You are on page 1of 6

Magnetic Saturation at Short-Circuit Tests on

Power Transformers
A.L.J. Janssen, L.H. te Paske, W.A. van der Linden, R.P.P. Smeets
KEMA High-Power Laboratory
Arnhem, the Netherlands. e-mail: hpl@kema.nl
Abstract: The post-set andpre-set short-circuit test methodsfor
power transformers are discussed, K&VIIb testing expedience
withspecial attention to the pre-set method, inrushcurrents and
prernagnetization, are highlighted, By means of pre-magneti-
i:ationinrushcurrents can be avoided to a large extend.
The saturation of the supply side transformers leads to a
distortionof the supplyside voltage, thuspreventing saturationof
lhe test objects. The dl~ferentsaturationmodes of the transformer
under test are shortly mentioned in relation to the dynamic
~:tresseson the windings.
Keywords: Power transformers, short-circuit tests, pre-set short
circuit, inrush currents, pre-magnetization, magnetic saturation
1. INTRODUCTION
In order to test the ability of large power transformers to
withstand short-circuit currents, two methods of applying
the short-circuit are defined in the Standards: the post-set
short-circuit and the pre-set short-circuit method.
A. Post-set short-circuit
With the post-set method the transformer is energized at
one side with the other side open. After the inrush currents
have disappeared the short-circuit is switched on at the
other side.
Before short-circuiting, the source voltage is applied at the
terminals of the transformer, so that the source voltage has
to be limited in relation to the rated transformer voltage.
To the IEEE :Standards [1] a maximum source voltage of
[10% related to the rated tap voltage of the transformer
under test is allowed and to the IEC Standards [4] 115/0.
13.Pre-set short-circuit
With the pre-set method one side of the transformer is
short-circuitecl before the application of the voltage. The
source voltage may differ to a large extend from the
transformer voltage, as long as the specified short-circuit
current flows. For testing, the required short circuit power
is much lower than the power for testing in post-set
condition. [6].
The post-set short circuit method is mostly applied in test
stations that are supplied from the grid. There is enough
power available, but the main drawback is the difficulty to
achieve the correct voltage, especially for the maximum
tap position. The main drawback of test stations with own
generators is the limit of the available power. At KEMA
High Power Laboratory, being the largest short-circuit
laboratory in the world with 4 generators and a short-
circuit power of about 10.000 MVA 3-phase and
5.800 MVA l-phase (at 60 Hz), the pre-set short-circuit
method is quite often applied, mostly in the 1,5 phase
method. Hereby, the phase under test is connected in series
to the other two parallel connected phases in a Y-con-
nection (for a schematic outline, see fig. A 1 in Appendix).
Note that the power required fi-om the test station is much
smaller than the short-circuit power in the network (20.000
MVA @ 245 kV to 83.500 @ 800 kV).
The fact that the inrush current is superimposed to the
short-circuit current is inherent to the pre-set method. But,
due to the short-circuited windings, the inrush current is
not comparable with a normal inrush current or the inrush
current of the post-set method. The effects of the inrush
currents during pre-set short-circuits are discussed in this
report.
II INRUSH CURRENTS
Both manufacturers and testing authorities worry about the
non-linear effects of the magnetic flux and the residual
magnetism, resulting in magnetizing currents and inrush
currents.
Under no-load conditions, the energization of a trans-
former leads to a sinusoidal magnetic flux, that is flowing
completely in the magnetic core: pm(ij. For a particular
transformer with a given magnetic characteristic, the flux
and therefore the inrush current is depending on the
amplitude of the voltage applied, on the phase angle of the
voltage applied and on the residual flux (@r):
V(t) = U sin(o t +u)
(jm(t) = @ [1- COS(Ot+cx)]+@r
With the post-set testing method the applied voltage can be
as large as 11So/O of the rated tapping voltage, thus gener-
ating inrush currents, which are far larger than those under
normal service conditions. A very accurate and very
reliable controlled switching (synchronous switching) of
the HV-circuit-breaker, to energize the transformer at
voltage maximum rather than at voltage zero, is a
prerequisite for the post-set testing condition [8], [9].
