs, J., Rockefeller, G.D.，
Testing and performance of transformer differential relays ，
IEEE , Volume: 3 , Issue: 4 , July-Aug. 1997 Pages:36 – 42。
[3] Stringer, N.T., Lawhead, L,; Wilkerson, T.; Biggs, J., Rockefeller, G.D.，
Real-time
transient testing and performance of transformer differential relays，
IEEE , Volume: 2 , 8-12 Oct. 1995 Pages:1142 - 1150 vol.2。
[4] Bronzeado, H., Yacamini, R.,”Phenomenon of sympathetic interaction
between transformers caused by inrush transients”, IEE , Volume:
142 , Issue: 4 , July 1995
Pages:323 – 329
[5] Yabe, K., “Power differential method for discrimination between fault and
magnetizing inrush current in transformers”, IEEE Transactions
on , Volume: 12 , Issue: 3 , July 1997 Pages:1109 – 1118
[6] IEEE Standard “Common Format for Transient Data Exchange
(COMTRADE) for Power Systems” , IEEE Std C37.111-1999 ,
[7] IEEE WG116 Report “Understanding Microprocessor-Based Technology
Applied to Relaying”, February 2004
[8] Casoria, S., P. Brunelle, and G. Sybille, "Hysteresis Modeling in the
MATLAB/Power System Blockset," Electrimacs 2002, École de
technologie supérieure, Montreal, 2002.
[9] K.S. Smith , C.Eng., MIEE ,”Transformer Inrush Studies for Wind Farm
Grid Connections”, International Conference on Power Systems
Transients (IPST’05 June 2005)
[10] Guzman, A.; Zocholl, S.; Benmouyal, G.; Altuve, H.J.,”a current-based
solution for transformer differential protection-part I: Problem Statement",
IEEE Transactions on Power Delivery ,Volume 16,October 2001
Page(s):485-491
[11] Guzman, A.; Zocholl, S.; Benmouyal, G.; Altuve, H.J.,”a current-based
solution for transformer differential protection-part II: relay description
and evaluation", IEEE Transactions on Power Delivery ,Volume
17,October 2002 Page(s):886-893
Li-Cheng Wu received his B.S and M.S. degrees in electrical
engineering from National Taiwan University of Science and
Technology, Taipei, Taiwan, in 1997 and 1999, respectively.
He is currently pursuing the Ph.D. degree at National Taiwan
University, Taipei, Taiwan. Between 1997 to 2002 he worked
as an electrical engineer in the Department of relay, Taiwan
Power Company.
His main research interests are power electronics, high voltage
test and power system protection.
Chih-Wen Liu (S’93-M’96-SM’02) was born in Taiwan, in
1964. He received the B.S. degree in electrical engineering
from National Taiwan University (NTU), Taipei, Taiwan, and
the M.S. and Ph.D. degrees in electrical engineering from
Cornell University, Ithaca, NY, in 1987, 1992, and 1994,
respectively.
Since 1994, he has been with NTU, where he is a Professor of
electrical engineering. His main research interests include
application of computer technology to power system
monitoring, protection, and control. His other research interests
include motor control and power electronics.
Shih-En Chien was born in Keelung, Taiwan in 1980. He
received his B.S. degree in electrical engineering from National
Taiwan University of Science and Technology in 2002, and
M.S. degree in electrical engineering from National Taiwan
University in 2004. He is currently working toward his Ph.D.
degree at Electrical Engineering Department of National
Taiwan University. At present, his interested research includes
power system protection and relay testing.
Ching-Shan Chen was born in Taichung, Taiwan, in 1976. He
received the B.S. degree in electrical engineering from National
Taiwan University of Science and Technology, Taipei, Taiwan,
and the M.S. and Ph.D. degrees in electrical engineering from
National Taiwan University, Taipei, Taiwan, in 1998, 2000,
and 2003, respectively.
At present, he works at Industrial Technology Research
Institute and his research interests include distributed
generation systems and computer relaying.