ANALYSIS OF THYRISTOR SOURCES OF REACTIVE POWER

V. N. MISHIN, D. F. ZENKOV, V. V. LEONOV

(Presented by the scientific-technical seminar of the Research Institute of Nuclear Power Engineering)

Thyristor sources of reactive power (SRP), unlike unregulated compensating devices, make it possible automatically to regulate and stabilize the value of the power factor, smoothly regulate the magnitude of the generated reactive power, maintain the stability of the supply-network voltage in both static and dynamic regimes; they are not critical with respect to the harmonic components of the network voltage and current and have a long service life.

The basic element determining the classification principle of SRP should be considered the working element of these devices, which serves for the accumulation and exchange of energy with the network. Continuous variation of the reactive power of an SRP containing, as its working element, a bank of capacitors requires a smooth change in voltage, which is achieved by introducing into the circuit an appropriate control device [1].

Another method of regulating reactive power is the use of an unregulated bank of static capacitors and a controlled reactor connected in parallel or in series with it. One such SRP was proposed at the All-Union Electrotechnical Institute named after Lenin [2]. An important advantage of the VEI circuit is its low inertia. Its disadvantages include the presence of higher harmonics in the current curve, which entails the need to protect capacitor banks, a negative regulating effect on voltage, and a short service life.

Instead of the three reactors required in the VEI version of the SRP, one reactor may be used, connected in the direct-current circuit of a controlled valve converter connected to the network in parallel with the capacitor bank. This version has all the disadvantages of the preceding one, with a worse harmonic composition of the current curve and somewhat lower speed of response. However, if the converter is supplemented with an artificial-commutation unit, the capacitor bank can be dispensed with, since, when operating with leading control angles, the converter generates reactive power into the network.

One of the possible SRP circuits based on an AC-to-DC valve converter with artificial commutation was developed at the Moscow Power Engineering Institute [3]. The disadvantages of the SRP according to the MEI circuit include the large ratios of internal transient—

voltages, the presence of a transformer with two sets of secondary windings and two reactors, the large installed power of the equipment, and the limited control range.

The RPS proposed at the Gorky Polytechnic Institute (GPI) [4] is based on a bridge inverter with clipping diodes. In order to reduce switching overvoltages in the network, a protection unit against switching overvoltages is used; it consists of a bank of electrolytic capacitors connected to the network through a diode bridge, and an inverter returning to the network, through a step-up autotransformer, the energy received by the filter capacitors at each commutation. Along with such advantages as fast response and simplicity of control, the circuit also has a number of shortcomings: the presence of clipping diodes and the power losses in them; the reverse voltages on the clipping diodes are equal to twice the value of the amplitude network voltage; the control range is limited.

Fig. 1

Figure 1 shows the circuit of an RPS developed at the Scientific Research Institute of AEM under the Tomsk Polytechnic Institute jointly with GPI and the Scientific Research Institute of the “Elektrovypryamitel” plant in Saransk. The RPS consists of a controlled three-phase bridge rectifier loaded by a choke \(L\), two forced-commutation units (FCU), a filter, and an inverter. Each FCU circuit switches off the valves of either the cathode or the anode group. To limit switching overvoltages, a device is used consisting of the rectifier bridge \(D_1\)—\(D_6\), a bank of filter capacitors \(C_{\phi}\), and an inverter that recovers the energy of the switching overvoltages to the network.

Let us consider the operation of the proposed RPS. Suppose valves \(T_1\) and \(T_6\) are conducting current. To commutate the current from the circuit of valve \(T_1\) into the circuit of \(T_3\), the commutating thyristor \(T_{kk}\) is switched on. The capacitor \(C_{kk}\) begins to recharge resonantly through the inductances \(L_1\) and \(L_2\). After a quarter-period of the natural oscillations of the circuit, the self-induction emf across the inductance \(L_1\) changes sign and is applied to valve \(T_1\) in the reverse direction, and under its action the load current transfers from the circuit of the operating

