cahier technique no. 101studiecd.dk/...of_sf6_puffer_circuit_breakers.pdf · fig. 3: pole of sf6...

..........................................................................

Cahier technique no. 101

The behaviour of SF6 puffercircuit-breakers underexceptionally severe conditions

J.C. HenryG. PerrissinC. Rollier

Collection Technique

"Cahiers Techniques" is a collection of documents intended for engineersand technicians, people in the industry who are looking for more in-depthinformation in order to complement that given in product catalogues.

Furthermore, these "Cahiers Techniques" are often considered as helpful"tools" for training courses.They provide knowledge on new technical and technological developmentsin the electrotechnical field and electronics. They also provide betterunderstanding of various phenomena observed in electrical installations,systems and equipments.Each "Cahier Technique" provides an in-depth study of a precise subject inthe fields of electrical networks, protection devices, monitoring and controland industrial automation systems.

The latest publications can be downloaded from the Schneider Electricinternet web site.Code: http://www.schneider-electric.comSection: Experts' place

Please contact your Schneider Electric representative if you want either a"Cahier Technique" or the list of available titles.

The "Cahiers Techniques" collection is part of the Schneider Electric’s"Collection technique".

ForewordThe author disclaims all responsibility subsequent to incorrect use ofinformation or diagrams reproduced in this document, and cannot be heldresponsible for any errors or oversights, or for the consequences of usinginformation and diagrams contained in this document.

Reproduction of all or part of a "Cahier Technique" is authorised with theprior consent of the Scientific and Technical Division. The statement"Extracted from Schneider Electric "Cahier Technique" no. ....." (pleasespecify) is compulsory.

no. 101The behaviour of SF6 puffercircuit-breakers underexceptionally severe conditions

ECT101 updated January 1983

J.C. HENRY

Ingénieur IEG.Chef du Service Technique branche Transport Haute Tension.

G. PERRISSIN

Ingénieur PST.Station d’Essais à Grande Puissance.

C. ROLLIER

Ingénieur ESE.Service Technique Très Haute Tension.

Cahiers Techniques Schneider Electric no. 101 / p.2


* The behaviour of SF6 puffercircuit-breakers under exceptionallysevere conditions

The development of transmission power systems and industrial powersystems places high-voltage circuit-breakers in operating conditions muchmore severe than those taken into account by the Standards. Twosituations are discussed:c the case of very long lines for which difficulties encountered upon theenergisation and de-energisation of no-load lines and upon shunt reactorswitching are reviewed;

c the case of powerful transformers with low impedance voltage whosehigh natural frequency is at the origin of a severe transient recovery voltagewhen a fault takes place on the secondary side of the transformer.

In both cases, it is demonstrated that the SF6 puffer circuit-breaker has asatisfactory behaviour though some of the conditions under considerationare exceptionally severe. It is generally unnecessary to resort to the use ofauxiliary resistors, except for the energisation of long HV lines.

*Report number 13.08 presented at CIGRE, Conférence Internationale desGrands Réseaux Electriques (International Conference on Large HighVoltage Electric Systems), session in 1978.

Appendix:Overvoltages upon low inductivecurrent breaking

3 Transformer secondary fault

Bibliography

Contents

1 Introduction p. 4

2 Long extra-high-voltage lines 2.1 Energisation of open-ended lines p. 5

2.2 De-energisation of open-ended lines p. 6

2.3 De-energisation of shunt reactors p. 8

3.1 Stresses p. 11

3.2 Tests results p. 11

3.3 Conclusions p. 12

Current chopping p. 13

Successive re-ignitions p. 14

p. 15


1 Introduction

The knowledge of the operation of high-voltagetransmission systems and of the phenomenataking place on them upon operation of theirprotective circuit-breakers has been steadilyprogressing over the last twenty years. Thetheoretical study of operating requirements, theanalysis of failures and CIGRE work haveresulted in the recapitulation of all the conditionswhich must be taken into account for the designand verification of the switching devices intendedto be used in high-voltage power systems. In theend, the process has been materialized by theinclusion of these switching conditions within thescope of international standards whose volumeand complexity reflect the extent of the workcarried out, a few points being still object ofactive work.However, from time to time, special situationsnot directly referring to the operating conditionscovered by the standards may occur. Forinstance, the extensive application of hydro-

electric resources in some countries,necessitating the installation of very long lines,explains why network designers have to definenon-standardized conditions for the verificationof circuit-breakers. These problems arediscussed in the first part of this paper.

