USING OVER-DAMPING METHOD TO SUPPRESS THE FERRO-RESONANCE OF POTENTIAL TRANSFORMER

Transcription

1 USING OVER-DAMPING METHOD TO SUPPRESS THE FERRO-RESONANCE OF POTENTIAL TRANSFORMER Lai Tianjiang, Lai Tianyu, Lai Qingbo Dalian Electric Power Company, China 1 Forward In power system, the phenomenon of Ferro-resonance caused by the non-linearity of electro-magnetic potential transformer (PT) may be classified into two categories according to the dissimilarity of physical model of the resonance network: One category is in neutral isolated kv distribution network, which is caused by a set of or several sets of PT together with the capacitance of network to ground, another category is on the HV idle bus bar of kV substations, i.e. the series resonance caused by a set of or several sets of equalizing capacitances of circuit breakers together with the bus bar PT (usually one set), the resonance weave is only restricted within the range of the bus bar in substation. This paper refers to the former as network resonance, and the latter as substation resonance. Essentially speaking, Ferro-resonance is caused by the non-linearity of PT, while the casualty is also of itself. If, however, explosion is resulted from damage of PT, then it is liable to endanger the safety of adjacent equipments, even induce more series aftereffect of system fault. Therefore, suppression of Ferro-resonance has become one of the tasks for global prevalent attention. There are manifold measures for suppressing Ferro-resonance, among which the damping method is the simplest one, traditional practice is to connect a resistance across two terminals of the secondary winding of every PT or across the open delta which is composed of the tertiary winding of 3 single-phase PTs. But practically, it is not allowed to do so due to restriction of thermal capacity of PT. N. Janssens etc [1] put forward an improved scheme, i.e. to connect a damping resistance (several kw) and a saturable reactor in series into the secondary side of PT to suppress the substation resonance; C. Stuckens etc also recommended to connect a resistance and a saturable reactor in series into the open delta of PT to suppress the network resonance. The damping performance of the resistance is thus improved after it is connected in series with a saturable reactor, under normal operation condition, non-saturated reactor has high-resistance characteristic, which may mitigate the influence of damping resistance against the accuracy of PT measurements. When resonance appears, the reactor reaches saturation prior to the PT so as to get the resonance suppressed. This kind of damping method with series reactor expands the applicable range of resistance damping method. But the practical effect of this improvement is still limited, unable to solve all problems, especially for the network of smaller critical damping resistance, it is still helpless. That is because the non-linearity of the reactor is impossible to reach idealization.(i.e. the reactance tends to infinite under normal operation, and zero when resonance appears). This paper introduces the strong damping method on the basis of power electronic technique. The basic practice is: to install a thyristor (SCR) of switching characteristic and a low resistance in series across the terminals of PT open delta. Normally, the SCR is in blocking state, isolates the damping circuit from PT to avoid any influence of low resistance to the accuracy of measurements. When resonance occurs, SCR is triggered to be conducting, and then the low resistance plays its strong damping role. This controllable strong damping method was put forward by one of the authors first in 1979 and successfully applied in the development of network resonance-suppression device (WX type) and DLP_TJLai_A1 Session 1 Paper No

2 substation resonance- suppression device (BX type, see Fig. 1) thereafter. Over the past two decades, the above devices are increasingly perfected through many times of improvements. The thought-way for every improvement is based on the information from on-site tests. We have carried out more than 80 times of tests in network and more than 30 tests in substations. At present there are several thousand of WX type and several hundred of BX devices under operation in power system. BX see Fig 2, Fig 3 respectively. We have accumulated many experiences from field test and operation, what caused us pleasantly surprised is that We accidentally found: when resonance appears on the idle bus bar of 110/220 kv Substation, the voltage of resonant phase is not an isolated one, it produces serious effect on the other phases through capacitance coupling between phases. If two phases or three phases appear resonance at the same time, then the mutual influence will be more intricate and complicate. Over many years, people always attribute the substation ferro-resonance to physical model of independent single-phase resonance, R. Gandre etc [1] and [3]. The new discovery of electric field coupling between phases challenges the traditional physical model of single-phase resonance. This paper shows in detail such influence by the oscillogram recorded on site and presents a new sort of single-phase resonance model, which takes the influence between phases into account. Fig.2 Block diagram of hardware circuit of WX type of resonance-suppression device Fig.3 Block diagram of hardware of BX type device 2.1 Connection-in of damping circuit Fig 1 BX type of s/s resonance suppressing device. 2 Design consideration of hardware circuit of the device For block diagram of hardware circuit of WX and Damping circuit of both devices is connected to the two terminals of open delta of PT. For BX type of device two additional auxiliary wires are needed, as shown in Fig.3 by dotted lines, so as to meet the requirement of split-phase damping. This connection method may obviously increase the damping effect as compared with the traditional one that is connected to the PT secondary side (three times increase). DLP_TJLai_A1 Session 1 Paper No

