Sunday, December 9, 2012

CHAPTER 1 THE BREAKING OF A.C. CIRCUITS





1.1       GENERAL


The purpose of any switching device is to make and interrupt current.  This function applies to any such device from the largest circuit-breaker down to the smallest relay.

The flow of electric current in a circuit is a form of kinetic energy, and it cannot be stopped instantaneously.  The energy must be taken out of the system for all current to cease, and this takes a finite time, no matter how small it may be.

1.2       D.C. INTERRUPTION


When the contacts of a switch separate while carrying direct current, that current cannot be stopped there and then but will continue to flow between the opening contacts in the form of an arc.  The arc is a column of white-hot gas which is ionized by its heat and provides a conducting path for the current.  This path has appreciable resistance which increases as the contacts continue to separate and the arc lengthens.  Eventually the resistance becomes so high that the current falls to the point where it can no longer maintain the arc temperature; ionization ceases, the air ceases to conduct and the arc goes out, finally stopping the current.  All energy has been dissipated in the heat of the arc.

1.3       A.C. INTERRUPTION




FIGURE 1.1
A.C. CURRENT BREAKING


The mechanism of current interruption in an a.c. switch is totally different from that of d.c.  With direct current the arc continues uninterrupted until it can no longer be maintained, after which it goes out and stays out.

With alternating current, however, at each half-cycle the current passes through a natural zero, and the arc which had been carrying that current goes out momentarily for lack of heat.  Under the action of the voltage which immediately reappears across the open contacts the gap breaks down, the arc ‘restrikes’ and will continue to carry current for the next half-cycle until, at the next current zero, it goes out again and the process repeats.  This is shown in Figure 1.1.

As the contacts continue to separate, the arc path becomes longer; eventually the gap at a current zero becomes too great for the voltage to break it down, so that the arc does not restrike.  The current has then been finally interrupted, and the energy has been dissipated in a series of successive arcs.

It should be particularly noted that, whereas in d.c. the arc will continue unbroken until it is finally suppressed, in a.c. it extinguishes itself naturally twice every cycle of current, restriking each time until it is no longer able to do so.

From here on only a.c. breaking is considered.

1.4       FAULT CURRENTS


When a short-circuit occurs, it may be between two of the three lines of a 3-phase system, or it may involve all three.  The fault current may pass between phases as an arc, which has some resistance and so limits the current, or there may be metal-to-metal contact, a


FIGURE 1.2

PARTIAL CURRENT ASYMMETRY AT ONSET OF A FAULT



so-called ‘bolted’ fault, where the impedance is zero.  As an item of switchgear must be able to deal with the most severe possible case it is always assumed that the fault is a 3-phase bolted one, and that the whole circuit is mainly inductive with little resistance.

It was shown in the manual ‘Fundamentals of Electricity 3’, Chapter 5, that with an inductive fault the current which immediately follows is in general partially asymmetrical.  The asymmetry will be complete (100%) if the fault occurs at the instant of a voltage zero.  If it occurs at a voltage peak, positive or negative, the asymmetry is zero (0%) - that is to say, the current wave is then wholly symmetrical.

Figure 1.2 shows the general case where the asymmetry is partial (between 0% and 100%).  The point on the voltage wave at which a fault may occur is of course entirely random.  So therefore is the degree of asymmetry which will occur in any particular case.

This asymmetrical current wave is regarded as resolved into two parts: a symmetrical a.c. wave plus a steady but decaying ‘d.c. component’ whose rate of decay is mainly the R/L ratio of the fault circuit; the d.c. component is shown in green in Figure 1.2.  The quicker the decay of the d.c. component, the quicker the fault current resumes symmetry.

With complete asymmetry the first peak of the asymmetrical current wave is almost double the amplitude of the a.c. component at that time - that is 2 x Ö 2 (= 2.82) times its rms value.

FIGURE 1.3
THREE PHASE CURRENT ASYMMETRY (GENERAL CASE)

However by the time the first current peak is reached there has already been some decay of the d.c. component, and it is usual to take the first current peak as approximately 2.55 times the rms value of the a.c. component.  This figure however may differ slightly in special cases.

