# Dynamical behaviour analysis of energetic transformers using DIgSILENT program.

Abstract: Dynamical behaviour analysis of energetic transformers is presented in this paper. Physical and basic mathematical background necessary for dynamical behaviour analysis is explained. Case of switching unbiased transformer on gird and switch current analysis are described. Switch voltage appearance in cases transformer switching and breaking from gird in normal drive is explained. Simulations of energetic transformer switching and breaking from gird using DIgSILENT program is given.

Key words: analysis, switching, transformer, simulation,

DIgSILENT.

1. INTRODUCTION

The main topic of this paper is model for testing power transformer in cases of inrush current appearance, when unloaded transformer is switched on gird, and case of appearance of switch voltage in normal drive. This testing is made in DIgSILENT program on 50 MVA power transformer. The term power transformer (Fig. 1) is used to refer on those transformers used between the generator and distribution circuits, and these are usually rated at 500 kVA and above ( Sim & Digby, 2004). Inrush current is the magnetizing current that creates during switching of unloaded transformer on gird (Kelemen, 1997). Inrush current appears because of residual flux that remains in the core due to the properties of the magnetic core materials. Magnitude of this current can be 3.5 to 40 times bigger then rated full-load current and can cause many problems. This inrush current can have an effect on the operation on relays and fuses located in the system near the transformer (Sim & Digby, 2004). Model for testing transformer on inrush current is also made in ATP program (Maric, et all., 2006) where presented results can help in prevention of undesirable effects that inrush current causes. Controlled switching can reduce the undesirable effect of inrush current. The actual controlled switching devices operate with high voltage circuits breakers but only in normal cases: usual manual switching of network components like power transformers, lines loads, etc. (Munteanu, 2006).

The other problem that is explained in paper is the appearance of switch voltage that creates when transformer is switched on or off gird. Because of its high frequency, speed and magnitude switch voltage can cause problems on grounded areas of transformer.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

2. INRUSH CURRENT

The waveform of typical inrush current (Fig. 2a) displays a large and long lasting DC component, rich in harmonics, assumes large peak values at the beginning (up to 30 times the rated value), decays substantially after a few tenth of a second but its fully decay occurs only after several seconds to the normal exaction level of 1 - 2 % of rated current (Maric, et all., 2006). The peak of inrush current is limited only by the air-core reactance (Fig. 2b) and it depends of coil inductance L. Therefore, the peak inrush current is expressed in the cgs system of units as follows (Winders, 2002):

[I.sub.peak] = [[phi].sub.r] + 2[[phi].sub.n] + [[phi].sub.s])l x [10.sup.-8] / 0,4[pi]N[A.sub.mt] (1)

where [[phi].sub.r] = residual flux, [[phi].sub.n] = normal flux change, [[phi].sub.s] = saturation flux and N = number of turns in the coil.

3. SWITCH VOLTAGE

Switch voltages are transition stages in gird and they appear (for example) when we have two circuits (Fig. 3) with transformer (presented with [R.sub.k] and [L.sub.[sigma]]) on one side and load on the other side (presented with [L.sub.2] and [R.sub.2]). After circuit brake, because of capacity [C.sub.1] and [C.sub.2] between transformer and load difference voltage is created. This difference appears on circuit breaker and can cause such amounts (2 to 3 time of nominal voltage value) that can depredate isolation of transformer, load and circuit breaker. For the voltages on capacitors [C.sub.1] and [C.sub.2] after circuit breaking relation (Kelemen, 1997) that shows significant values of switch voltage is:

[u.sub.C(t)] = [u.sub.Cs(t)] + ([DELTA][u.sub.0] cos vt + [DELTA][I.sub.0] [square root of L/C] sin vt)[e.sup.t/T] (2)

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

where is [u.sub.C(t)] stationary voltage value on capacitor, [DELTA][u.sub.0] is difference between voltage value before circuit brake (t = 0 s) and [u.sub.Cs(0)], [DELTA][i.sub.0] is momentarily current value trough circuit breaker in the moment of opening, [square root of L/C] is wave resistance of circuit, v is frequency and T - 2L / R is time constant of circuit. Time duration of wave is few hundred micro seconds.

4. EXPERIMENTAL MODEL

Experimental model (Fig. 4) for switching unloaded transformer on gird (where inrush current is been recorded) and breaking of loaded transformer (where switch voltage is been recorded) consists of generator (60 MVA), generator transformer (Tr 1, 50 MVA), bus (110 kV) and tested transformer (Tr 2, 50 MVA) with consumers on LV 10,5 kV side.

Testing of transformer is done in DIgSILENT program that is engineer software tool for analysis industrial, distributional and commercial electrical girds. Programming in DIgSILENT starts by defining of all elements given in scheme on Fig. 4. Work window (Fig. 5) consists of Main window--1, Data manager window--2, Graphic window--3 and Output window--4. Basic data about tested and generator transformer are entered by pressing 2-Winding Transformer order. The model of tested transformer (Tr 2) and other elements are real and DIgSILENT program provides entering of all known data of tested model elements by simple click on each element. By pressing on command Load Flow data about magnetize current and magnetize losses are given. This data is relevant for inrush current determination.

5. RESULTS

Switch of unloaded transformer was started at 0.5 s from the beginning of simulation on HV--110 kV side (Fig. 6). In the beginning current is 2.8 times higher from nominal value (262 A). Time that is necessary for inrush current droop on 50% value of maximum amplitude is 72 s for 50 MVA power transformer (Kelemen, 1997). Loaded transformer switch off is done at 0.5 s on LV-10.5 kV side. On Fig. 7 switch voltage and current amplitude drop is given. Amplitude of switch voltage is 6.07 kV. Voltage amplitude drops on nominal value after few hundred micro seconds.

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

6. CONCLUSION

DigSILENT program provides real testing of observed 50 MVA energetic transformer on appearances of inrush current and switch voltage. By settings on the tested model the worst case scenario for inrush current can be tested. This is the case when switch of transformer occurs when voltages in all three phases are passing trough zero. Also important is transformer testing on switch voltage because this testing provides information about switch voltage magnitude that can occur on circuit breaker and transformer. Combined testing on model and implementation of results provided by real measurement can give good results and evaluation of transformer dynamical behaviour analysis.

7. REFERENCES

Kelemen, T. (1997). Transformator, In: Tehnieka enciklopedija (Technical encyclopedia) HRZ Sv. 13, Podhorsky, R. (Ed), pp. 149-168, HLZ, ISBN 953-6036-50-9, Zagreb

Maric, P.; Nikolovski, S. & Baus, Z. (2006). Simulation of 300MVA Transformer Energization in new substation 400/110kV Ernestinovo using ATP-EMTP, Proceedings of EEUG, Hopfner, S. (Ed.), pp. 57-65, Tehnische Universitat Dresden, September 2006., EEUG Committee, Dresden

Munteanu, F.T. (2006). EMTP Methods and Algorithms for Inteligent Switching, Proceedings of EEUG, Hopfner, S. (Ed.), pp. 30-40, Tehnische Universitat Dresden, September 2006., EEUG Committee, Dresden

Sim, H.J. & Digby, S.H. (2004). Power Transformers, In: Electric power transformer engineering, Harlow, J.H. (Ed.), pp. 2-23, CRC Press LCC, ISBN 0-8493-1704-5, Boca Raton, Florida, USA

Winders, J.J. (2002). Power Transformers Principles and Applications, Marcel Dekker, Inc, ISBN 0-8247-0766-4, New York, Basel