3 2.2 Selection of damping resistor The connecting cable between tertiary winding of PT and this device acts concurrently as a damping resistor, therefore no more dedicated damping resistor will be installed in the device. Under certain special occasion, if the resistance of the above connecting cable still cannot meet the requirement of the resistance of network or bus bar whereabouts, then resistance-reduction measure should be taken to the cable. artificial single-phase ground fault. Altogether 23 times of excitation operations were conducted, among which 16 operations excited stable 1/2 sub-harmonic resonance. The probability of occurrence of resonance is 70%. Whenever resonance occurs, the device always can operate to eliminate them. Fig. 4 reflexes the different results under different value of damping resistances. 3 Test of WX type of device in network In 1981, more than 50 times of resonance-suppression tests were carried out in 10kV networks of Tieling old city, Liaoyang Lighthouse etc. Considering that the higher is the network voltage level, stricter is the requirement of damping resistance value; therefore, in order to strictly check the function of the device, in 1982, the 66kV Tieling network was once more selected for performing another 20 resonance-suppression tests. The purpose of these tests is: understanding the probability of resonance generated by excitation. Checking the validity of WX type device for resonance suppression. Seeking for the critical resistance of network. (b) (a) The neutral point of network is originally grounded through arc-suppression coil, in order to carry out Ferro- resonance test, all arc-suppression coils were taken out of service temporally. The tested networks have the following main parameters: capacitive current to ground 28A altogether 6 sets of PTs Ratio of capacitance to ground to reactance of PT X CO /X L =0.023 In order to facilitate estimation of critical resistance of network, only one set of PT open delta was installed with WX device, other 5 sets of PTs were without any damping means. The mode of resonance excitation: using pre-determined circuit breaker to switch off the (c) Fig 4 Different effects result in suppressing 1/2 sub- harmonic resonance with different damping resistance (a) R Z =0.197Ω, resonance cleared by damping 0.1s after connection of R Z (b) Rz=0.247Ω, resonance cleared by damping 0.26s after connection of Rz (c) Rz=0.377Ω,damping invalid Where, Ua, Ub, Uc are secondary phase voltage of PT; U is open delta voltage, I is the current when SCR is conducting. From Fig.4 it can also estimate the critical resistance DLP_TJLai_A1 Session 1 Paper No

4 of that network is about 0.247, if the device is installed indoor, then the connecting cable between the device and PT open delta may be too long, then it is very difficult to meet the requirement of resistance value. If,however, the damping means are adopted at 3 sets or 6 sets of PTs in the same time, then the critical damping resistance sensed by each set of PT can increase to 0.741Ω or 1.48Ω. 4 Test of BX device in substation Tests of resonance-suppression were carried out in three 220kV substations of Diaobingshan, Liaoyang and Ganxi, and four 110kV substations of Ganxi, Fuding, Lufeng, Tangshan one after another. The main purpose of test in early stage was to check whether the device could reliably suppress the resonance of a specified substation. I.e. to check the pertinence and validity of the device; the purpose of the test in recent years since 1993 is to lay particular attention to check whether the device, after many years improvement, has the function of suppressing resonance of all substations. I.e. to check its universal validity. There are many unexpected problems happened in field tests, which forced us to modify the technical scheme for resonance-suppression, and the hardware circuit and software are correspondingly modified along with. The tested circuit breakers, whether they are of 110kV or 220 kv, are all with double breaks; unit capacitance of the equipotential capacitor is about 1800 PF. Except in Liaoyang substation the bus bar is connected to power source through two sets of circuit breakers, the other are all through one set of circuit breaker. Resonance excitation mode: to switch off the last circuit breaker, which is connected with the idle bus bar. 4.1 Test of 220kV substation Liaoyang substation: On Nov. 15, 1993, 10 times of opening operation were conducted, among which 6 times of operations initiated stable resonance. Mainly basic frequency resonance, 1/3 sub-harmonic resonance second. And these two resonance weaves often present interactively. Fig 5 is the oscillogram recorded during opening of the circuit breaker. It shows: After opening, Both phase A and B present 1/3 sub-harmonic resonance at the same time, about 2 cycles later, transit to basic frequency resonance simultaneously. Under strong damping of the device, resonance weaves of phase B is quickly eliminated, that of phase A is eliminated gradually after a longer time too. Fig 5 Two-phase basic frequency resonance, Liaoyang 500kV substation Mengda power plant substation: On April 18, 2002, altogether 2 times of opening operation were conducted, stable basic frequency resonance was initiated on phase A and B one after another during second time of opening, and vanished successively under strong damping of the device. (See fig 6) Fig. 6 Suppressing phase A and B basic frequency resonance, Mengda power plant substation 4.2 Typical oscillogram of 110kV substation test Fig 7 Successively suppressing phase A and C basic frequency resonance, Ganxi Shangzhuang substation, June 1,2000 Fig. 8 The course of excitation of 3-phase basic DLP_TJLai_A1 Session 1 Paper No

5 frequency resonance, Ganxi Shangzhuang substation, Dec. 3, 2001 Fig.9 Fig. 9 The course of 3-phase basic frequency resonance Being damped, Lufeng Nantang substation, Jan. 6, 2002/12/29 Fig 10 the course of excitation and damping of 3-phase 1/3 sub-harmonic resonance, and variation of each harmonic content during the course. Tangshan power plant central substation. Nov. 28, Fig Discussion In the resonance test of substation, some problems were found worth discussing: Fig. 7 Fig. 8 1) Viewing the physical model of single-phase resonance from the influence between phases. It can be seen from Fig.7, after 0s opening 0f circuit breaker, basic frequency resonance presented on phases A, C successively. At the time of 0.27s, the device started up to implement damping over phase A. Along with attenuation of resonance voltage on phase A, the peak value of resonance voltage on phase C dropped from 100V to 68.9V, voltage weave of phase B changed along with. The phase voltage, after reduction of amplitude, conversely influences phase A, the resonance of which had just been damped, and the adjacent phase B. At the time of 0.5s, phase B was damped, the above influence became unstable and variable along with, up to 0.93s the resonance was eliminated. It can be seen from this course, the two resonant phases A, C generated strong mutual influence through coupling of spatial capacitance, and they also influenced the non-resonant phase B at the same time. So that we have reason to doubt the preciseness of the physical model for traditional substation single-phase resonance (Fig.11). We consider that the single-resonance model should be as shown in Fig.12 after considering the influence between phases. DLP_TJLai_A1 Session 1 Paper No

6 power system caused by PT. 2) Due to that the device makes use of the external connecting cable as damping element concurrently, the damping strength is enhanced to the greatest limit. In practical application, it fully uses this characteristic: Fig.11 Traditional physical model of single-phase resonance For the 10kV network with higher critical damping resistance, generally it only needs to install WX device on one or two PTs, to save the expenses. C A For 35kV network, generally one piece of WX device may be installed on every 2 or 3 sets of PTs. C A C B Fig. 12 Physical model of Single-phase resonance after considering influence between phases 2) Under general condition, the amplitude of resonance has something to do with the basic frequency, the higher the frequency, the higher the amplitude and vice versa, For example, the amplitude of basic frequency resonance weave is obviously higher than that of power supply voltage, the amplitude of 1/3 harmonic resonance weave is obviously lower than that of power supply voltage, This basic rule has been proved by many tests. But there is exception anyhow, such as shown in Fig.7, the amplitude of phase C is less than 0.8 P U during the interval from 0.27s to 0.5s, obviously lower than system phase voltage, for this further research is needed. 5 Conclusion 1) The strong damping method on the base of power-electronic technique has the unique effect of suppressing ferro-resonance. In principle, it can suppress all of the ferro-resonance phenomenon in On 220kV idle bus bar, if there are multi-sets of circuit breakers equipped with equalizing capacitance connected to the bus bar, then the critical damping resistance may be about 1 Ω. This moment, if the resistance value of the external connecting cable is too high, it will be likely to influence the damping effect. Substation resonance presents as a phenomenon of single-phase resonance. Therefore, the traditional physical model of single-phase resonance is the core. But only when the mutual influence between phases is taken into consideration, will the model further accord with the practical condition. That is to say, the electric field should be introduced into the model on the basis of electric circuit. References: 1 N.Janssens, V. Vanderstockt, H. Denoel, P.A.Monfils, 1990, Elimination of temporary over voltage due to ferroresonance of voltage transformers: design and testing of a damping system, CIGRE 1990 Paper C. Stuckens, P.A. Monfils, N.Janssens, TH. Van Craenenbroeck, D. Van Dommelen, 1997, Risk of ferroresonance in isolated neutral networks and remedies, CIRED 1997, Paper DLP_TJLai_A1 Session 1 Paper No

7 3 R.G.Andrei, P.E.B.R.Halley, P.E.: Voltage transformer ferroresonance from an energy transfer standpoint, IEEE 89WM PWRD 4 CIGRE 1990, Paper The origin, effect and estimation of transient overvoltage. 5 Gao Qiang, Lai Tianyu, Lai Qingbo, 1997, Study on features of ferromagnetic resonance of 110kV substation and its eliminating method, CIRED 1997, Paper DLP_TJLai_A1 Session 1 Paper No