Although Figure 1.2 shows the a.c. component as having constant amplitude, it does in fact gradually reduce in size as the current moves from its initial subtransient value towards the transient.

For reasons which it is not necessary to go into here, an asymmetrical current is less difficult to break than a symmetrical one.  Therefore, in order that the circuit-breaker is able to deal with the most difficult case, it is required, when testing, that the bolted fault shall continue long enough for the d.c. component to decay to a specified level.  This level depends on the opening time of the breaker itself (i.e. from instant of trip signal to separation of contacts) and may be of the order of 15%, after which the current is regarded as ‘symmetrical’.  At Switchgear Testing Stations deliberate delay is introduced between the onset of the fault and the trip signal to the breaker on test to ensure that this is so.

In 3-phase switching the asymmetry will in general be partial in all phases, as shown in Figure 1.3, and the percentages, taking account of sign, will always add up to zero.  (In Figure 1.3 they are - 27 + 97 - 70 = 0).  If one phase happens to be symmetrical (0% asymmetry) the other two, being displaced 120°, must both be partially asymmetrical.

The breaking capacities of circuit-breakers are always rated in kA (or MVA) ‘rms symmetrical’.  The peak asymmetrical current rating in kA may additionally be given.  All system fault calculations are made to determine the required rms symmetrical breaking current rating of the switchgear to be installed (see manual ‘Electrical Protection’).  From here on only symmetrical faults will be considered.

1.5       RESTRIKING VOLTAGE


As previously mentioned, the cause of the contact gap breaking down and the arc restriking after a current zero is the voltage which immediately reappears across the gap.  It is necessary now to look more closely at the nature of that voltage.

Figure 1.1 shows in simple form the steps in the interruption of an a.c. current supplying a load.



                                         
FIGURE 1.4

LOAD SHORT CIRCUITED


In the worst case the contacts might have to break the full fault current if the load were short-circuited; the contacts would then be interrupting a dead short-circuit across the generator, as shown in Figure 1.4.  This worst case will be considered first; also for clarity, discussion will be kept to single-phase circuits, as in Figure 1.4, although it applies equally to 3-phase circuits.


FIGURE 1.5

RESTRIKING VOLTAGE


Consider the idealised circuit at the top of Figure 1.5.  Supply is taken from an a.c. generator (50Hz or 60Hz) which is being short-circuited through the circuit-breaker contacts and whose current I the breaker has to interrupt.  The generator has internal reactance (and also some resistance, which is neglected here).  This reactance is represented by the series inductance L.  Figure 1.5 shows the breaker with its contacts separated but with the arc momentarily extinguished at a natural current zero.

The circuit also has a certain amount of self-capacitance, especially in any cable feeder system as well as in the generator itself.  Such capacitance is small and is of course distributed, but for simplicity suppose it to be ‘lumped’ as a capacitor C as shown in red in the figure.

At the onset of the fault the contacts are closed and are carrying the fault current I; they also short-circuit C which is therefore in a discharged state.  Even when the arc is established as the contacts begin to separate, C remains effectively short-circuited by the low-resistance arc and therefore discharged.

In Figure 1.5 the upper waveform shows the generator’s 50Hz or 60Hz alternating voltage.  Before the fault a small current was passing, but with the short-circuit a large current starts to flow, limited only by the generator’s reactance.  The fault current is therefore lagging almost 90° on the voltage, as shown in the middle waveform of Figure 1.5.

At the first current zero, which therefore occurs at a voltage peak, the arc momentarily goes out but immediately restrikes.  This may occur for several zeros, but eventually the gap becomes long enough to prevent a restrike, and the arc stays extinguished.

The short-circuit is now removed from C, which at that instant is still discharged.  Suppose the generator’s voltage peak at this instant of current zero was such that the positive and negative polarities were as shown in the figure.  Then the positive terminal of the generator, which now has C straight across it, will start to charge C, through L, to its own potential, the

charging current being i.  When C reaches the same potential as that of the generator, the charging current i is flowing strongly through the inductance L and cannot stop.  It continues to overcharge C until it reaches double the potential of the generator, by which time i has stopped.  But then the overcharged C starts to discharge through L back into the generator. Again i, now reversed, cannot stop, and C discharges again almost to zero.  Then the whole cycle repeats.

