Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
3, MAY 1995
i~ R1 L1
Ferroresonance is usually initiated by a system disturbance of
some configuration. For example, one or more of the supply
conductors to an unloaded or very lightly loaded power
transformer are switched or interrupted, leaving a transformer
coil energize through the capacitive coupling with the other
phases or with parallel lines. Thus, in order to be able to 1 I
predict whether or not ferroresonance will occur, the system Fig. 1. Single phase equivalent scheme of transformer.
conditions immediately prior to the disturbance must be
accurately defined. In a no-load test high voltage transformer, the rms voltages,
The ferroresonance problem, whether in a general circuit or in currents, and core losses are measured in secondary (low
a power-system circuit, is usually tackled in one of two voltage). In this case, the magnetizing current io is equal to
ways: circuit configuration (i.e. series and/or parallel) and the secondary current i2. The total current io comprises two
nonlinearity (i.e. saturation and hysteresis of the transformer), parts: the main saturation current is and the hysteresis current
since, fundamental factor to the occurrence of ferroresonance ih (current of loss part). The main saturation flux-current
is the nonlinear magnetic behaviour of the transformer [11. curve is = f(h,A) can be represented by a pth odd order
Ferroresonant phenomena have been observed or presented in polynomial :
[l-71, but the magnetization characteristics in these papers are
very difficult to obtain. A1 k. + Ap hP p = 3, 5, 7 ...
i= @.U
Therefore this paper presents a new method for determination The hysteresis current is determined by a qth even order
of instantaneous magnetization characteristics of transformers polynomial :
(saturation and hysteresis loop) by taking into account only
the classical rms values and no-load losses. This model is
used to accurately study ferroresonance phenomena.
The magnetizing current can be determined by :
11. PROPOSED MODEL OF NONLINEAR MAGNETIZATION
CHARACTERISTICSOF A TRANSFORMER @.3)
From measured rms voltages, rms currents and power losses, Where I is flux linkage; A and D are sets of constants and
the main saturation curve and the equivalent hysteresis loops these constants will be determined from the measured data by
are obtained. The required data for the present method are using the least squares curve fitting method.
easily available from the final test in the manufacturer's Assume that the flux varies sinusoidally as a function of
report [8]. After the main saturation curve has been obtained, time:
it can be combined with the dissipation function to establish
equivalent hysteresis loop. The dissipation function h = hm sin rot W.4)
determines total energy losses in the transformer core The no-load losses can be found as equivalent to the area of
(without separation between eddy current and hysteresis hysteresis loop:
losses). It is possible to improve the main saturation curve
by taking into account no-load losses. This model is
characterized by only two passive nonlinear components (II.5)
R
(RNL and LNL in Fig. 1).
Authorized licensed use limited to: Florida Institute of Technology. Downloaded on January 31, 2009 at 11:01 from IEEE Xplore. Restrictions apply.
2061
where v=
dt
*=
k m cos at = V,, cos at
In this paper, a three-phase 62.5/20 kV, 36 MVA wye-wye
ungrounded power transformer is used. By using the proposed
method, the main saturation curve of this transformer is
.
I-RMS (A)
Fig. 2. V-RMS - I-RMS curve in f a phase a (only 3 measured points).
2062
-80
-5 I I
0 100 200 300 400 500 600
Time (ms)
Fig. 8. Primary voltage waveform of the transformer in phase a.
Fig. 6 shows the calculated hysteresis loop of the transformer
at V = 1.1 pu.
111. FERRORESONANCESTUDY
In order to study ferroresonance phenomena, in this part, as in
part 11, a three-phase 62.5120 kV, 36 MVA wye-wye
ungrounded power transformer is used. This transformer is
supplied over a 2.2 km underground cable from a 63 kV bus,
through a three phase breaker. This circuit (Fig. 7) is used to -1
observe ferroresonance phenomena in case of energization . .
................. i................. :..I..!...!.......i.................4 .................4 .................
(switching) or de-energization (opening) of the transformer.
-3 ..................i.................i..................i .................i.................i..................
..................-..................................................... .-.................-.................
same, the overvoltages reach to 4.7 pu. After the closure of waveform of the transformer in phase c. There is a high
phase a, there is a high frequency oscillation in phase c. Then distortion in secondary voltage of the transformer and this
after the moment of closing the phase c, the voltages in voltage tends slowly to steady state. Fig. 13 shows the
phase a, b, and c tend to steady state (1.Opu). primary current waveform of the transformer in phase b. The
magnetizing current tends slowly to symmetric steady state.
v-tn For the second case, the transformer is de-energized. The
I
phase a is opened (switched-off)at the instant the voltage in
phase a passes zero. The ferroresonancephenomenon occur.
Fig. 14 shows primary voltage waveform of the transformer
in phase a. The overvoltage can be reached to 4.0 pu.
. V(DU) v-Pa
Time (ms)
Fig. 11. Neutral voltage waveform of the transformer.
v (PU) v-sa
2.5
................. ........... ....i.............. ..................i..................i.................
-0.8I
0 100 200 300 400 500 600
Time (ms)
Fig. 13. Primary current waveform of the transformer in phase b.
Fig. 11 shows the neutral voltage waveform of the
transformer (see a subharmonic component, with the same
form in primary voltage in phase a and c), the overvoltage
can be reached to 10.0 pu. Fig. 12 shows secondary voltage
Authorized licensed use limited to: Florida Institute of Technology. Downloaded on January 31, 2009 at 11:01 from IEEE Xplore. Restrictions apply.