of the next valve into the parallel circuit \(D_1—T_{\mathrm{kk}}—L_1—D_7\). The reverse recharge of \(C_{\mathrm{kk}}\) is carried out along the circuit \(C_{\mathrm{kk}}—L_2—L_1—R_{\mathrm{kk}}—D_{\mathrm{kk}}—C_{\mathrm{kk}}\); in this case the load current passes from the circuit \(T_{\mathrm{kk}}—L_1\) into the circuit \(C_{\mathrm{kk}}—L_2\). The commutating thyristor is given the possibility of restoring its blocking properties owing to the applied reverse voltage arising on the chain \(D_{\mathrm{kk}}—R_{\mathrm{kk}}\) from the flow of the reverse-recharge current of \(C_{\mathrm{kk}}\). After \(C_{\mathrm{kk}}\) has been recharged to the line voltage between the conducting and the next phases, valve \(T_3\) is switched on, and the load current passes into the next phase. When the voltage on \(C_{\mathrm{kk}}\) becomes equal to the voltage on \(C_\phi\), the load current flowing through the commutating unit begins to flow through \(C_\phi\), charging it. In view of the large capacitance of \(C_\phi\), the current-commutation process proceeds rather smoothly, without causing large overvoltages in the network.

The advantages of an RPS (Fig. 1) are the small overvoltages and voltage distortions in the ac network, and the insignificant increase in the rating class of the valves used in comparison with an analogous converter but without artificial commutation, owing to the divider of the commutating voltage on inductances \(L_1\) and \(L_2\). For a comparative estimate of the installed power of the electrical equipment of various RPSs, the calculated data are summarized in Table 1.

Table 1

The RPS variant according to the MEI circuit, because of the presence of a transformer and two reactors, is obviously inferior to the other variants; therefore the table gives data for three variants.

The installed power of the thyristors and diodes was determined as the product of the number of thyristors in the RPS circuit by the average value of the current flowing through them and by the maximum value of the reverse voltage. The installed power of the thyristors and diodes in the artificial-commutation units and in the RPS inverter was taken into account with a coefficient of 1.1. The installed power of the reactors was determined on the basis of [5]. In this case the required value of the reactor inductance was determined according to [6], proceeding from a regulation range of \(1:10\), and it was assumed that in the regulated mode the RPS is at the boundary of discontinuous current. The installed power of the commutating capacitors in the circuits was determined as the product of the capacitance by the angular frequency of the ac network and by the square of the maximum voltage on the capacitors, divided by two.

It follows from the table that the RPS according to the VEI circuit is inferior because it requires capacitors rated for the full power. The installed powers of the electrical equipment of the RPS according to the GPI circuit and Fig. 1 are approximately the same, but the RPS based on an inverter with clipping diodes has greater specific losses because of the presence of additional diodes in the power circuit.

Therefore, the reactive-power source shown in Fig. 1 appears to be the most preferable.

Thus, the approximate analysis presented above of the installed capacity of power equipment for thyristor reactive-power sources shows the advisability of implementing the reactive-power source on the basis of transformerless AC-to-DC converters with forced commutation. For the final substantiation of one variant or another, detailed technical-and-economic calculations are required, taking into account all the equipment and the losses in the elements of the circuit.

REFERENCES

1. L. A. Zhukov, I. I. Kartashov, E. A. Pankratova, Yu. P. Ryzhov. Methods of smooth regulation of reactive power in electrical networks. Author’s certificate No. 230959, 1968. Bulletin of Inventions, No. 35, 1968.

2. V. V. Khudyakov, V. A. Chvanov. A controllable static source of reactive power. Elektrichestvo, 1969, No. 1.

3. Yu. P. Ryzhov. Some characteristics of a controllable static source of reactive power operating with forced commutation of valves. Energetika i transport, 1968, No. 1.

4. B. Ya. Gumanovskii, V. G. Titov, O. I. Voropaev, A. A. Poskrobko. A thyristor source of reactive power. Elektrichestvo, 1970, No. 11.

5. E. A. Man’kin. Calculation of transformers with a steel magnetic core and gaps. Elektrichestvo, 1959, No. 7.

6. A. I. Zaitsev, V. N. Mishin, D. F. Zenkov, L. V. Aleksandrova. On the calculation of thyristor reactive-power sources. Reports to the VII Scientific and Technical Conference. Static converters in automation and electric drive. Tomsk, 1971.