ON ONE FEATURE OF DIFFERENTIAL PROTECTION OF GENERATORS WITH EQUALIZATION OF THE NO-LOAD EMF OF CURRENT TRANSFORMERS

L. V. Baginskii, student V. G. Koinova, L. G. Nekrasova

(Presented by the scientific-technical seminar of the Department “Electric Power Stations, Networks, and Systems”)

Experiments, the results of which are given in Fig. 5 [1], show that the unbalance current of the protection under consideration depends very little on the primary current. Therefore, when one and the same pair of current transformers is used for the protection of generators of different ratings, the maximum values of the unbalance currents will be practically identical; that is, for equal values of the reliability coefficients, the absolute values of the primary pickup currents will also be equal for any ratings of the protected generators. This means that the value of the primary pickup current, expressed in fractions of the rated current, will be the smaller, the greater the rating of the protected generator.

Moreover, as the rating of the protected generator increases, one may also expect an increase in the sensitivity coefficient owing to an increase in short-circuit currents.

Fig. 1.

To verify these assumptions, corresponding experiments were carried out in which a pair of current transformers was used with numbers of turns of the primary and secondary windings \(w_1 = w_2 = 200\), for—

with a mean magnetic path length of the core \(l_{\mathrm{av}} = 41.5\ \mathrm{cm}\) and a toroidal-core cross section \(q = 3\ \mathrm{cm}^2\). The no-load characteristics of these transformers are presented in Fig. 1, from which it is evident that the difference in the characteristics was very considerable: 23% at low inductions, 14% at an induction of about \(11\ \mathrm{kG}\), and 10.7% at high inductions of about \(15\ \mathrm{kG}\). Such a large difference in the characteristics in the region of high inductions is not typical for current transformers made of steel of the same grade. Therefore it may be assumed that at least one core was made from strips of steel of different grades. This substantially worsened the operating conditions of the scheme.

Fig. 2.

Figure 2 gives the dependences of the unbalance current \(I_2\) on the field strengths in the cores, calculated as \(H_{\mathrm{eff}} = I_1 w_1 : l_{\mathrm{av}}\), for different values of the relay-circuit resistances. Curve 1 was taken with active and inductive resistances \(r_2 = 2.12\ \Omega\) and \(\omega L_2 = 0.7\ \Omega\), and curve 2 with \(r_2 = 3.49\ \Omega\) and \(\omega L_2 = 2.8\ \Omega\), the inductive resistance in both cases being formed by the coils of relay ET521/0.6 (parallel and series connection). From Fig. 2 it follows that the current \(I_2\) is practically unchanged beginning with values \(H_{\mathrm{eff}} = 50\ \mathrm{A\!-\!turns/cm}\). Since high-power generators are equipped with current transformers having a mean core-line length of the order \(l_{\mathrm{av}} = (75\text{--}100)\ \mathrm{cm}\), it may be considered that the short-circuit currents of a generator of any rating will produce field strengths in the cores of the protection scheme under consideration of \(H_{\mathrm{eff}} > 50\ \mathrm{A\!-\!turns/cm}\). Therefore the unbalance current of the scheme is practically independent of the rating of the generator being protected.

Figure 3 shows the dependences of the primary operating current \(I_{\mathrm{op}}\), expressed in fractions of the rated current of the protected generator \(I_{\mathrm{N}}\), and of the sensitivity coefficient \(k_s\) on \(I_{\mathrm{N}}\). These dependences were determined for the same relay-circuit resistances as the curves in Fig. 2. The relay setting was determined as \(I_{\mathrm{set}} = k_{\mathrm{r}} \cdot I_2\), where the reliability coefficient \(k_{\mathrm{r}} = 1.5\), owing to the independence of \(I_2\) from \(I_{\mathrm{N}}\) under the conditions considered. The value \(k_s\) was determined from the values of the current in the relay for two-pole short circuits at the terminals of the protected generators. Hydrogenerators without damper windings were assumed as the protected generators. Therefore the multiple of the short-circuit current was taken as \(2.4 I_{\mathrm{N}}\). The protected hydrogenerator

was considered to be operating in isolation. The currents in the relay circuit were measured with the aid of carefully calibrated relays of type ET521.

Fig. 3.

The results of these experiments show that the sensitivity of the protection scheme under consideration increases substantially as the power of the protected generator is increased. For example, with an increase in the rated current of the generator from 2000 to 6000 A, the current \(I_{1cp}\) decreased by more than a factor of three, with a noticeable increase in the sensitivity coefficients.

Conclusions

1. The effectiveness of applying differential protection with balancing of the no-load e.m.f. of the current transformers increases with an increase in the power of the protected generator.

2. The scheme under consideration provides high sensitivity even with an unfavorable ratio of the characteristics of the current transformers \((I_{1cp} < 0.5 I_{ng},\ K_4 > 3.75)\).

3. In this scheme it is rational to use the ET521/0.6 relay, since the resistance of its windings limits the unbalance current sufficiently well. At the same time, in order to achieve the reliability-coefficient values required for selective operation, it may become necessary to move the relay pointer to the right beyond the limits of the setting scale.

Literature

1. Baginskii L. V. On the use of current transformers with a small volume of steel for differential protection of generators, Proceedings of the Tomsk Polytechnic Institute, vol. 94, 1958.