By D.F. Peelo
Introduction
In the past, circuit breaker closing resistors were used as the chief method to control switching surge levels on EHV systems. These resistors added to the mechanical complexity of the circuit breakers and generally required a greater degree of maintenance than circuit breaker interrupters. Driven by a desire to eliminate closing resistors, BC Hydro adopted controlled high-speed auto-reclosing on its newest 500 kV line.
Line auto-reclosing on an EHV system is essential. Controlled closing is not simply the provision of a control device because it may malfunction, blocking the reclose operation or operate improperly by closing the circuit breaker at a time other than the "optimal instant." In the scheme adopted by BC Hydro, the closing signal to the circuit breaker is guaranteed. Malfunction or misoperation of the control device is mitigated by staggering the close signals to the control device using special features in the control device itself, and by providing special low protective level metal oxide surge arresters at the line terminals and in the middle of the line.
This paper describes the studies undertaken to determine the requirements and specification for the surge arresters, the selection and type testing of the surge arresters, and, finally, the results of actual line swtiching tests to demonstrate the effectiveness of the switching surge control scheme.
EMTP Studies
The intent of the swtiching surge control scheme was to ensure that the switching surge over-voltage level would not exceed 1.7pu at any point along the line (1pu = 550/2/ 3 = 450kV peak). The line itself is both series and shunt compensated (Fig.1). The series compensation operates such that if the line trips out, the series capacitor bank is automatically bypassed and thus is not a factor in the reclose. The shunt compensation is fixed at one end (reactor neutral is isolated through a metal oxide arrester for secondary arc control) and switchable at the other end. This means that the control device must be capable of coping with low and high levels of shunt compensation and associated wide variations in the line side voltage used to derive the optimal closing instant.
Fig. 2 shows the cumulative frequency distribution of switching surge over-voltages under the following conditions: no control; control by closing resistors; control by staggered closing; control by staggered closing and two or three (1.5pu protective level) surge arresters; and control by controlled closing and three of the above-noted surge arresters. In this instance, controlled closing means closing at a source side 60Hz voltage zero crossing rather than at minimum or zero voltage across the circuit breaker as is the goal of the control device. It is clear that the only scheme which will meet the 1.7pu limit requirement at the line terminals is the controlled switching option.
Fig.3 shows the over-voltage profile along the line for the cases of closing resistors and two or three surge arresters. The value of providing the surge arresters is evident, but not sufficient to meet the 1.7pu limit at all points along the line. The addition of controlled closing was judged to be capable of reducing the highest point over-voltage at about 75 per cent of the line length to between 1.6 and 1.7pu. No specific voltage profile was calculated for this case, however, field tests confirmed this judgment to be accurate.
A number of statistical studies were run to determine the requirements for the surge arresters. The studies were based on surge arresters with a protective level of 1.5pu at 2kA. Fig.4 shows the results of the studies of surge arrester currents and absorbed energies. Based on these results, the surge arrester basic requirements became a protective level of 1.575pu @ 2kA and a single shot energy absorption capability of 2.5 MJ. Detailed requirements for the surge arrester are discussed below.
The specification for the surge arresters was based on IEC 99-4 rather than IEEE C62.11. The reason for this is that the IEEE rating basis is invalid because energy absorption capability is not related to rated voltage. Metal oxide surge arresters are thermally limited and must be rated on this basis as is done in the IEC approach. According to the IEC, rated voltage is an over-voltage and rating is calculated on the basis of absorbing rated energy and demonstrating thermal stability at rated voltage applied for ten (10) second followed by the (maximum) continuous operating voltage (COV) for thirty minutes. The merit and wisdom of this approach is clearly evident in this case.
The specification of COV, rated voltage, duty cycle, protective level and TOV capability for the surge arresters was as follows:
COV: The COV demanded by the system is 318 kV rms. This COV value is the same for standard station arresters at 500 kV.
Rated Voltage: The rated voltage chosen for the surge arresters was 372kV. This is lower than the rated voltage value used for station arresters. Had 396 kV rated surge arresters been used, then the arresters would comprise ten or more columns in order to achieve the desired protective level. At 372 kV, two column arresters were envisaged; in reality, two and four column arresters were supplied by the two manufacturers qualified to provide them.
Duty Cycle: The duty cycle was chosen as 2.5MJ - 60s - 2.5MJ. This reflects the IEC approach as noted above and discussed further below. The chosen cycle provides a degree of margin since two 100 per cent single shot events in succession are highly improbable.
Protective Level: The protective level was chosen as 1.575pu at 2kA. This value was somewhat higher than that used in the studies (1.5 pu) but the difference was judged to be insignificant. The value was based on prior discussion with potential suppliers in anticipation of two column arresters.
Temporary Over-voltage (TOV): The studies showed that the maximum TOV would not exceed the rated voltage of the arresters and the associated application time of ten seconds.
A comparison of the requirements for standard station arresters at 500kV and the line surge arresters is provided in the table below.
The most important point with regard to the rating of the surge arresters is that the rating is not reliant on correct functioning of the closing control device. The rating is based on staggered closing of the circuit breaker, and thus in the event of malfunction of the controlled closing device, over-voltages will be mitigated to the degree shown in Fig.2. With near optimal performance of the controlled closing device, energy absorption by the surge arresters will of course be minimal.
Surge Arrester Selection and Type Testing
Two main concerns needed to be addressed in the selection and type testing of the surge arresters. These concerns were, chiefly, the ability to meet the duty cycle requirements and, secondly, the ability to survive in the long-term under steady state conditions.
Tenders were received from a number of suppliers and each was technically reviewed and then ranked in terms of compliance and technical merit. In regards to the latter, the arresters were ranked on the basis of energy absorption capability, TOV capability and losses. The losses were compared by calculating the loss per unit volume per longitudinal voltage stress at COV. This process ultimately resulted in the awarding of two contracts.
The IEC Operating Duty Test confirms the rating basis of an arrester. For standard station class arresters, IEC defines energy absorption capability on the basis of five line discharge classes. Each line class number is the energy per shot in kJ/kV rated voltage that an arrester with a switching surge protective level equal to two times rated voltage will absorb in the operating duty test. Following a conditioning part, the test samples are heated to 60ūC and two shots of rated energy (line class) are injected 60 seconds apart, followed by rated voltage of 10 seconds and COV for 30 minutes. Applying this approach to the subject surge arresters, it could be said that the arresters have an equivalent line discharge class of 6.7kJ/kV rated voltage at the required protective level. The results of the operating duty tests are shown in Fig. 5 in terms of tested values superimposed on standard IEC requirements (Fig E.1 of IEC 99-4). The tested values exceed the required values due to waveform requirements and the individual V-I characteristics of the tested arresters.
In addition to the above test, arresters from both suppliers were successfully subjected to all the other type tests as required by IEC 99-4.
To demonstrate the performance of the overall controlled switching scheme, a series of line swtiching and auto-reclosing tests were carried out with one and two reactors connected. Dependent on the degree of closing event accuracy with respect to minimum voltage across the circuit breaker, the surge arresters absorbed varying energy magnitudes, the highest single absorption being about 1 MJ.
Due to an unplanned occurrence during the course of the testing, the surge arresters were exposed to one of the highest TOVs ever encountered on the BC Hydro 500 kV system. The arresters actually limited the voltage excursion (Fig.6).
Conclusion
The project demonstrated the viability of eliminating circuit breaker reclosing resistors by a combination of controlling closing and low protective level surge arresters. Using the IEC approach to metal oxide surge arresters, it is possible to specify and test the special arresters to the true requirements of the application. ET
D.F. Peelo is with B.C. Hydro