0-7803-7031-7/01/$10.00 (C) 2001 IEEE
0-7803-7173-9/01/$10.00 2001 IEEE 1830
Opposite to the no-load situation, the energization of a
short circuited transformer gives a large component of the
magnetic flux flowing outside the magnetic core,
The magnetic flux consists of three components:
@m the part flowing completely through the core and
forming the coupling between the windings
@ow the part of the leakage flux that flows through the
gap between the windings and returns through the
magnetic core
@ol the part of the leakage flux that flows through the
gap between the windings and does not return via
the magnetic core, but via oil or tank.
Supposing that the applied voltage is identical to that of the
no-load energization, the at-component of @m under
short-circuit conditions is much smaller than under no-load
conditions (40 to 600/0), caused by the voltage drop across
the leakage impedance. As a part of the leakage flux
(@om) returns via the magnetic core, that part of the core
has to conduct @m + @rim. But, even then the ac com-
ponent of @ZU+ CPgmis smaller than the ac component of
@m under no-load conditions. So, with the same state of
residual flux and the same voltage angle at energization a
short-circuited transformer gives less saturation effects,
only local saturation (see chapter 6) and lower inrush
currents.
Compared to the random energization under normal
service conditions, the above is an advantage for the pre-
set testing method. In addition, the applied voltage is 100/0
in comparison to the 110 to 115/0 used with the post-set
method. The main drawback of the pre-set testing method
is that the voltage angle has to be the most onerous one: a
100% offset of the flux is inherent to the requirement of
testing with a.full asymmetrical short-circuit current.
With the post-set method a magnetic flux is already flow-
ing through the core, so that at the initiation of the short-
circuit (at voltage zero), the flux has the maximum ampli-
tude belonging to the terminal voltage (that can be as high
as 115%0). Ccmsequently, after short-circuiting the flux will
follow smoothly a COSfunction, superimposed to -1 15%.
Again a vet-y accurate and reliable closing of the high-
vohage circuit-breaker, now applied at the short-circuit
side, is required.
Post-set
a. Initiation of magnetization (most onerous moment):
V(t) =1. 15i7 sin(m t)
(p(t) = 1.15@[l-cos((o t)]+mr =>q)Peak =2.3@+@,
with U and @ as 1 pu.
b. Initiation c}fmagnetization (optimal moment):
v(t) = 1.15U Cos((llt)
T(t) =l.15@sin(of)+@r $pea/c 1.15@+@r
c. At initiation of short-circuit:
V(t) =1. 15U sin(o t)
(f(f) = 1.15@cos(ol t)
d. After initiation of short-circuit:
V(t) = U sin((ir t)
(p(f) = -o.15m Ocos(co f)
Pre-set
V(t) = U sin(~ t)
Q(t) =@@COS(13t)+@r >~peak =2@+@7
L
+
1 4 pre.magnltlzingshot + full test I
0
Fig. 1 Pre-magnetization of magnetic core
III. PRE-MAGNETIZATION
The extra magnetizing current forced by the residual flux,
that results fi-om the former test, can be influenced by
controlling the direction or polarity of the residual flux.
KEMA takes advantage of the residual or remanent flux
(remanence) by deliberately pre-magnetizing the magnetic
core in the opposite polarity, see fig. 1.
When necessary, the correct pre-magnetization is forced
by one or two short 70~o to SOY.tests preceding the succ-
essive 10OO/. tests. Now the remanent flux will be subtract-
ed instead of added to the transient flux. In this way,
during the 100V0test, the saturation effects are limited.
Such intermediate tests are regarded as harmless for the
transformer under test as the forces and stresses are less
than 50% compared to the full tests. See [11].
IV CONTROLLING THE RESIDUAL FLUX
A possibility to estimate the residual flux is the appearance
of inrush currents. Inrush currents appear when the mag-
netic core gets saturated, meaning that the residual flux
after switching off is at maximum or close to its maximum
possible value. An example of a test series is shown in fig.
2.
Starting with the residual flux from the former test, the
voltage excitation (which is a sine function in order to
achieve the full asymmetric short-circuit current) leads to
an ac component in the flux and a de-component:
0-7803-7031-7/01/$10.00 (C) 2001 IEEE
0-7803-7173-9/01/$10.00 2001 IEEE 1831
I I
k
.
Irl
1s2
l!!L_-
.

. ..
I
test 06
A T ------------ ______, ._,
I
test 05 I
1
Ust
,---------------
Irl
1s2
Fig. 2: Test results with various degrees of saturation (see text). Upper traces (Ust): voltage (20 kV/div); middle traces (Irl): source side current (20
kA/div); lower traces (1s2): current at short circuited side (20 kA/div). Time: 20 ms/div.