Exceptional conditions also affect someinstallations incorporating powerful transformerswith low impedance voltage. These specialinstallations generate heavy stresses for thecircuit-breakers; these have not been taken intoaccount in the standards because they differ toomuch from the severity conditions generallyencountered by power system circuit-breakers.This second problem is dealt with in the secondpart of this paper.

In both cases, the stresses withstood by thecircuit-breaker, as well as the test proceduresapplied to check the satisfactory behaviour of anSF6 puffer circuit-breaker, are reviewed.


2 Long extra-high-voltage lines

The SF6 gas puffer technology used long sincein switchgear for high-voltage distributionsystems has been progressively extended to theswitchgear of extra-high-voltage power systemsowing to its advantages [1].

The very good experience gained in the field sofar has led system users to try and expand thenumerous advantages of this switchgear tosystems of higher voltages, in particular to the525 kV systems which constitute the backboneof transmission systems of numerous countriesin the american continent where the heavypowers to be conveyed and the remoteness ofload centers have been in favour of the decisionfor such a high voltage level. It must be recalledthat the choice of a high voltage level is notreally advantageous unless provision has been

made for limiting the temporary overvoltages andthe switching surges likely to occur in a high-voltage system. Without this limitation, theadditional cost of the insulation to be providedfor the system that would have to withstandheavy overvoltages reduces to nothing thesavings achieved with the reduction of losses.

The development of systems having ratedvoltages of 525 kV or above has therefore, inparticular, necessitated the consideration ofthree operational conditions that are likely togenerate the highest overvoltages:

c closing and reclosing of open-ended lines;

c line charging current switching at exceptionallyhigh voltages;

c shunt reactor switching.

2.1 Energisation of open-ended lines

A circuit-breaker protecting a line may have toenergize it under open-ended conditions. Theovervoltages due to the reflection at the openend must absolutely be mastered. Theovervoltage levels to be complied with are notcovered by international standards at presentand their specification remains within theprovince of the system designer. Among all themethods which have been suggested to limitclosing surges, the simplest one consists inenergising the line through a resistor chosen incompliance with the characteristics and length ofthe line.

The circuit-breakers able to protect 525 kV and765 kV systems must therefore be equipped withauxiliary interrupters allowing the resistors to beinserted during a predetermined time. Thisservitude, quite practicable with air-blast circuit-breakers, is also suited for SF6 puffer circuit-breakers. It is indeed possible, to the very simplemechanism of these circuit-breakers, to add alinkage system driving the resistor insertioncontacts during a closing operation (see Fig. 1 ).These contacts automatically move back to theiropen position immediately after the maincontacts have closed. Such a linkage systemassures an excellent accuracy of the insertiontimes of the resistors upon closing.

The value of the resistor may be selected bymeasurements on mode1 networks or bycalculation. In particular, measurementsperformed on a transient analyzer have made itpossible to determine the maximum values of theresistors and the minimum insertion times to be

Fig. 1: sectional view of a pole unit of 525 kV circuit-breaker equipped with closing resistors.1. break2. grading capacitor3. closing resistor4. support insulator

Cahiers Techniques Schneider Electric no 101 / p 6

provided to limit the overvoltages during areclosing operation on a 400 km long 525 kV lineto 2.2 p.u., if the line is not compensated, and to2 p.u., if the line is compensated (see fig. 2 ).

