On the Issue of Relay Protection of Synchronous Condensers

Yu. E. Nebolyubov
Submitted 1945 | SovietRxiv: ru-194501.08486 | Translated from Russian

Abstract Generated abstract

This article examines relay protection for synchronous condensers, arguing that schemes copied from generators and motors fail to account for condenser starting equipment and possible loss of synchronism. It proposes extending longitudinal differential protection to include the autotransformer starting device, and develops a special loss-of-synchronism protection based on detecting the alternating current induced in the excitation winding during asynchronous operation. The proposed relay arrangement is described with its operating sequence, intended delay to allow possible resynchronization, and measures to avoid false tripping during faults and starting. A calculation method for selecting relay settings is then outlined using simplified expressions derived from Park’s theory, and laboratory tests are reported to have shown clear operation at various slips.

Full Text

On the Issue of Relay Protection of Synchronous Compensators

Yu. E. Nebolyubov
Associate Professor, Candidate of Technical Sciences

This article presents some considerations concerning the development of relay protection for synchronous compensators.

At present, protection schemes for compensators are implemented in the same way as for generators and motors, and therefore certain features of the operation of these machines and of their switching schemes are not taken into account. Here one should keep in mind the presence of starting devices and the possibility of compensators falling out of synchronism.

In the journal literature of recent years, articles have appeared concerning changes in the protection of compensators against overcurrents, but the standard schemes of differential protection have not been discussed. Likewise, the question of creating special protection for compensators against loss of synchronism has not been considered at all.

Longitudinal differential protection with circulating currents is implemented according to the scheme in Fig. 1. In this case the protection zone includes one compensator without the starting device. The power of autotransformers in modern starting schemes reaches a considerable value and, for example, for a 7.5 MVA compensator it is 2800 kVA.

Fig. 1

Fig. 1

The development of phase-to-phase faults in the starting device may lead to a prolonged shutdown of the compensator.

It is quite clear that the zone of longitudinal differential protection must be changed with the calculation that the entire starting device be included in it. For this purpose it is necessary to move the current transformers 1, installed at the terminals (Fig. 1), placing them before the oil circuit breaker VM-1, and to connect the neutral points of the compensator and autotransformer

(Fig. 2). In the event of phase-to-phase short circuits in the autotransformer and in the connecting cables, the differential protection will instantaneously disconnect the compensator.

Similar schemes with the neutral points connected together may be encountered in synchronous motors, but in the literature there are no indications of their application to synchronous compensators.

Starting a compensator through an autotransformer is accompanied by the appearance of transient-condition currents in the stator winding. These currents considerably exceed the rated current of the machine. According to the materials on

Fig. 2

Fig. 2

the testing of a 7.5 MVA, 6 kV synchronous compensator of the Elektrosila plant, the peak surge of current, when the compensator is switched on at a difference between the network and machine voltages, is 1550 A, while the rated current of the machine is 722 A.

The starting currents will, of course, increase the unbalance current in the differential-protection circuit. It is extremely difficult to take exact account of the change in the unbalance current, but there is every reason to suppose that it will not reach the pickup current of the protection and will not cause a nonselective disconnection of the compensator. Therefore there is no point in coarsening the protection, and the pickup-current setting may be left unchanged \((i_{\text{pt}} = 1\text{–}2\ \text{A})\).

Turning to the question of relay protection of compensators against loss of synchronism, it should be noted that their falling out of synchronism and transition to a stable asynchronous mode is quite possible [5], despite the known possibility of self-synchronization (pulling into synchronism) under certain ratios of the system parameters.

The asynchronous mode of compensators is dangerous for a whole series of reasons. First, significant currents flow in the stator winding, in the damper circuits, and in the field winding, which may cause inadmissible—

heating of these windings and disturb the thermal regime of the machine. Secondly, owing to the varying values of the compensator power during the slip period of the asynchronous regime, dangerous mechanical pulsating forces arise. Thirdly, in the event of improper operation of the field-suppression automatic device or of a break in the excitation circuit, the voltage at the terminals of the excitation winding at large slips may reach a considerable value. Finally, fourthly, if the compensator power is large, then the swings arising at the moment when its normal operating regime is disturbed may lead to swings of nearby stations and to their falling out of synchronism.

All existing protections will not respond clearly to the appearance of an asynchronous regime. Undervoltage protection can operate when the asynchronous regime is caused by a sharp reduction in the system voltage, but not by damage to the machine’s excitation circuit.

In some cases the possibility is not excluded that overcurrent protection will operate. For this it is necessary that the current and voltage reach values sufficient to pick up the relay. However, the voltage at the compensator terminals in an asynchronous regime cannot always fall to the pickup setting of the undervoltage relay, and the protection will fail to operate. This may occur at small slips of a stable asynchronous regime, when the current in the stator winding only slightly exceeds the rated current of the machine.

Thus, all the considerations presented lead to the idea of creating a special protection for synchronous compensators against falling out of synchronism, which should disconnect from the network compensators that have entered a stable asynchronous regime. Such a protection scheme can be implemented on the current principle, using the induced alternating current that appears in the excitation winding in the case under consideration.

If the compensator is in synchronism, then only direct current flows in the excitation winding. With the occurrence of an asynchronous regime, an alternating current appears in this winding, having the slip frequency. This alternating current can be filtered out by means of a current transformer whose primary winding is connected in the circuit of the excitation winding, and to whose secondary winding an electromagnetic current relay is connected (Fig. 3).

Fig. 3

Fig. 3

The diagram shows a special three-winding current transformer [2], eliminating strong saturation of the core and the possibility of incorrect relay operation during processes of regulating the compensator excitation. Windings a and b (Fig. 3) have the same number of turns and are connected so that, under normal conditions and with any changes in voltage, there is no flux in the core.

Consequently, a total resistance \(z\) is connected in series with winding \(b\), giving the circuit of this winding the same time constant as that of the compensator excitation winding (circuit \(a\)). Under asynchronous operation, the electromotive force induced in the excitation winding produces currents in windings \(a\) and \(b\), the ampere-turns of which add. A flux will arise in the core, and the relay connected to winding \(c\) will operate.

In addition, the protection circuit includes three auxiliary relays (2, 3, 5) and a time relay (4). Auxiliary relay (2) has three pairs of contacts: one pair of contacts operates to open, and the other two pairs operate to close the direct-current control circuit. Auxiliary relays (3 and 5) each have two normally open contacts. The time relay (4) is provided to remove the direct current after a time interval of 5–10 seconds and ensures that the circuit returns to its initial position.

The purpose of the auxiliary relays becomes clear from an analysis of the operation of the protection. If the nature of the change in the alternating current in the secondary winding of the current transformer is represented in accordance with Fig. 4, then the current relay (1) will operate when the current reaches the values marked by the ordinates \(aa'\) and \(cc'\), corresponding to the pickup current. The period of the alternating-current curve is determined by the magnitude of the slip and, obviously, the greater the slip, the more slowly the current rises and falls during one quarter-period, and the more slowly relay (1) closes and opens its contacts.

Fig. 4

Fig. 4

The protection circuit operates in the following sequence. If the current in relay (1) reaches the value \(aa'\), equal to the pickup-current setting (Fig. 4), the contact of this relay will close and the direct current will be supplied to auxiliary relays (2), (3), and to the time relay (4). Relays (2), (3) operate instantaneously, and relay (3), which has a self-holding circuit, is reliably held in the energized position.

When the alternating current decreases to the value determined by the ordinate \(bb'\), corresponding to the return current, relay (1) opens its contact, and auxiliary relay (2) returns to its initial position. At the same time the middle contact of relay (2) opens, but the circuit for energizing auxiliary relay (5) has already been prepared for further operation of the protection circuit.

With further variation, the alternating current again reaches the value of the pickup current (ordinate \(cc'\)), and relay (1) again closes its contact. Following this, relay (2) and relay (5), which has self-holding, are energized.

At the same moment, the direct current enters the trip coil of the main oil circuit breaker, and the synchronous compensator is disconnected from the network.

Thus, the protection circuit under consideration ensures disconnection from the network of a compensator that has lost synchronism and makes it possible for the machine to pull into synchronism. Indeed, when the compensator is able to pull into synchronism during the first cycle of asynchronous operation, the protection circuit provides the time delay necessary for this and does not produce instantaneous tripping.

The time delay is determined by the slip frequency and by the intrinsic operating time of relays (2), (3), and (5), while the duration of the first cycle of the asynchronous mode is often determined by a very small time and will be less than the protection time delay.

With an insignificant deviation of the machine speed from synchronous speed, the protection likewise must not operate, because the current in relay (1) will not reach the relay pickup current value, which is selected from the condition of a steady-state asynchronous mode. Thus the possibility of false tripping of compensators from the network is excluded.

The basic requirements for all protections—simplicity of the scheme, reliability, speed of operation, sensitivity, and selectivity of operation—are fully satisfied by the considered scheme for protecting compensators against falling out of synchronism.

The protection scheme includes one starting element—an ordinary maximum-current relay of type ET-60 or RE-100, three intermediate relays of type EP-240 or RP-2, and a time relay EV-182.

All the apparatus and the DC control circuits are distinguished by their simplicity.

The simplicity of the scheme ensures its reliability, since the fewer the elements, the more reliable the scheme.

With proper selection of the pickup current, the protection must have sufficient sensitivity and will respond to a comparatively small induced current of the asynchronous mode.

The selectivity of the protection operation is characterized by the fact that it responds only to the asynchronous mode of the compensators.

In the case of symmetrical and asymmetrical short circuits in the network, induced alternating currents associated with the course of the fault process will also appear in the excitation winding. However, the compensator will not be disconnected from the network, because in this case the alternating component of the current in the excitation circuit will vary at a frequency of 50–100 Hz, and the moving system of current relay (1) will not be able to close and open the relay contacts in accordance with the current variation.

The loss-of-synchronism protection must be removed from service during asynchronous starting of compensators in order to exclude the possibility of unnecessary tripping of the machine while it is accelerating to synchronous speed. For this purpose the positive pole of the control current should be passed through the manual-control contactor (K) or through the block washers of the main oil circuit breaker.

The considered compensator protection scheme was tested under laboratory conditions and, at various slips, showed clear operation.

Calculation of the protection of compensators against loss of synchronism

The protection current relay (Fig. 3) must respond to the alternating component of the current in the excitation winding under a steady-state asynchronous mode. This alternating current \((I_{fd})\) corresponds to the average slip \((s)\) of the established asynchronous mode, determined with account of the specific operating conditions of the compensator in the system [1]. With some allowance, the slip \((s)\) for possible cases of occurrence of an asynchronous mode of compensators may be taken equal to 0.05–0.1, and these values may be used in the protection calculation.

Determination of the current \(I_{fd}\) is carried out by a method based on Park’s theory (see the author’s article in the present collection, “Construction of Asynchronous Characteristics of Compensators”).

To reduce the computations of the current \(I_{fd}\), a number of simplifications may be introduced into the calculation and generalized average values of the reactive—

...of capacitances and resistances entering into Park’s formulas. As a result of simplifications, which amount to neglecting insignificant terms, we obtain.

1) The operational reactances in complex form:

\[ x_d(js)=x_d-j0.2, \]

\[ x_q(js)=x_q-j0.1, \]

where \(s\) is the slip of the established asynchronous operating condition of the compensator, adopted in the calculation; \(x_d\) and \(x_q\) are the synchronous reactances of the compensator along the direct and quadrature axes.

2) The direct-axis component of the stator current in the asynchronous condition

\[ i_d=\frac{jx_q(js)-\dfrac{0.01}{1-s}}{2x_d(js)}. \]

3) The induced alternating current in the field winding

\[ I^{*}_{fd}=\frac{(-s^2 0.2+js0.06)i_d}{-s^2 0.2+js0.1}. \]

The current \(I^{*}_{fd}\) is obtained here in relative units and is numerically equal to the excitation current \(I_f\), expressed in fractions of the excitation current \(I_{fk}\), which gives the normal current in the stator under a three-phase short circuit,

\[ I^{*}_{fd}=\frac{I_f}{I_{fk}}. \]

In turn, the following relation holds:

\[ \frac{I_{fk}}{I_{fo}}=x_{afd}, \]

where \(I_{fo}\) is the excitation current (in amperes) at no load and normal voltage of the compensator (according to catalog data); \(x_{afd}=(0.6—0.8)\) is the mutual reactance of the stator winding and the field winding along the direct axis. Then the current \(I_{fd}\), in amperes, is

\[ I_{fd}=I_f=I_{fk}\cdot I^{*}_{fd}. \]

The effective value of the total current in the field winding under the asynchronous operating condition of the compensator, taking into account the constant component, is

\[ I_{\mathrm{ef}}=\sqrt{\left(\frac{I_{fd}}{\sqrt{2}}\right)^2+I^2_{fn}}. \]

The current transformer of the protection is selected according to the effective value of the current in the excitation circuit \((I_{\mathrm{ef}})\).

The pickup current of the electromagnetic relay is calculated from the known relation

\[ i_{p\tau}=k_n\frac{I_{fd}}{n_\tau}, \]

where \(k_n=(0.6—0.7)\) is the reliability coefficient; \(n_\tau\) is the transformation ratio of the current transformer. The error of the current transformer when operating in the region of low slip frequencies increases, but this increase in error is compensated by introducing the reliability coefficient.

REFERENCES

  1. Lebedev S. A. and Zhdanov P. S.—“Stability of Synchronous Compensators” (VEI Bulletin No. 5, 1941).

  2. Pines N. M.—“Protection of Generators against Falling out of Synchronism” (Proceedings of KII No. 2, 1937).

Submission history

On the Issue of Relay Protection of Synchronous Condensers