Thus there is an oscillating charge/discharge current between C and the generator which will continue until it is damped out by such resistance as is present.  (Electronics students will recognise this as a ‘ringing circuit’.)  On each ‘swing’ the charge voltage of C oscillates between zero and twice the generator peak voltage.  But the voltage across C is also the voltage between the open contacts of the breaker and is in fact a ‘restriking voltage’ appearing across the gap, trying to break it down and re-establish the arc.  It succeeds in doing so for several half-cycles, but eventually the gap becomes too long and the arc does not restrike; the full voltage oscillation then takes place.  After damping has caused the oscillation to die out, the voltage across C, and so across the open contact gap, settles down to the generator peak voltage.  This high-frequency voltage reappearing across the gap is shown in the lowest of the three waveforms of Figure 1.5.

The frequency of the oscillation is very high, chiefly because C is very small. It is given by


and typically may be of the order of 20kHz.  It is so fast that, by comparison, the 50Hz or 60Hz of the generator may be regarded as ‘d.c.’ and considered not to alter during the very short time of the oscillation.  At a frequency of 20kHz the time from current zero to the first peak of restriking voltage is only 25 ms - a very short time indeed.

This double-height, high-frequency restriking voltage appearing across a breaker when a large a.c. current is broken will travel along all connected conductors as a steep-fronted


FIGURE 1.6

EFFECT OF POWER FACTOR ON SWITCHING


voltage wave where it could, like a lightning strike, cause insulation problems to connected apparatus with high inductance, such as transformers.  This is the well-known ‘switching surge’ found in all electrical systems both large and small.

The high-frequency oscillating current will also radiate and may give rise to radio interference.

Up to this point only the worst case has been considered - that is, with the circuit-breaker clearing a generator short-circuit.  The current is then limited only by the generator reactance and therefore lags nearly 90° on the voltage as in Figure 1.6(a) - a very low power factor.

If the circuit-breaker, instead of clearing a short-circuit, were merely opening a normal load current where the load may have a high power factor, the switching situation would be much eased.

The current is of course lower, which means that there is less energy to dissipate.  But also the power factor is much higher (say 0.8), so that the current lags by only a small angle on the voltage (37° if pf = 0.8).  Consequently the current zero occurs not opposite a voltage peak but at a point well down the voltage curve, as shown in Figure 1.6(b).

After extinction of the arc at a current zero the discharged capacitance C recharges not from the generator peak voltage but from a voltage much less than this - see Figure 1.6(b).  The restriking voltage wave, which oscillates to twice the generator voltage at that moment, is therefore much smaller than in the low power-factor case.

Breaking a high power-factor current thus results in a considerably lower restriking voltage peak and a generally smaller restriking voltage wave.  The lower peak means less liability to restrike, and the breaker will therefore clear in fewer half-cycles of arcing.  The travelling restriking wave will also be less intense and is less likely to pose insulation problems on the system.

In short, therefore, the duty on a circuit-breaker is much eased if the power factor of the load to be broken is high.  Conversely it is increased with low power factors, that is with highly inductive loads (for example an open-circuited transformer) and especially with short-circuits.

The extra strain with inductive loads is well known in d.c. switching.  It occurs also with a.c. switching, but for a totally different reason, as explained above.

It is for this reason that testing authorities lay down that, when a circuit-breaker undergoes short-circuit tests, those tests shall be carried out at a power factor of not more than 0.15.

1.6       OIL CIRCUIT-BREAKERS (OCB)


Oil circuit-breakers are little used on offshore installations but are very common onshore.  They are usually confined to high-voltage systems.

An oil circuit-breaker consists of sets of breaking contacts for each phase totally immersed in oil, either in a common tank or in three separate tanks.  A mechanism, which may be mechanical, hydraulic or pneumatic in operation, drives all the moving contacts in unison to close the circuit.