SF6 puffer circuit-breakers equipped with a set ofauxiliary interrupters (see fig. 3 ) incorporatingresistors are therefore able to meet therequirement of limitation of reclosing surges,which requirement is of major importance todetermine the insulation level of extra-high-voltage systems.

Fig. 3: pole of SF6 puffer circuit-breaker with 4 breaksand closing resistors.rated voltage = 525 kV; rated breaking current = 50 kA;rated current = 3,150 AFig. 2

2.2 De-energisation of open-ended lines

Stresses

The severity of the conditions imposed upon acircuit-breaker when opening an open-ended linemay be such that it is these breaking conditionswhich dictate the size of the circuit-breaker, inparticular the selection of the number of breaks.The major fact is that, half a cycle after theinterruption, the circuit-breaker must accept avoltage across its terminals at least equal totwice the peak value of the phase-to-earthvoltage of the system prior to the interruption.Unfortunately, at the time of opening, it mayhappen that the phase-to-earth voltage of thepole which has to open has reached valuesmuch higher than the values stipulated in thestandards for testing the circuit-breaker in suchinterrupting conditions.

This dynamic voltage rise may be the result of anumber of causes. In particular, the opening of acircuit-breaker sited at the receiving end of a lineconveying a heavy load will leave the line openat its end. The voltage of the line increasesowing to the sudden disappearance of the loadthat is not immediately compensated by thevoltage regulation and owing to the capacitive

load constituted by the line. Consequently, thecircuit-breaker sited at the sending end may beled to de-energise the line while the phase-to-earth voltage at the sending end hassubstantially exceeded the normal value.

The special conditions of some systems haveevidenced the possibility of high dynamicovervoltages, of about 1.5 p.u., in spite of thefavourable effect achieved on the limitation ofdynamic overvoltages by compensation reactors.For instance, the phase-to-phase voltages of a525 kV system and of a 765 kV system maytemporarily reach 750 kV and 1,100 kVrespectively.

These conditions are exceptional and it is quitenormal that such situations are excluded fromthe verifications stipulated in the standards forline-charging current interruptions. However, thefact that such situations may actually occur hasmade it necessary to check the ability of thecircuit-breakers to withstand such voltages. Eventhough it can be admitted that such verificationsare carried out on the system in the field, themanufacturer must a priori demonstrate theability of his switchgear.

Non Compensatedcompensated lineline 40 % 70 %

Resistor value 360 360 1,000(ohms)

Insertion time (ms) 10 8.4 10

Overvoltage factor 2.2 2 2(98 % cumulativeprobability) p.u.


Test procedures

For system voltages under 245 kV, a direct testcan generally be made in a testing station andwith an actual line. As soon as the systemvoltage reaches 420 kV, the direct test becomesmore difficult due to the operational conditionsoften preventing the availability of a no-load lineof sufficient length.

Another test procedure consists in simulating theno-load line by means of a capacitor bank. Hereagain the maximum possibilities of thelaboratories are rather rapidly reachedconsidering the size of the bank necessary toobtain the high currents simulating lines of longlength with the high voltages previouslymentioned. The manufacturer is thus led toperform tests no longer on a complete pole, buton a portion of a pole, even on one break. Thesetests can be made either on a direct circuitincluding capacitor banks of high capacitance, orwith a synthetic circuit. It is the latter methodwhich we have used with the diagram shown inFigure 4. Such a circuit offers the advantage ofusing only capacitor banks of small size. Indeed,in the “current” circuit, where a high capacitancevalue is required, a comparatively low voltage issufficient; in the “voltage” circuit, a capacitanceof low value insulated for the full voltage issuitable.

Fig. 5: oscillogram recorded during capacitive currentbreaking tests on the circuit illustrated in Figure 4.

Fig. 4: diagram of synthetic circuit for line-chargingcurrent breaking tests.- Dp: back-up circuit-breaker- Db: auxiliary circuit-breaker- De: circuit-breaker under test- T: transformer with 2 secondary windings- G: generator- S1: “current” circuit- S2: “voltage” circuit- C1: capacitor bank of “current” circuit- C2: capacitor bank of “voltage” circuit- U: recovery voltage at the termmals of De- I: breakmg current of DeI = I1 + I2; U = U2 - UC2

The unit-testing method for the line-chargingbreaking currents is not explicitly provided for instandards and its application requires someprecautions. For circuit-breakers with a shortminimum arcing time (this is the case of SF6puffer circuit-breakers), the deviation in thesimultaneous operation of the breaks of one polemust not exceed 2 ms approximately. Theovervoltage on the first break to open mustindeed be negligible. A rapid calculation showsthat, in the event of a minimum arcing time of1 ms, a 2 ms deviation involves a 7 percentovervoltage on the first break to open in a4-interrupter circuit-breaker, and a 12 percentovervoltage in a 6-interrupter circuit-breaker.Consequently this method is quite applicable tothe circuit-breaker referred to in this reportwhose simultaneity of operation of theinterrupters is properly ensured and whichincorporates only a small number of breaks.

Test resultsThe tests have been made on a break of the SF6puffer circuit-breaker illustrated in figure 3. Thetests represent the stress withstood by onebreak of a 4-break circuit-breaker should thephase-to-phase voltage of the 525 kV systemreach 750 kV.Figure 5 shows the oscillogram of such abreaking operation.

The test voltage is determined by a relationshipas follows:

U x x kV2 1 2750

3

14

. =


Fig. 7: connection diagram of shunt reactors.

The tests have been made in compliance withthe standards as regards the instant of contactseparation. The results are given in the table infigure 6.

The results, that do not show any restrike, provethe ability of the circuit-breaker to interrupt no-load lines under the severe conditions describedpreviously.

Fig. 6: results of line-charging current breaking tests.

2.3 De-energisation of shunt reactors

The application of shunt compensating reactorson the lines is practically always necessary inextra-high-voltage systems. Indeed they make itpossible to avoid too high over-voltages alongthe line if this line is at no load or with lightload.The extremely favourable influence ofcompensating reactors on the dynamicovervoltages produced by load rejection at theend of a long line is also known. Lastly, thereactors also have a favourable influence on thelimitation of switching surges when closing orreclosing lines at no load. Various possibilitiesare thus available to benefit from the advantagesof shunt reactors:

c the permanent coupling, in parallel with eachphase of the line, of an inductor whose value ischosen to be acceptable under all the operatingconditions of the system;

c or the connection of the inductor to the linethrough a circuit-breaker whose controlledclosing or opening allows a greater flexibility inthe application of the reactor according to theload conveyed by the line (see Fig. 7 ).These circuit-breakers operate in specialconditions since they have to interrupt a smallcurrent (a few hundred amperes) and since theyoperate very frequently. Their mechanicalreliability must therefore be very high and theymust not generate abnormal overvoltages whenbreaking this current.The excellent reliability of SF6 puffer circuit-breakers has already been mentioned inpublications [1] ; it is mainly due to the simplicityof their design.

On the contrary, the behaviour of circuit-breakerswhen breaking small inductive currents is ratherpoorly known and it appears it is difficult to laydown the corresponding test standards owing tothe great number of parameters likely to beinvolved and to the contingent nature of theresults generally obtained.

In addition, it is extremely rare to be in a positionin a laboratory to correctly represent the actualoperating conditions of the circuit-breakersdesigned for very high voltages.

Two difficulties are generally encountered:

c test voltage too low;

c inherent capacitances of the test circuit toohigh.

It is therefore strongly desired to predeterminethe overvoltages likely to occur in any operatingconditions, on the basis of the results of testsmade in definite conditions, if possible with a

No. U2 U I Arcing(kV) peak (A) time

value (ms)(kV p)

1 128 345 815 2

2 128 340 815 9

3 128 353 815 8

4 128 340 815 6

5 128 352 815 5

6 128 352 815 3

7 128 353 815 2

8 128 355 815 1

9 128 332 815 9

10 128 340 815 8

11 128 332 815 7

12 128 340 815 5


Fig. 8: diagram of circuit for inductive current breakingtests.

small number of breaks in series. Some authorshave already demonstrated some laws ofvariation of the chopped current, an overvoltagegenerator [2]. It will be seen that the resultsobtained in a testing station with an SF6 puffercircuit-breaker of the type illustrated in Figure 3corroborate those laws and that it is thuspossible to evaluate the maximum overvoltageslikely to be generated by such a circuit-breaker.

Test conditionsTwo test series, totalling over 100 breakingoperations, were made on several breaks of thecircuit-breaker.

In both series, the test circuits were single-phaseand their main characteristics were as follows(see Fig. 8 ):

c Series No 1U = 235 kVf = 50 HzI = 245 - 517 - 1,100 AC1 = 1 mFC2 = 46 - 127 µFNumber of breaks in series = 3.

c Series No 2U = 20 - 40 kVf = 50 HzI = 250 - 500 AC1 =17 nFC2 = 1.9 to 12 nFNumber of breaks in series = 1 or 2.

Series No 1, performed on a 3-interrupter circuit-breaker, is representative of the operation of a4-interrupter circuit-breaker on a 525 kV powersystem. The voltage of the test circuit, i.e. 235 kV,was the highest voltage available in thelaboratory. During the series, it was not possibleto reduce the capacitance of the load-side circuitto a value low enough to be representative of theinherent capacitance of a shunt reactor.Therefore tests were performed at a reducedvoltage, with low-value capacitances on theload-side, in order to study the influence of thenumber of breaks and of the capacitance on thechopped current value.

Test resultsTwo phenomena are likely to take place whenbreaking small inductive currents: currentchopping and successive re-ignitions(see appendix 1).

During both tests series, current chopping wasobserved almost systematically, but no breakingoperation gave rise to successive re-ignitions.This result is very important because it meansthat the overvoltages produced by the circuit-breaker can be predetermined with certainty if itis possible to know the law of variation of thechopped current in relation to the parameters ofthe circuit. Some results previously obtained bythe authors during tests of SF6 puffer circuit-breakers, together with some theoretical andexperimental studies already published, showthat the chopped current would be determinedby a relationship as follows:

I0 3 '= λ n C (6)

where Io, is the chopped current, λ a factorspecific to the circuit-breaker, expressed inAmpere (Farad)-1/2, n the number of breaks inseries per pole, C’3, the capacitance in parallelwith the pole.To determine factor λ by means of the testresults, only the results obtained for arcing timesequal to or greater than 5 ms are taken intoaccount. Shorter arcing times give rise tochopped currents which are not worth studyingdue to their low values and dispersion, as well asto the inaccuracy of their measurement.

The table in Figure 9 shows the average valuesof λ obtained on the different test circuits, eachaverage value being generally calculated on 5tests. It can be seen that the values thusobtained are very similar, whereas the testconditions cover a very wide range of values forparameters n and C’3; this proves thatrelationship (6) is quite applicable to this type ofcircuit-breaker.

Fig. 9: average value of factor λ for arcing times longer than 5 ms.

Number of interrupters 1 1 1 1 1 2 2 2 2 3 3 3

Voltage (kV) 20 20 20 20 40 20 20 20 40 235 235 235

Breaking current (A) 250 250 250 250 500 250 250 250 500 245 517 1,100

C’3 (nF) 4.2 5.2 9.2 9.5 9.2 3.9 7.9 8.3 7.9 110 45 47

λ.10-3 94 89 95 90 92 81 84 96 96 81 74 92


Fig. 10: histogram of factor λ.

Fig. 12: calculated probability of overvoltage factors.

Since factor λ does not depend upon the testcircuit, it is interesting to analyze its statisticaldistribution for all the tests under consideration.

Figure 10 represents the histogram of thevalues of λ and shows a Gaussiandistribution :

c the average value λ . /= −88 5 103 1 2x A F ;c the standard deviation σ = 14 x 103 A F-1/2.

The cumulative frequency curve of l is shown inFigure 11 in which the normal lawcorresponding to λ and σ is also plotted. It canbe seen that the distribution of λ follows anormal low, in particular for the values higherthan the average value. This makes it possible tocalculate the probability of occurrence of highvalues of chopped current.

Calculation of overvoltages

In the general case where the source-sidecapacitance is high compared with the othercapacitances of the circuit, the overvoltage factoris given by the following relationship as workedout in appendix 1:

kn L

Um = +1

22

2λ

(10)

where L2 is the load-side inductance and Um theamplitude of the phase-to-earth voltage.

Since factor λ is known statistically, it is possibleto calculate the probability of occurrence of theovervoltages for the 4-break circuit-breaker usedto control shunt reactors in a 525 kV powersystem. For instance, let us consider 3 values ofreactance corresponding to single-phase powersof 37, 75 and 150 MVA respectively. Whenapplying the normal law defined in the precedingparagraph to λ, the results listed in the table inFigure 12 are obtained.

These predetermined overvoltage levels, thoughbased on a comparatively narrow sampling(about 100 tests) nevettheless show that thecircuit-breaker under test will not generateabnormally high overvoltages when in service onthe power system.

For comparison purposes, tests made on an air-blast circuit-breaker have made it possible todetermine an average value of 230 A F-1/2 forfactor λ, i.e. three times the value correspondingto the SF6 puffer circuit-breaker. The applicationof the method of predetermination of inductive-

Fig. 11: cumulative frequency of values of factor λ.experimental resultsnormal law: λ = 88.5 x 103; σ = 14 x 103

current breaking overvoltages to a 6-break air-blast circuit-breaker, for the values of reactancespreviously considered, shows that theovervoltage factors would substantially exceedthe permissible levels.

Consequently, it would be necessary to limit theovervoltages by means of resistors. In thisrespect, results of comparative tests performedon Hydro-Quebec power system are available;they show that the de-energisation of reactors byair-blast circuit-breakers not including openingresistors is featured by unacceptableovervoltages.

It is therefore conclusive that this majoradvantage of the SF6 puffer circuit-breaker underconsideration will be appreciated, since it makesit possible to use a circuit-breaker withoutresistors for reactor switching, without any riskfor the insulation of the reactors.

k

P 150 MVA 75 MVA 37 MVA

10-2 1.27 1.5 1.87

10-3 1.32 1.57 2

10-4 1.35 1.62 2.07


Fig. 13: oscillogram of the TRV recorded during theinterruption of a fault on the secondary side of a220/60 kV transformer.P= 150 MVA; Impedance voltage = 10.3 percent.

3 Transformer secondary fault

3.1 Stresses

The severity imposed upon a circuit-breaker bythe conditions produced upon interruption of ashort-circuit occurring on the secondary side of atransformer has already been described [3] [4]. Ithas been demonstrated that the interruption ofsuch a short-circuit, though its magnitude isdefinltely below the interrupting capability of thecircuit-breaker, could cause difficulties to sometypes of circuit-breakers. In particular, somesensitivity of small-oil-volume circuit-breakershas been explained by the fact that, forcomparatively low fault currents, the de-ionizingcapacity, which depends upon the magnitude ofthe breaking current, was not high enough owingto the rate at which the voltage recovers in acircuit oscillating at very high frequency. It hasalso been noticed that, in some installationconditions, the air-blast circuit-breaker was not ina position to break a short-circuit currentcorresponding to 40 percent of its breakingcurrent, owing to the value of the recoveryvoltage frequency that mainly includes the20 kHz oscillation of a 150 MVA transformer(see Fig. 13 ).

