RECONSTRUCTION OF AN RVS-220 VALVE-TYPE SURGE ARRESTER INSTALLATION

I. I. KALYATSKII, P. G. KOVCHAVTSEV, N. M. TORBIN

(Presented by the scientific seminar of the Research Institute of High Voltages and the Department of High-Voltage Engineering)

The existing standard design of RVS-110 and RVS-220 arresters for networks with a grounded neutral provides for the installation of individual arresters of the RVS-30 type in a single column.

As the operating practice of RVS-110 and RVS-220 arresters has shown, in the Kuzbassenergo system there have been accidents at substations as a result of the failure of the indicated arresters. Thus, in November 1960, at one substation, during sharp gusts of wind, the three upper elements of RVS-220 arresters fell (the elements located above the guy wires).

A similar case of the fall of three upper RVS-220 elements occurred in June 1961 at one of the stations of the system during hurricane-force wind. The cause of the accidents is breakage of the porcelain of the arrester housings because of its insufficient mechanical strength and the appearance, during operation, of annular cracks in the porcelain at the reinforcement or beneath the reinforcement of the flanges. Operating experience with reconstructed RVS-110 arresters in the Kuzbassenergo system over a period of four years showed the absence of arrester falls, despite the fact that defective elements with annular cracks at the flanges are found annually [1].

In Kuzbassenergo, new methods have been proposed for installing RVS-220 arresters in order to increase mechanical strength.

1. Of the eight RVS-30 elements, where the first two elements from the ground are installed in a column on their own base, the next four elements are suspended in a spiral between three columns made up of six KO-400 support insulators, and the following two elements are installed on the upper frame of the three support columns.

2. Of the eight RVS-30 elements, where the first two elements from the ground are installed in a column on their own base, four elements are suspended in a zigzag pattern between two support columns made up of six KO-400 insulators, and the following two elements are installed on the upper frame of the support columns (Fig. 1).

3. The eight RVS-30 elements are separated into pairs; in this case the first two arresters are fastened on their own base on the foundation, the next two on three KO-400 support insulators, the fifth and sixth elements on three KO-400 insulators, and the seventh and eighth on four KO-400 insulators (Fig. 2).

The comparative dimensions of the complete structure for the variants considered are presented in Table 1.

Fig. 1. RVS-220 arrester in an arrangement on two supporting columns made of KO-400 insulators (variant 2).

Fig. 2. RVS-220 arrester in an arrangement of 4 columns (variant 3).

Calculation of the mechanical strength of the arrester installation structure according to the second variant (on two columns of 6 KO-400 insulator units each) showed that the moment of forces produced by a wind with a velocity

Table 1

of 40 m/sec and directed perpendicular to the plane of the arrester installation is 150.3 kgm, while the permissible moment for KO-400 insulators is 300 kgm. Thus, the safety factor is

\[ K_z=\frac{300}{150.3}=2. \]

Since, in the proposed variants of the modified RVS-220 arrester structures, the partial capacitances of the elements of the structures differ from the capacitances of the elements in the factory design, in the high-voltage laboratory of the Tomsk Polytechnic Institute investigations were carried out of the electrical characteristics of the modified RVS-220 arrester structures [3] according to the following program:

1. Determination of the volt-second characteristics of the arrester under the action of a standard impulse wave of 1.5/40 μsec.

2. Determination of the arrester sparkover voltage under the action of an impulse wave with a front length from 3 to 15 μsec.

3. Determination of the arrester sparkover voltage under simultaneous action of impulse voltage and power-frequency voltage of nominal phase value and maximum permissible value.

4. Determination of the voltage distribution over the arrester elements at nominal phase voltage and maximum permissible voltage, 50 Hz.

5. Testing of the arrester with the maximum permissible voltage, 50 Hz, for 2 minutes.

6. Investigation of the influence of the dimensions of the shielding rings on the impulse characteristics of the arrester.

The indicated types of tests were performed for the RVS-220 arrester in the factory installation (without spacer guy wires) and in the designs proposed by Kuzbassenergo—variants 2 and 3.

Variant 1 was rejected because of its more complex structural design in comparison with the other variants, despite its greater mechanical stability.

Investigation of the impulse characteristics of the arresters was carried out with the aid of a 3-million-volt impulse-voltage generator with an energy-storage capacitance of 7200 pF.

Voltage measurement was performed with the aid of a capacitive voltage divider and an OK-19M oscillograph, having additional time-sweep ranges up to 100 μsec.