It will be recalled from para. 1.5 that the time from a current zero to the first restriking voltage peak is of the order of 25 ms.  After an arc has gone out naturally at a current zero, it leaves behind in the gap a mass of vaporised and ionised oil of poor dielectric strength which the reappearing restriking voltage will easily break down.


To prevent a restrike it is necessary to flush out this polluted oil and oil vapour from the gap and to replace it with cool, fresh, clean oil - and all within a matter of microseconds.

Various oil circuit-breaker manufacturers have developed ingenious devices for doing this.  They are nearly all, however, variations of a basic concept called the ‘cross-jet explosion pot’.  This is an insulated, fire-proof pot which surrounds the contacts under the oil and is arranged with ports.  It uses the energy of the arc itself to build up internally a great pressure of oil which, at the instant the arc goes out at a current zero, is released and causes a clean, cool supply of unburnt oil to blast away the polluted oil through the ports.  The explosion pot, in one form or another, is a well-established arc control device for oil circuit-breakers.

After an oil circuit-breaker has cleared a fault it should be able to continue in service without attention.  But it is nevertheless customary to lower the tank at the first opportunity and examine the contacts for wear or burning, and also to replace the oil.

1.7       AIR BREAK CIRCUIT-BREAKERS (ACB)


Air-break circuit-breakers operate with their contacts in free air.  Their method of arc control however is entirely different from that of oil circuit-breakers, for it depends on the suppression of the restriking voltage. This gives them, as will be explained, a very different performance characteristic.  They are always used for low-voltage interruption and are now tending to replace high-voltage oil breakers up to 11kV and even higher, very largely because of their performance.  HV circuit-breakers on most offshore installations are of the air-break type.

Figure 1.7 illustrates the principle of air-break operation.  There are differences in detail between various manufacturers, and the method shown is only typical.

The fixed contact is on the left, and the moving contact assembly is hinged.  While the breaker is closed, the ‘main’ contacts (1) which carry the steady load current are held tightly closed by the operating mechanism or latch.


FIGURE 1.7
AIR BREAK CIRCUIT-BREAKER INTERRUPTION

When released by tripping, the hinged assembly moves to the right, first separating the main contacts across which an arc will form.

Under the electromagnetic forces due to the fixed-contact/arc/moving-contact current loop the arc is driven upwards.  During this stage it transfers from the main to separate ‘arcing contacts’ (2).  The main contacts are thereby relieved of any further burning.  Still being driven upwards the arc transfers to metal horns (3) on the base of a box called the ‘arc chute’.

This box is made of insulating and fire-proof material.  It is divided into many sections by barriers of the same material, as shown in Figure 1.7(a).  At the bottom of each barrier is a small metal conducting element between one side of the barrier and the other.

When the arc, driven upwards by the electromagnetic forces, enters the bottom of the chute, it is split into many sections by the barriers, but the metal pieces ensure electrical continuity between the arcs in each section; the several arcs are thus in series.

The electromagnetic forces within each section of the chute cause the arc in that section to take up the form of a helix, as shown in Figure 1.7(b).  All these helices are in series, so that the total length of the arc has been greatly extended, and its resistance is much increased.  This has the effect of reducing the current in the circuit.

Figure 1.7(a) shows the progress of the arc from the time it leaves the main contacts until it is within the arc chute.

When the current next ceases at a current zero, the ionised air in the path of where the arc had been is in parallel with the open contacts and acts as a shunt resistance across both the contacts and the self-capacitance C, shown in Figure 1.8 in red as a high resistance R.

When the oscillation starts between C and L as described for the oil circuit-breaker and shown in Figure 1.5, this resistance damps the oscillation heavily.  Indeed it is usually so


FIGURE 1.8
AIR BREAK CIRCUIT-BREAKER RESTRIKING VOLTAGE

heavy that the damping is ‘critical’; the oscillation cannot then take place at all, and the restriking voltage, instead of appearing as a high-frequency oscillation, rises ‘dead-beat’ to its eventual value of peak generator voltage.  This is shown on the lowest waveform of Figure 1.8, which should be compared with that of Figure 1.5.

The effect of this damping is important.  Because there is now no restriking voltage peak, and because there is no arc ‘residue’ across the small main gap, there is less likelihood of the contact gap restriking.  The air-break circuit-breaker’s performance is consequently good, in that it is able to break circuits in very few half-cycles.