Such situations are more and more frequentlyencountered in installations fed at voltages from72.5 kV to 170 kV and their severity increasesdue to the power rating of the transformersinstalled. The severity of the breaking conditions

also increases with transformers of lowimpedance voltage that are used for the supplyof some industrial installations.

The conclusions drawn from studies conductedon such power systems show that the TRV’srecorded with a fault fed through the transformerare definitely more severe than those stipulatedin standards for short-circuit currentscorresponding to 10 percent and 30 percent ofthe breaking current. It was therefore importantto make sure that the SF6 puffer circuit-breakeris not led into difficulties in such cases.

3.2 Tests results

The unit under test is a one-break SF6 puffercircuit-breaker (see Fig. 14 ) intended for use inpower systems with voltages from 72.5 to170 kV. The use of an air reactor, located on theload-side of the circuit-breaker (see Fig. 15 )allowed the tests to be performed for low currentvalues (1 and 2.5 kA).

Fig. 15: test circuit including a high reactance on theload-side of the circuit-breaker.

Fig. 14: SF6 puffer circurt-breaker with 1 break forvoltages from 725 kV to170 kV.


Fig. 16: results of break tests at high frequency.

The possibilities of adjustment of the current onthe circuit were limited and a current-injectionsynthetic circuit was used after-wards, fora range of breaking currents from5 kA to 20 kA.

The table in Figure 16 summarizes the testconditions and the results.

The tests performed represent severityconditions much in excess of those stipulated inthe standards.The results prove the ability of this circuit-breaker to overcome the most severe stresseslikely to take place when faults are fed bypowerful transformers, such as encountered insome types of installations.

3.3 Conclusions

It has been shown that, for circuit-breakers usedin power systems or industrial installations,exceptional operating conditions may occur,these being widely different from or not yet fallingwithin the scope of standardized conditions.

They are more particulary related to:

c the breaking and making of line chargingcurrents at exceptionally high voltages;

c the breaking of shunt reactor currents;

c the breaking of transformer secondary short-circuit currents.

In each case it has been possible, by laboratorytests, to demonstrate that the SF6 puffer circuit-breakers under consideration are capable ofcoping with extremely severe conditions underwhich circuit-breakers of former technologiescould have difficulties.

Tests Breaking Number 1st peak Frequency T1 or T3 RRRVcurrent of tests voltage F (kHz) (µµµµµs) (kV/µµµµµs)I (kA) U1 or Uc

(kV)

Tests with 1.1 5 140 17 26 5.4reactor on 2.5 6 85 28 15 5.7load side

Synthetic 5 10 126 33 13.5 9.3tests 10 3 126 22 20 6.3

10 2 126 50 9 1415 3 139 22 20 7

5 3 250 18 24.5 10.210 5 250 18 24.5 10.215 3 250 19 23 10.920 2 250 21 21 11.9


Appendix:Overvoltages upon low inductive current breaking

The phenomena likely to generate overvoltagesupon low inductive current breaking are wellknown. There are two types as follows:

c premature current interruption, commonlytermed “current chopping”;

c successive re-ignitions.

These two phenomena can in fact take placesuccessively during the same operation(see Fig. 17 ).

In both cases, the current id is interrupted wheni2 is not zero, owing to high frequencyoscillations that are superimposed on the power-frequency component of the current in thecircuit-breaker.

Fig. 17: representative diagram of an inductive currentinterruption.

Fig. 18: interruption with current chopping:id: current in circuit-breakeri2: current in load-side inductoru2: load-side voltage

Current chopping

Current id is interrupted when current i2 is equalto i0, and voltage u2 to U0 (see Fig. 18 ).If the damping of the load-side circuit can beconsidered as negligible during 1/4 cycle of itsnatural oscillation, the calculation of the over-voltage is straightforward :

U ULCc

'= +0

2 2

20

2I (1)

where C’2 is the capacitance in parallel oninductance L2 after the break:

C CC C

C C'

2 21 3

1 3= +

+ (2)

If Um is the amplitude of the load-side voltageprior to the break:Uo ≈ Um (3)

the overvoltage factor kUU

c

m =

is then written as follows:

k = +1 2ε (4)

where ε '

=I0 2

2ULCm

(5)

It is worth studying term ε when the value of thechopped current confirms the law:

I0 3 '= λ n C (6)

where n is the number of breaks in series in onepole and C’3 the capacitance in parallel in one pole:

C CC C

C C'

3 31 2

1 2= +

+(7)

Relationship (4) becomes:

ε λ

=

++

nL

U

C C

C Cm

2 1 3

1 2(8)


Fig. 19: interruption with repetitive re-ignitions.

As a rule, the values of the capacitances aresuch that:

C1 >> C2 and C1 >> C3

Relationship (8) therefore becomes:

ε λ =

nL

Um

2 (9)

wherefrom kU

nLm

= +12

2 2λ

(10)

There is however an upper limit for the value ofL2 beyond which this relationship does not applyany more. This limit is reached when thechopped current is equal to the amplitude of thebreaking current.Finally, if a current chopping takes place withoutany re-ignition, the overvoltage factor can bepredetermined. Moreover, if the source-sidecapacitance is high compared with the othercapacitances, the over-voltage level dependsonly on the number of breaks per pole and onthe value of the load-side inductance, for a givenvoltage.

Successive re-ignitions

The phenomenon of successive re-ignitionsillustrated in Figure 19 has been described intechnical literature [5] . It must essentially benoted that, in this case, the overvoltage is due tothe transfer, into the load-side capacitance, ofthe energy that is re-injected into the load-sidecircuit on each re-ignition. The highest voltage isnot necessarily reached upon final interruption; itmay occur earlier, according to the exchange ofenergy between the source-side circuit and theload-side circuit. The overvoltage level dependson numerous parameters such as:c the natural frequency of the load-side circuit;

c the point on current wave of contactseparation;

c the rate of rise of dielectric strength acrosscontacts;

c the characteristics of the high-frequencycurrent oscillation which in their turn dependupon the distance between the source-side andload-side capacitances.

The nature of some of these parameters givesthis phenomenon a very uncertain character andit appears it is very difficult to predetermine theovervoltage level likely to be reached in a givenpower system. This is still made worse by theinteraction which may occur between phases.


Bibliography

[1] “Towards the maintenance-free circuit-breaker”CIGRE 1974 – Report 13-06.R. Michaca, J. Verdon, J.C. Okerman, C. Rollier,J. Daillet, B Trolliet.

[2] “Switching of shunt reactors-comparisonbetween field and laboratory tests”CIGRE 1976 – Report 13-04.S. Berneryd, C.E. Solver, R. Erikson, L. Ahlgren.

[3] “Transient recovery voltage conditions to beexpected when interrupting short-circuit currentslimited by transformers”CIGRE 1970 - Report 13-07.Study Committee No. 13.

[4] “Transient recovery voltage associated withpower system, three-phase transformersecondary faults.IEEE PAS 91 - September/October 1972, pages1887 to 1896.R. Harner and J. Rodriguez.

[5] “A comparison of overvoltage phenomena inhigh-voltage vacuum and air-break switches”Cahier Technique no. 66.J.L. Mircovich, G. Perrissin.

Schneider Electric Direction Scientifique et Technique,Service Communication TechniqueF-38050 Grenoble cedex 9Télécopie : 33 (0)4 76 57 98 60

DTP: HeadLines - MeylanEdition: Schneider Electric- 20 € -

© 2

002

Sch

neid

erE

lect

r ic

12-02

cahier technique no. 101studiecd.dk/...of_sf6_puffer_circuit_breakers.pdf · fig. 3: pole of sf6...

Documents