The voltage distribution over the arrester elements was determined with the aid of a sphere gap with spheres 120 mm in diameter.

Test Results

Impulse tests of the RVS-220 arresters in accordance with the program presented above were carried out for the factory installation and for the modified designs according to variants 2 and 3.

Before the tests of the RVS-220 arrester, each of the eight elements of the arrester was tested separately.

The sparkover voltage of all RVS-30 arresters proved satisfactory and had a value of 110–120 kV.

Figure 3 presents the volt-second characteristics of the RVS-220 arrester in the factory design for different depths of insertion of the shielding ring. It is apparent from Fig. 3 that reducing the insertion depth of the shielding ring from 1200 mm (normal design) to 900 mm considerably lowers the discharge voltages at discharge times of less than 1 μsec (curve 2).

Fig. 3. Volt-second characteristics of the RVS-220 arrester in the factory design:
1 — shielding ring \(D = 1880\) mm with insertion depth 1200 mm; 2 — the same with insertion depth 900 mm. ×× — negative voltage impulse, ○○ — positive voltage impulse.

Figure 4 presents the volt-second characteristics for variant 2, and Fig. 5 for variant 3.

Changing the insertion depth of the shielding ring within the same limits (from 1200 mm to 900 mm) in the case of variant 2 leads to an even greater reduction of the discharge voltage (Fig. 4, curve 2), and the discharge voltages fall below the permissible values. The volt-second characteristics for variant 3, with shielding rings from RVS-110 arresters, at voltage exposure times of less than 1 μsec decrease to 300 kV and lie considerably below the permissible values. The reduction of the discharge voltages with a shallower insertion of the shielding ring is explained by a more nonuniform voltage distribution over the arrester elements, since at the initial moment, for small voltage exposure times, the voltage is distributed over the capacitances of the arrester [2].

Variant 3 is the most unfavorable in this respect, since the breakdown voltage of the arrester is determined mainly

by the value of the voltage on the first elements, and the presence of a shielding ring with a shallow seating depth does not provide a sufficient reduction of the voltage on the first element. Within the limits of measurement error, the polarity of the voltage impulse has no effect on the discharge voltage of the arresters.

Fig. 4. Volt-second characteristics of the RVS-220 in variant 2:
1 — normal ring \(D = 1880\) mm, seating depth 1200 mm;
2 — the same with seating depth 900 mm.

Fig. 5. Volt-second characteristic of the RVS-220 arrester in the design according to variant 3.

The sparkover voltage of the arrester under the action of power-frequency voltage and impulse voltage of positive and negative polarity, both when the polarities of the voltage half-wave and the impulse wave coincide and when they are opposite, is determined by the impulse discharge voltage and correspondingly

will affect the arrester sparkover voltage only under the action of an impulse voltage.

Table 2 presents the results of measuring the voltage distribution over the arrester elements with a normal grading ring and with a grading ring set at a smaller depth, for variant 2.

Table 2

As can be seen from Table 2, with a normal grading ring, on each RVS-30 element, when the nominal phase voltage \(U_{\phi}=127\) kV rms is applied, the voltage does not exceed the maximum permissible value of 24 kV rms; at the highest permissible voltage, equal to 192 kV rms, the voltage on the first and second elements is 30 kV rms, which is considerably lower than the puncture voltage of the gaps of the RVS-30 arrester at power frequency, which is 56 kV rms.

The RVS-220 arrester in the design according to variant 2 also successfully withstood a two-minute test at the highest permissible voltage of 192 kV rms.

Conclusions

1. The studies carried out on modified designs of RVS-220 arresters showed that, when the arrester is structurally made in the form of two support columns (variant 2), the arrester, in its electrical characteristics, satisfies the necessary requirements and can be installed for trial operation in existing 220 kV outdoor switchgear installations.

2. The modified arrester design has better mechanical properties and is more convenient in operation when preventive inspections are carried out.

Literature

1. P. G. Kovchavtsev, V. A. Kostrygin, K. S. Storozhuk. Reconstruction of valve-type arresters of the RVS-110 type, Elektricheskie stantsii 2, 1959.

2. A. M. Zalesskii. Insulation of High-Voltage Apparatus. Gosenergoizdat, 1960.

3. I. I. Kalyatskii, N. M. Torbin. Study of the characteristics of valve-type arresters in modified designs, Certificate of Registration in the USSR State Committee for Inventions and Discoveries No. 41204 of December 6, 1963.