The other great advantage is that, because there is no steep-fronted restriking voltage wave, there is virtually no switching surge to travel down the system and strain the insulation of connected apparatus.  There is, of course, practically no radio interference.

It is for these reasons that the air-break circuit-breaker is becoming increasingly attractive to system engineers, and design efforts are being directed to extending it to ever higher voltages and breaking capacities.

There is one feature about the air-break circuit-breaker which however should be mentioned.  Its operation depends upon the arc being driven electromagnetically upward from the main contacts into the chute.  Whereas with heavy currents these forces are more than adequate, with light currents they are weaker and the drive is less positive.  Air-break circuit-breakers are therefore usually fitted with a ‘puffer’.  This is a small air tube under each arc.  Air is compressed in a cylinder operated by the tripping mechanism and blows any reluctant arc upwards into the chute.  The puffer can be seen in Figure 1.7(a).

When breakers are tested for performance, tests are always made at light currents as well as with full rated current to prove this feature.

1.8       VACUUM CIRCUIT-BREAKER (VCB)


The vacuum circuit-breaker is becoming increasingly popular, especially in the medium ranges of voltage, because of its good performance and its compactness.

Its method of arc control differs from those of both the oil and the air-break circuit-breaker.  Whereas the oil breaker functions by flushing out the ‘combustion’ products of the arc, and the air-break type by suppressing the restriking voltage wave, the vacuum breaker operates by denying the arc any medium in which to re-form.

Figure 1.9 shows the elements of a vacuum interrupter, which is described in greater detail in Chapter 2.  It consists of a glass bottle evacuated to a very high degree; the vacuum area is shaded blue.  The contacts inside the vacuum space are of the butt type, and the moving contact, sealed through bellows, travels only a very short distance from the fixed contact when opening.  When the contacts first separate, normally in mid-cycle, an arc is drawn and the current continues to flow between them in the arc, supported by vaporised and ionised metal from the contacts.

At a current zero the arc momentarily ceases, and the vaporised metal instantly recondenses onto the contact surfaces, leaving no gas present in the vacuum chamber.  When the restriking voltage reappears across the open contacts there is no gas dielectric to break down and therefore no vehicle to support a new arc.  The arc consequently does not restrike.  It is claimed that vacuum interrupters will break an a.c. circuit in one to two half-cycles - quicker than any other type of breaker.

Because of the recondensing of the vaporized material onto the contact surfaces, there is very little net loss of contact metal over a large number of operations, and therefore very


FIGURE 1.9
VACUUM INTERRUPTION

little contact ‘wear’.  This renders them highly suitable for repeated use as contactors.  Such wear as there is is measured by feeler gauge on the operating linkage, and, when it reaches a certain level specified by the manufacturer, the whole vacuum interrupter element must be replaced.  Because the contact travel is so small it is recommended that all three bottles should be replaced and the new ones set up and aligned together.

1.9       OTHER TYPES OF CIRCUIT-BREAKER


Circuit interruption by oil, air-break and vacuum circuit-breakers has been dealt with in detail in this chapter, but these three do not exhaust the list.

Air-blast circuit-breakers are widely used onshore, but they are confined to extra high voltage and high breaking capacity transmission systems.  They function similarly to oil breakers except that the arc products are forcibly blasted away by the release of compressed air instead of by oil.

Another type is the ‘SF6’ breaker, which uses sulphur hexafluoride in place of oil.  The advantages of using this substance as an insulating and interrupting medium in circuit-breakers arise from its high dielectric strength and outstanding arc-quenching properties.  SF6 circuit-breakers are much smaller than air-break circuit-breakers of the same rating, the dielectric strength of SF6 at atmospheric pressure being equal to that of air at 10 atmospheres.  The SF6 decomposition products, discharged as a gas following extinction of the very hot arcs, are harmless, but they contain a small amount of fluorine which may react with metallic parts of the breaker.

Neither the air-blast nor the SF6 breaker is at present used in installations either offshore or onshore, and they are not further discussed here.

No comments: