Calculation of Transformer Inrush Currents

occurring during the Energizing of the Public

Grid after a major Black Out
C.P.J. Jansen; J.G. Slootweg and R.A.C.T. de Groot

strongly non-sinusoidal and can last for several minutes.

Abstract—The paper describes simulations and a live test of There is a risk that protection devices respond to these
the black-start procedure of an industrial power plant in the currents, causing the generating unit to trip. In addition,
North of the Netherlands. After a total blackout of the Public resonance overvoltages can occur caused by the transformer
Grid in the Netherlands, the restoration process shall start with
non-linear impedance and the line or cable capacitance
so-called black-start units. To increase the probability that the
re-energizing of the Public Grid indeed succeeds, the Dutch together with the weak damping of the system [2], [3].
transmission system operator obliges units to prove their black- One way to energize the transformers is by performing a
start capability by performing a live test. so-called soft start of the complete circuit generator-step-up
A crucial step in the re-energizing of the Public Grid is the transformer-grid transformer. For a soft start, first the
energizing of large transformers. Problems can be expected due connections between generator and transformers are made.
to the inrush current of the transformers. For one such Then the generator is started without excitation voltage. If the
generating unit extended simulations of the inrush phenomena of
the transformers and calculation of inrush currents and
generator runs at no load and full speed, the excitation current
overvoltages were performed. Based on the simulation results of the generator is increased. The gradual increase of the
well-founded decisions were made on the most suitable start-up applied voltage avoids inrush currents. Drawback is that the
procedure. The simulations were followed by live tests of the procedure deviates from the normal start-up procedure of the
black-start procedure. unit and also affects the protection scheme and possible
interlocks.
Index Terms—Inrush Current; Grid Black Out; Grid To increase the success rate of the re-energizing of the
Restoration; Transformer Energizing.
Public Grid after a total black out, the Dutch transmission
I. INTRODUCTION system operator (TenneT) obliges contracted generating units
to prove their black-start capability by performing a live test.
A FTER a total blackout of the Public Grid, and separation
of all connected power plants, the restoration process
For one such generating unit, an industrial power plant in
the North of the Netherlands, extended simulations of the
begins by starting the generating units designated (or inrush phenomena of the transformers and calculation of
contracted) as black-start units. Black-start units are capable inrush currents and overvoltages were performed. Based on
of starting without external power supply. A black-start is the simulation results, well-founded decisions were made on
usually performed with the aid of a diesel unit or a gas turbine the most suitable start-up procedure. The simulations were
unit to supply the auxiliary systems of the main unit. Once followed by live tests of the black-start procedure.
running, the generating unit is connected to the Public Grid
[1]. II. ORIGIN OF INRUSH CURRENTS
A crucial step in the re-energizing of the Public Grid is the
The working principle of a transformer is governed by
energizing of the main unit's step-up transformer or the first
Faraday’s and Ampère’s laws [4-6]. Faraday’s law states that
grid (step-down) transformers. These grid transformers are of
induced voltage is equal to the derivative of flux linkage, or,
similar or even higher rating than the generating unit.
stated differently, that the coupled flux equals the integral of
Problems can be expected due to the inrush current of the
the applied voltage. Ampère’s law relates the flux linked by a
transformers. These currents are large in amplitude, are
coil to the current flowing through it.
During normal operation, the transformer current and
This work is done by KEMA T&D Consulting and Essent Netwerk B.V. voltage, as well as the flux in the iron core, vary sinusoidally.
and was supported by Delesto B.V. and by TenneT bv, the Dutch When the transformer is not loaded, i.e. one of the terminals is
Transmission System Operator.
open, the phase shift between voltage on the one hand and
C. P. J. Jansen and R. A. C. T. de Groot are with KEMA T&D Consulting
Arnhem, The Netherlands (e-mail: [email protected]). flux and current on the other, is approximately 90 degrees.
J. G. Slootweg is with Essent Netwerk B.V. ’s-Hertogenbosch, The Assume now that the transformer is idle and a voltage is
Netherlands (e-mail: [email protected] ).
applied. When the presence of remanent magnetism is TABLE I
GENERATOR PARAMETERS
neglected and the connection happens at the instant at which
the flux would equal zero during normal operation, i.e. at the Quantity Value Quantity Value
voltage’s maximum, this does not give rise to any exceptional Gen 1 Gen 2 Gen 1 Gen 2
current or flux. The flux is the integral of the applied voltage Snom 34.08 MVA 50 MVA Xq ’ 0.5 p.u.* 0.5 p.u.*
Unom 10.5 kV 10.5 kV Xq’’ 0.131 p.u.* 0.151 p.u.
and is exactly equal to what it would be if this had not been Xl 0.10 p.u.* 0.10 p.u.* X0 0.075 p.u. 0.085 p.u.
the first period of the applied voltage. Rl 1.481 mp.u. 1.317 mp.u. Td0’ 6,54 s 5.49 s
However, when the transformer is connected at the moment Xd 2.007 p.u. 1.910 p.u. Td0’’ 0,04 s* 0.018 s
that the voltage equals zero, the flux will reach twice its Xd’ 0.194 p.u. 0.226 p.u. Tq0’ 0,50 s* 0.04 s*
Xd’’ 0.131 p.u. 0.151 p.u. Tq0’’ 0,15 s* 0,15 s*
normal height when integrating the voltage. Had the
Xq 1.8 p.u.* 1.72 p.u.*
transformer already been connected, the integration would
have started from the negative maximum of the flux and end TABLE II
at the positive maximum of the flux. However, because this is TRANSFORMER PARAMETERS
the first period of the voltage, the integration starts at zero flux Quantity Value
and the flux reaches twice the maximum value occurring Transformer 1, 2 Transformer 3 Transformer 4
during normal operation. Snom 20 MVA 50 MVA 160 MVA
Because these high flux densities, the iron core is driven UHV 66 kV 70.5 kV 220 kV
ULV 10.6 kV 10.5 kV 66 kV
into saturation. Therefore, the transformer starts to exhibit Uk 14.7 % 11.0% 17.8 %
non-linear behavior. As a result, the current is no longer PCu 87.6 kW 116 kW 459 kW
approximately proportional to the flux. Instead, much more P0, 100% 10.3 kW 31.3 kW 71.4 kW
current is needed to increase the flux and the current can I0, 100% 0.77 mp.u. 2.3 mp.u. 1.04 mp.u.
easily reach much more than two times its maximum. This
large current is referred to as the inrush current. As stated 34.0 MVA 20.0 MVA 20.0 MVA
10.5 kV 10.6/66.0 kV 66.0/10.6 kV
earlier, inrush currents can lead to undesirable effects, such as
the triggering of protection devices and resonant overvoltages.
G
~

The presence of remanent magnetism can increase this effect, Transformer 1 Transformer 2
leading to even larger inrush currents.
10.5 kV 66.0 kV 10.5 kV

III. MODELING System a

A. System description 50 MVA 50 MVA

10.5 kV 10.5/70.5 kV
The systems studied in this paper consist of transformers
G

and generators in island operation, as depicted in figure 1. The

~

160 MVA
generators have a rating of 34 MVA and 50 MVA at a voltage Transformer 3 66.0/220.0 kV
of 10.5 kV. The first generator is connected to a 66 kV busbar
10.5 kV
through a 20.0 MVA 10.6/66 kV transformer. It is driven by a
gas turbine. The second generator is connected to the same Transformer 4
34.0 MVA 20.0 MVA
66 kV busbar through a 50 MVA transformer and is also a gas 10.5 kV 10.6/66.0 kV 220.0 kV
turbine unit.
G
~

To investigate the inrush phenomenon and to validate the

developed models, firstly a second 20.0 MVA 10.6/66 kV Transformer 1
transformer is connected to the 66 kV busbar (system a in
10.5 kV 66.0 kV
figure 1). In the next case also a 160 MVA 66/220 kV
transformer is connected to this 66 kV busbar (system b in System b
figure 1). In the first case, only one generator is in operation, Fig. 1. Investigated systems a and b
whereas in the second case, both are in operation. The relevant
parameters of the generators are given in table I and of the
transformers in table II. B. ATP Model
The parameters not given in the generator documentation have The ATP model was developed using ATPDraw. The
been estimated and are marked with *. The one-line diagrams generators were modeled as synchronous generators using the
of the systems are depicted in figure 1. model SM59 for which all relevant parameters are given in
The generators and transformers were each modeled in two table I. The transformers were modeled using the BCTRAN
different software packages, namely ATP and DIgSILENT dialog box. For those transformers connected directly to the
Power Factory. generators, which were assumed to be already energized at the
start of the simulation, all nonlinearities were neglected. In
case of the transformer to be energized, the non-linear
behavior was modeled using the pseudo-nonlinear hysteretic
inductor component Type 96 [7]. Φ
The Φ-I characteristic of the nonlinear inductor proved saturated

critical for matching the simulation results with the

Φknee
measurements. However, it was unfortunately not available
for the investigated transformers. Therefore, it was calculated
using the supporting routine HYSDAT that comes with ATP.
The values for Φ and I to be inserted in HYSDAT were unsaturated
derived using the “View+” function of the BCTRAN dialog
box. It was assumed that the positive saturation point, to be
inserted in HYSDAT, lied at a voltage of 1.18 p.u., and the
corresponding no-load current was assumed to equal 10 times iM
that at a voltage of 1.1 p.u. The assumption of a factor 10 Fig. 3. Flux saturation model in Power Factory
increase in the current was based on measurements taken from
one of the transformers, of which no-load measurements up to IV. SIMULATION RESULTS AND MEASUREMENTS
a voltage of 1.18 p.u. were available. These were used to
A. Simulation results
extrapolate the value for the other transformers of which only
measurements up to a voltage of 1.1 p.u. were available. Now, simulation results obtained with the models described
earlier are presented. Figure 4 shows simulation results
C. Power Factory Model obtained with ATP and Power Factory. It shows the phase
In Power Factory the generators were modeled using the currents for the switching of the 20.0 MVA transformer
standard synchronous machine model. The parameters for the (transformer 2 of system a in figure 1).
generators are given in table I. The transformer model in 200
[A]
Power Factory used for the inrush simulations, is the two
150
winding transformer model shown in figure 2 [8]. The model
contains the leakage reactances at the low and high voltage 100

side together with the winding resistance at low and high 50

voltage side. The model also shows the magnetizing reactance
0
Xm and the ion losses resistance RFe.
Rprim Xprim Rsec Xsec -50

-100

Uprim Usec -150

XM RFe -200
0.015 0.035 0.055 0.075 0.095 [s] 0.115
n1 : n2 (f ile SW_20_2.pl4; x-v ar t) c:X0055C-X0051C c:X0055A-X0051A c:X0055B-X0051B

Fig. 2. Transformer model in Power Factory

200.0

To model the effect of inrush currents, the magnetizing

reactance must be modeled as a function of the magnetizing
100.0
current. For that, the magnetic flux is given as a function of
the magnetizing current in figure 3.
The unsaturated part of the curve is determined by the 0.0
simulation software from the no-load characteristics of the
transformer. The saturated part and also the knee flux, the
normal operating point of the transformer, is given by the Φ-I -100.0
characteristic of the transformer. As already mentioned, this
characteristic was not known for the transformers. For the
knee flux a typical value 1.04 p.u. is used. For the ratio -200.0
0.00 25.00 50.00 75.00 [ms] 100.
between the gradient of the saturated and unsaturated curve, a 38-2T02: Phase Current A/HV-Side in A
38-2T02: Phase Current B/HV-Side in A

typical value of 500 is used. This value is tuned to get a better 38-2T02: Phase Current C/HV-Side in A

match between simulated and measured curve of the inrush Fig. 4. Simulation of the connection of transformer 2 with ATP (above) and
current during the simulations. Power Factory (below)
To model the remanent magnetism in Power Factory, the
value of the transformer magnetizing flux is set at a In figure 5 the phase currents for the switching of the
predefined value (between 0 and 0.7 p.u.) at the start of a 160 MVA transformer (transformer 4 of system b in figure 1)
simulation. are depicted.
 The shape and amplitude of the inrush currents depend The tuning showed that the amount of remanent magnetism
strongly on the instant of connection for both transformers. and the value of the saturated flux were up to a certain extent
The cause of this is the physical mechanism that causes the interchangeable.
occurrence of the inrush current, which makes them
B. Comparison of simulations and measurements
dependent on voltage at connection instant as well as on
remanent magnetism, as described earlier. In figure 6, one of the measurements of the connection to
the 66 kV busbar is shown for transformer 2 and transformer
2000 4. When these figures are compared to figures 4 and 5, it can
[A] be concluded that for transformer 2 both the amplitude and
1500
shape of the measured inrush currents are quite similar to the
1000 simulation results obtained with ATP and Power Factory.
500 200

-500 100

Current [A]
-1000
0
-1500

-2000
-100
0.008 0.028 0.048 0.068 0.088 [s] 0.108
(f ile SW_160_2.pl4; x-v ar t) c:X0024B-X0001B c:X0024C-X0001C c:X0024A-X0001A

-200
2000.0
0 0,02 0,04 0,06 0,08 0,1
Time [s]

1000.0 2000

1000
0.0
Current [A]

0
-1000.0

-1000

-2000.0
0.00 25.00 50.00 75.00 [ms] 100.
38-1T261: Phase Current A/LV-Side in A
38-1T261: Phase Current B/LV-Side in A -2000
38-1T261: Phase Current C/LV-Side in A 0 0,02 0,04 0,06 0,08 0,1
Time [s]
Fig. 5. Simulation of the connection of transformer 4 with ATP (above) and
Power Factory (below) Fig. 6. Measurements of the connection of transformer 2 (above) and
transformer 4 (below)
Further investigations with ATP on the impact of the
presence of remanent magnetism (which was created by first For transformer 4, the observed discrepancies both with
starting up the simulation with the transformer connected and respect to shape as well as to amplitude are larger, particularly
then disconnecting and again reconnecting it) have shown that in ATP, but to a lesser extent also in Power Factory. It proved
when this was included, the simulation results varied even not feasible to get a better match. Explanations for this
more widely [9]. However, in the rest of this paper, remanent observation could be the that the magnetizing characteristic of
magnetism is not included in the ATP simulations in order to the iron differs more from the assumed characteristic than in
reduce the number of quantities to be varied. case of transformer 2 or that the effect of the simplifications in
The simulations with Power Factory also showed that the the transformers model (e.g. neglecting the third winding of
inrush current is dependent on the instant of connection and transformer 4, which was not present in transformer 2)
on the assumed amount of remanent magnetism. In the Power negatively affect the similarity between simulation results and
Factory model, the instant of connection and the amount of measurements. Further, in some cases the phases had to be
remanent magnetism were tuned together with the value of the exchanged to get a better correspondence. This is, however,
slope of the saturated flux (figure 3) in order to get the best not seen as a major weakness of the model, because in
possible agreement between the simulation results and the practice it does not matter very much which phase exhibits
measurements described in the next section. which behavior, as long as the behavior of each of the phases
in the measurements is reflected in the simulation results as
well.
 The maximum resonance overvoltages found in the
V. SIMULATIONS OF RE-ENERGIZING THE GRID simulations, caused by the transformer impedance and the
The high voltage grid in the North of the Netherlands is a cable capacitance, were 1.9 per unit. Similar voltages
220 kV grid. According to the test procedures for re- appeared at the station side of the cable.
energizing the grid, the black-start unit has to energize the Such voltages are not acceptable, particularly not because
220 kV station it is connected to, and the first load station. in the experiment the grid side of the station was in normal
The latter requires energizing of a 220/110 kV transformer. operation, but also not during a black-start. Therefore another
Therefore the model is extended with the 2.1 km cable to the black-start procedure had to be developed.
220 kV station and the 220/110 kV transformer (transformer
400.0
5). A simplified single line diagram is shown in figure 7.
Before the first large size transformer is energized, the two
generating units run in parallel as shown in figure 7. The 200.0
66/220 kV transformer is energized together with the 220 kV
cable.
0.0

50.0 MVA 50.0 MVA

10.5 kV 10.5/70.5 kV
160.0 MVA cable -200.0
G
~

66.0/220.0 kV 2.1 km

Transformer 3
-400.0
Transformer 4 0.000 0.025 0.050 0.075 [s] 0.1
Term38-1T261: Phase Voltage A in kV
10.5 kV Term38-1T261: Phase Voltage B in kV
220.0 kV Term38-1T261: Phase Voltage C in kV

34.0 MVA 20.0 MVA

10.5 kV 10.6/66.0 kV 350.0 MVA 220.0 kV Fig. 9. Phase voltages during energizing 66/220 kV transformer with cable
220.0/110.0 kV
G

VI. SOFT ENERGIZING OF TRANSFORMERS

~

Transformer 5

Transformer 1 Inrush currents of transformers and resonance overvoltages

can be avoided by gradually increasing the transformer's
10.5 kV 66.0 kV terminal voltage to its nominal value. This is accomplished by
110.0 kV
slowly increasing the field voltage (and thus the terminal
Fig. 7. Total re-energizing scheme voltage) of one of the generators in the system. The ramp-up
feature is often standard available on modern automatic
The simulation of this event is shown in figures 8 and 9 voltage regulators (AVR).
and gives the phase voltages at the 220 kV terminals of the The normal electrical schemes of a power plant are often
transformer. equipped with interlocks to prevent switching operations
Figure 8 shows the energizing of the transformer if the based on potential unsafe or unwanted conditions. As an
cable is not connected. Figure 9 shows the voltages with the example, the logic of the circuit breaker prevents the closing
2.1 km cable connected. of the breaker if the voltage is zero. For a soft energizing of
the transformers it is therefore necessary to override these
400.0
interlocks, if applicable. This means that special procedures
have to be made for a soft energizing of the transformers at
200.0
the power plant or the transformers in a nearby switching
station.
A live test was performed to test the soft start procedure.
0.0 The 50, 160 and 350 MVA transformers were connected to
the 50 MVA generator. The 34 MVA generator and the
20 MVA transformer were disconnected. Next the 50 MVA
-200.0 generator is run up without excitation. The initial voltage at
the terminals of the generator and thus at the terminals of the
transformers are caused by remanent magnetism of the rotor
-400.0
0.000 0.025 0.050 0.075 [s] 0.1 of the generator. The excitation system of the 50 MVA
Term38-1T261: Phase Voltage A in kV
Term38-1T261: Phase Voltage B in kV generator is then started with the AVR in the soft energizing
Term38-1T261: Phase Voltage C in kV
mode. The ramp up time was around 10 s. The test system was
Fig. 8. Phase voltages during energizing 66/220 kV transformer without cable successfully energized without overvoltages occurring at the
terminals of the transformers or at the busbars of the
substations. The results of the test are shown in figure 10 and
figure 11. Figure 10 shows the 66 kV side of the transformer, assumptions must be made. In the latter case, the modeling
figure 11 shows the 220 kV side. At both sides of the approach presented in the paper could be used advantageously
transformer the voltage rise is smooth and reaches its end for formulating these assumptions.
value without overvoltages. Simulation results obtained with the developed models
were presented and it was shown that the shape and the
60 amplitude of the inrush currents strongly depend on the
connection instant, as well as on the assumptions with respect
40
to the amount of remanent magnetism. The simulation results
20 were compared with measurements. It was concluded that the
degree of correspondence was satisfactory, although
Voltage [kV]

0 differences remain, which vary from case to case. Factors

explaining these could be discrepancies between the shape of
-20
the modeled and the real magnetizing characteristic and the
-40 impact of remanent magnetism. The developed models were
used to simulate the re-energizing of the grid.
-60 Based on the simulation results, the procedure was
0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 9,0
Time [s]
modified to use soft energizing instead of direct switching.
Live tests were performed using the new procedure, resulting
Fig. 10. Phase voltage during soft energizing at the 66 kV side of the in the confirmation of the unit's black-start capability.
66/220 kV transformer