F
resldud
start
TEST 00
excitation +0,5
F
residual
start
TEST 04
excitation +1.0
- start
-resldua(
SATURATIDb(
TEST 01
excitation -0,7
max.
resldua
.~
start
SATURATION
- start
- residual
TEST 02
excrhtlon -1,0
Flax,
resldua~ start
- residual
- start
TEST 03
excitation +1,0
star7
SATURATID~

~esldua(
TEST 05 TEST 06 TEST 07
excitation +1.0 excrhitlon +1.0 excltmtlon -1.0
Fig, 3: Schematic representation of variation of remanent magnetization in the course of the tests, shown in fig. 2.
0-7803-7031-7/01/$10.00 (C) 2001 IEEE
0-7803-7173-9/01/$10.00 2001 IEEE 1832
V(t) = U sin(m t)
(p(t) =CD, +CD-alcos(cot)
After a few cycles the de-component of the short-circuit
current has disappeared to a large extend. (But, like for the
inrush current, the time constants for the remanent flux and
the de-component of the flux are much larger and its
decrease is therefore neglected.) The short-circuit current
will be interrupted at a current zero, meaning that cos(cot)
is zero. The new residual flux is then ~r + ~.
This is illustrated in fig. 3, where the evolvement of the
residual flux during succeeding tests (fig.2) is given.
Ailer saturation one or two full tests are possible before
again saturation (but of opposite polarity) is reached (tests
03 and 04). The residual flux cannot be larger than the
maximum given in the BH-curve, as shown at tests 05 and
06.
P
Irl
.
-,,
I
int(Ust)
Im
Fig. 4: Test result with high saturation
Upper oscillogram: upper trace (Ust): voltage (20 kV/div); middle trace
(Irl): source side current (20 kA/div); lower trace (1s2): current at short-
circuited side (20 kA/div).
Lower oscillogram: upper trace: integrated voltage Ust (Int(Ust)) (flux)
(33 Vs/div); lower trace (Ire) magnetizing current 10 kA/div.
Time: 20 ms/div,
The magnetization current can be determined by subtract-
ing the secondary current multiplied with the transformer
ratio, from the primary current. In figure 4 the inrush cur-
rent belonging to test 02 is shown. Also the flux is shown
as the integral of Ust, but the starting point (i.e. the residual
flLlxof the former test) is missing.
An alternative way to force the transformer into saturation
is by means of a low voltage de-source. Such a method is
presently under investigation.
V SATURATION OF THE SOURCE
The pre-magnetization procedure explained in the former
chapters helps also to prevent saturation of the step-up
transformers of the High-Power Laboratory, used to
supply the transformer under test. However, another aspect
has to be mentioned.
Saturation of the supply side transformers gives inrush cur-
rents at the generator side, but at the load side the current is
decreased. In fig. 5 the load side current of the saturated
transformers of the High-Power Laboratory can be seen.
This current, flowing into the transformer under test,
shows no inrush peaks at all and is flattened in comparison
to the source side currents (Irl ) in the figs. 2 and 4. This
means that saturation of the source leads to lower stresses
on the test object.
To evaluate the short-circuit tests, the total short-circuit
current is measured, including components that can be des-
cribed as magnetization currents. The peak value of the
current measured is compared with the specified current
peak in the Standards.
In case of saturation in the transformers at KEMAs test
plant, it maybe a difficult task to achieve the required peak
values of the short-circuit current. Caretid estimation of
the residual flux and adequate pre-magnetization together
with a full utilization of the capabilities of the High-Power
Laboratory are necessary.
VI SATURATION EFFECTS OF THE TRANSFORMER
The purpose of short-circuit tests is to verify the ability of
a certain design to withstand large currents. The local
stresses in a winding are determined by the current flowing
through that winding and the magnetic induction at that
location. The magnetic field direction and the magnetic
field strength are influenced by saturated parts of the mag-
netic core.
The leakage flux will partly follow the magnetic core and
stimulate saturation. The pattern of the leakage flux is de-
pending on the winding that is short-circuited (inner or
outer winding) and the tapping windings (and tap used).
Split windings give even more complicated patterns.
In the gap between the primary and secondary windings
the leakage fluxes of the two windings flow in the same
direction. The leakage flux of the inner winding returns
mainly through the limb. That of the outer winding flows
partly through the yokes, partly through other iron parts of
the transformer and partly through the free space in oil.
The returning leakage flux of the short-circuited winding
will flow opposite to the main magnetic flux, so reducing
0-7803-7031-7/01/$10.00 (C) 2001 IEEE
0-7803-7173-9/01/$10.00 2001 IEEE 1833
the saturation, if any. The returning leakage flux of the
supplying winding will flow in the same direction, in this
way stimulating the saturation in either the yoke (supply to
the outer winding) or the limb (supply to the inner wind-
ing). Short-circuiting the outer winding and supplying the
inner winding is therefore regarded as giving larger
impedances, different currents, different electro-dynamic
stresses.
7. Saturation of the test plant transformers has the effect to
reduce the inrush current in advantage to the transformer
under test.
magnetizing currents than the other way around.
REFERENCES
I
Ust
1s2
Fig. 5: Test result with saturation of the source. Upper trace (Ust):
voltage (20 kV/div); lower (1s2): current at short-circuited side (1O
VII. CONCLUSIONS
1. At short-circuit tests, switching at or around voltage
zero is a must to achieve the proper peak values of the
short-circuit currents. But the voltage drop across the
leakage impedance reduces the flux in the magnetic core in
comparison to the no-load energization.
2. Based on the direction of the flow of the leakage flux
through the magnetic core, it is, generally speaking, prefer-
red to supply the outer winding and to short-circuit the
inner winding in order to reduce its contribution to local
saturation.
3. An accurate and reliable controlled switching is a means
to reduce the inrush currents during the energization of
transformers subjected to the short-circuit test under post-
set conditions where voltages of 110/0 to 115/0 are
applied.
4. With the pre-set testing method, the problem of the
remanent magnetism can be turned into an advantage by
controlling it in a way that the residual flux is subtracted
instead of added.
5. The polarity of the short-circuit current at the instant of
making has to be choosen properly. In case necessary,
intermediate premagnetization shots may be required to
ensure a considerable residual flux.
6. One has to be aware that the saturation effects are dif-
ferent for no-load and short-circuit energization. More-
over, the (local) saturation is depending on supply to the
inner or the outer winding, on the tap position, on the
configuration of the windings and on the magnetic history
of the core. The (local) saturation leads to different
[1] IEEE Std C57. 12.00-1993: IEEE Standard General Requirements
for Liquid-Immersed Distribution, Power, and Regulating Transform-
ers
[2] IEEE Std C57. 12.90-1993, Prut I: IEEE Standard Test Code for
Liquid-Immersed Distribution Power, and Regulating Transformers
[3] IEEE Std C57. 12.90-1993, Part II: IEEE Guide for Shofi-Circuit
Testing of Distribution and Power Transformers
[4] IEC Standard 60076-5 (2000): Power transformers - Part 5: Abili-
ty to withstand short circuit
[5] G. Leber, Investigation of Inrush Currents during a Short Circuit
Test on a 440 MVA, 400 kV GSU-Transformer, CIGRE SC12 Sess-
ion 2000, report 12-104
[6] A.L.J. Janssen, L.H. te Paske, Short-circuit Testing Experience
with Large Power TransformersCIGRE SC12 Session 2000, report
12-105
[7] G.Bertagnolli, Short-circuit Duty of Power Transformers Book
printed on behalf of ABB Trasformatori,Legano (Milrrno), Italy
[8] J.H.Brunke, K.J.Frohlich, Elimination of Transformer Inrush
Currents by Controlled Switching, Part I - Theoretical Considerations
IEEE-PES, preprint PE-019PRD (09-2000)
[9] J.H.Brunke, K.J.Frohlich, Elimination of Transformer Inrush
Currents by Controlled Switching Part 11 - Application and Per-
formance Considerations IEEE-PES preprint PE-120PRD (09-2000)
[10] G,Bertagnolli, Effects of Short-circuit Currents in Power
Transformers Key Note speech at the 8th International Symposium on
Short-Circuit Currents in Power Systems, Brussels 1998
[11] A.L.J. Janssen, L.H, te Paske, R,P,P, Smeets, A,G.A, Lathouwers,
Short-Circuit Testing: a Crucial Instrument for the Design Asessment
of Large Power TransformersCEPSI 2000, Conference Proceedings
pp. 344-351
[12] N.V.C. Sastry, H. Gupta, A.L.J, Janssen, A.G,A, Lathouwers,
Considerations about Short-Circuit Tests on EHV Power Trans-
formers, 9th Symposium on Short-circuit Currents, Cracow (Poland)
2000, paper 2.7
[13] N.V.C. Sastry, H.Gupta, Short Circuit Tests on EHV Trans-
formers 9th Symposium on Short-circuit Currents, Cracow (Poland)
2000, paper 2.6
[14] G,Bertagnolli. et al., Power Transmission Reliability - Technical
and Economic Issues Relating to the Short-circuit Performance of
Power Transformers CIGRE SC12 Session 2000, report 12-107
[15] J.Foldi, D,BLrub4, G. Bertagnolli, R.Maggi, P,Riffon, Recent
Achievements in Performing Short-circuit Withstand Tests on Large
Power Transformers in Canada CIGRE SC12 Session 2000, report
12-201
B1OGRAPHIES
Anton Janssen received his M.Sc.-de-
gree in electrical engineering from the
Eindhoven University of Technology in
1977, After a short period in the
consultancy & contracting business, he
joined PNEM, a power company in the
southern part of the Netherlands, and
became manager of the department of
Transmission, Iateron the department of
Projects (Transmission and Co-generat-
ion). In 1993 he joined KEMA and
0-7803-7031-7/01/$10.00 (C) 2001 IEEE
0-7803-7173-9/01/$10.00 2001 IEEE 1834
became general manager of KEMA High-Power Laboratory. Mr.
Janssen is a member of CIGRE (SC13, Special Reporter, CIGRE
WG13.06 and 13.08) and IEEE (IEC/IEEE 17A/WG23).
Henk te Paske was born in Winterswijk,
the Netherlands in 1953. He received the
B.SC. degree for electrical engineering
from the Technical College of Enschede
in 1977. He joined KEMAs High-Power
Lab, in 1979 as a test engineer and is
since 1987 employed as a senior test-
engineer. He has been actively involved
in testing, quality surveillance, on site
testing and failure analysis. He is a mem-
ber of the Dutch standardization com-
mittees NEC 17A/C and NEC.
Wim A. van der Linden was born in
Escharen, the Netherlands in 1941. He
received the B.SC. degree for electrical
eng. from the Technical College of s-
Hertogenbosch in 1963. He joined
KEMAs High-Power Lab. in 1965 as a
test engineer and is since 1972 employ-
ed as a senior test-engineer. He has been
actively involved in the design and
construction of new test equipment in
KEMA High Power Lab. He is a mem-
ber of the IEEE Working Group on syn-
thetic testing of generator circuit-breakers.
Ren6 Peter Paul Smeets received the
M.SC. degree in physics from the Eind-
hoven Univ. of Technology in 1981. He
obtained a Ph.D. degree for research
work on vacuum arcs. Until 1995, he
was an assistant professor at Eindhoven
University. During 1991 he worked
with Toshiba Corporations Heavy
Apparatus Engineering Laboratory in
Japan. In 1995, he joined KEMA. At
present, he manages the R&D activities
of KEMAs High Power Laboratory. He
[3.04, WG13.12 the Current Zero lrb>
IEEE and CIGRE
TEST-CIRCUIT DIAGRAM
EPOFIT No.
m,l CIRCIJIT Mo.
CIRCUIT COMPONENTS
MEASUREMENTS
G = Generator TO = Test Object
U = Voltege Measurement
MB = Master Breaker
I = Current Measurement
MS= Make Switch
PT = Power Transformer
R = Resistor
C = Capacitor
L = Inductance
di/dt lR
n
Transformer under test
I I ~- - .,
11 I
E
]rIr
: : ?
~::.,
u: ;~;
~::
J[z u
=
L .1
N
I&
$
I ~~
Phase R under test
v
7 v =
-,----- -1-
YNdl 1 transformer in single-phase test circuit 1.5 test method KEMA~
Fig. Al Three-phase transformer under test in a single-phase test circuit (1.5 phase method)
0-7803-7031-7/01/$10.00 (C) 2001 IEEE
0-7803-7173-9/01/$10.00 2001 IEEE 1835

You might also like