200
VIII. ACKNOWLEDGMENT
150
Delesto B.V. is acknowledged for carrying out the
measurements, providing the equipment data and supporting
100
the work reported in the paper. Assistance of TenneT bv and
50 Essent Netwerk B.V. was available for the switching
Voltage [kV]

0 operations in the 220 kV and 110 kV Public Grid of the

-50
Netherlands.
-100
IX. REFERENCES
-150
[1] C. Saldana, G. Calzolari, ‘’Methodology Utilized in Black-start Studies
-200 on EHV Power Networks,’’ presented at the IV International Conference
0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 9,0 on Power System Transients, IPST’2001, Rio de Janeiro, Brazil, 2001
Time [s] [2] C.P. Cheng, Shihe Chen, ‘’Simulation of Resonance Over-voltages
during Energizing of High Voltage Power Network,’’ presented at the V
Fig. 11. Phase voltage during soft energizing at the 220 kV side of the International Conference on Power System Transients, IPST’2003, New
66/220 kV transformer Orleans, Louisiana, 2003
[3] M.M. Adibi, R.W. Alexander, B. Avramovic, ‘’Overvoltage Control
During Restoration,’’ IEEE Transactions on Power Systems, Vol 7, No4,
VII. CONCLUSION Nov 1992
Re-energizing of the grid after a major blackout can cause [4] M.J. Heathcote, 1998, The J&P Transformer Book, Reed Educational
and Professional Publishing, Oxford.
problems because of the need to energize large transformers [5] J.J. Winders, 2004, Power Transformers, Marcel Dekker, New York.
with relatively small generators. Together with the cable that [6] http://www.allaboutcircuits.com/vol_2/chpt_9/12.html
connects the transformer to the grid, overvoltages can occur at [7] Can/Am EMTP User Group, 1995, Rule Book-Alternative Transient
the terminals of the transformer and at the busbars of Program, Oregon.
[8] DIgSILENT, 2004, Power Factory Version 13; User Manual,
substations. Such problems can be avoided by soft energizing Gomaringen.
the transformers. This requires the black-start unit to be [9] M. Rioual, C. Sicre, ‘’Energization of a no-load transformer for power
equipped with possibilities to make the appropriate restoration purposes: Impact of the sensitivity to parameters,’’ presented
at the IV International Conference on Power System Transients,
connections (interlock overrides) and a soft energizing mode IPST’2001, Rio de Janeiro, Brazil, 2001
on its automatic voltage regulator.
Simulations were performed to test energizing procedures X. BIOGRAPHIES
of a particular power plant prior to live testing. First, the
physical origin of inrush currents was described. Then, it was Kees P.J. Jansen received the MSc degree in Electrical
discussed how inrush currents can be simulated using two Engineering from the University of Technology
state of the art power systems simulation software packages. Eindhoven (TUE), The Netherlands, in 1982. From 1983
to 1986 he was with ASEA in Sweden. Since 1986 he
Particular attention was paid to the modeling of the works at KEMA in Arnhem, The Netherlands. He holds a
transformer including magnetic saturation. It was indicated senior consulting position with KEMA. He is involved in
that in most cases not all necessary data is available, so that system studies and network design and calculations for
public networks and industrial installations.
either measurements must be available to tune the model or
Han G. Slootweg received a MSc degree in electrical
engineering in 1998 (cum laude) and a PhD degree in
2003, both from Delft University of Technology, the
Netherlands. He also holds a MSc degree in business
administration. Currently, Han works with Essent
Netwerk B.V., one of the largest Distribution Network
Operators of The Netherlands, where he is responsible
for the coordination of the maintenance and the long
term planning of the company's 110 kV and 150 kV
networks.

Richard A.C.T. de Groot MIEE received the MSc

degree in Electrical Engineering from Delft University
of Technology, The Netherlands, in 1992. Until 2000 he
was with Holec Projects (now Strukton Systems) as
electrical project/design engineer for power plants and
substations. He is currently with KEMA in Arnhem,
The Netherlands, as senior consultant in Power System
Design and Analysis. He is involved in system studies,
network design and connection studies.