Circuit breakers

Introduction

For the analysis of simple switching transients and for carrying out large system studies, it is often sufficient to model a circuit breaker as an ideal switch. When studying arc–circuit interaction, wherein, the influence of the electric arc on the system elements is of importance, a thorough knowledge about the physical processes between the circuit breaker contacts is absolutely necessary.

A high-voltage circuit breaker is an indispensable piece of equipment in the power system. The main task of a circuit breaker is to interrupt fault currents and to isolate faulted parts of the system. Besides short-circuit currents, a circuit breaker must also be able to interrupt a wide variety of other currents at system voltage such as capacitive currents, small inductive currents, and load currents. We require the following from a circuit breaker:

  • In closed position it is a good conductor;
  • In open position it behaves as a good isolator between system parts;
  • It changes in a very short period of time from close to open;
  • It does not cause overvoltages during switching;
  • It does not cause overvoltages during switching;

The electric arc is, except from power semiconductors, the only known element that is able to change from a conducting to a nonconducting state in a short period of time. In high-voltage circuit breakers, the electric arc is a high-pressure arc burning in oil, air, or sulphur hexafluoride (SF6). In medium-voltage breakers more often, the low-pressure arc burning in vacuum is applied to interrupt the current. The current interruption is performed by cooling the arc plasma so that the electric arc, which is formed between the breaker contacts after contact separation, disappears. This cooling process or arc-extinguishing can be done in different ways. Power circuit breakers are categorised according to the extinguishing medium in the interrupting chamber in which the arc is formed. That is the reason why we speak of oil, air-blast, SF6, and vacuum circuit breakers.

In 1907, the first oil circuit breaker was patented by J. N. Kelman in the United States. The equipment was hardly more than a pair of contacts submersed in a tank filled with oil. It was the time of discovery by experiments and most of the breaker design was done by trial and error in the power system itself. In 1956, the basic patent on circuit breakers employing SF6 was issued to T. E. Browne, F. J. Lingal, and A. P. Strom. Presently the majority of the high-voltage circuit breakers use SF6 as extinguishing medium.

J. Slepian has done much to clarify the nature of the circuit breaker problem, because the electric arc proved to be a highly intractable and complex phenomenon. Each new refinement in experimental technique threw up more theoretical problems. The practical development of circuit breakers was, especially in the beginning, somewhat pragmatic, and design was rarely possible as deduction from scientific principles. A lot of devel- opment testing was necessary in the high-power laboratory. A great step forward in understanding arc–circuit interaction was made in 1939 when A. M. Cassie published the paper with his well-known equation for the dynamics of the arc and then in 1943 Otto Mayr followed with the supplement that takes care of the time interval around current zero. Much work was done afterwards to refine the mathematics of those equations and to confirm their physical validity through practical measurements. It becomes clear that current interruption by an electrical arc is a complex physical process when we realise that the interruption process takes place in microseconds, the plasma temperature in the high-current region is more the 10 000 K, and the temperature decay around current zero is about 2000 K/μs per microsecond while the gas movements are supersonic.

The understanding of the current interruption process has led to SF6 circuit breakers capable of interrupting 63 kA at 550 kV with a single interrupting element.

The switching arc

The electric arc in a circuit breaker plays the key role in the interruption process and is therefore often addressed as switching arc. The electric rc is a plasma channel between the breaker contacts formed after a gas discharge in the extinguishing medium. When a current flows through a circuit breaker and the contacts of the breaker part, driven by the mechanism, the magnetic energy stored in the inductances of the power system forces the current to flow. Just before contact separation, the breaker contacts touch each other at a very small surface area and the resulting high current density makes the contact material to melt. The melting contact material virtually explodes and this leads to a gas discharge in the surrounding medium either air, oil, or SF6.

When the molecular kinetic energy exceeds the combination energy, matter changes from a solid state into a liquid state. When more energy is added by an increase in temperature and the Van der Waals forces are overcome, matter changes from a liquid state into a gaseous state. A further increase in temperature gives the individual molecules so much energy that they dissociate into separate atoms, and if the energy level is increased even further, orbital electrons of the atoms dissociate into free moving electrons, leaving positive ions. This is called the plasma state. Because of the free electrons and the heavier positive ions in the high- temperature plasma channel, the plasma channel is highly conducting and the current continues to flow after contact separation.

Nitrogen, the main component of air, dissociates into separate atoms (N2 → 2N) at approximately 5000 K and ionises (N → N+ + e) above 8000 K. SF6 dissociates into sulphur atoms and fluorine atoms at approx- imately 1800 K and ionises at temperatures between 5000 and 6000 K. For higher temperatures, the conductivity increases rapidly. The thermal ionisation, as a result of the high temperatures in the electric arc, is caused by collisions between the fast-moving electrons and photons, the slower- moving positively charged ions and the neutral atoms. At the same time, there is also a recombination process when electrons and positively charged ions recombine to a neutral atom. When there is a thermal equilibrium, the rate of ionisation is in balance with the rate of recombination.

The relation between the gas pressure P, the temperature T, and the fraction of the atoms that is ionised f is given by Saha’s equation with

(2)

With e = 1.6 ∗ 10−19 the charge of an electron;Vi = potential of the gaseous medium and k = 1.38 ∗ 10−23 Boltzmann’s constant

Saha’s relation is shown in graphical form for oxygen, hydrogen, and nitrogen and for the metal vapours of copper and mercury in Figure 1

Oil circuit breakers

Circuit breakers built in the beginning of the twentieth century were mainly oil circuit breakers. In those days, the breaking capacity of oil circuit breakers was sufficient to meet the required short-circuit level in the substations. Presently, oil and minimum-oil circuit breakers still do their job in various parts of the world, but they have left the scene of circuit breaker development.

The first oil circuit breakers were of simple design – an air switch that was put in a tank filled with mineral oil. These oil circuit breakers were of the plain-break type, which means that they were not equipped with any sort of arc quenching device. In 1901, Joseph N. Kelman of the United States built an oil–water circuit breaker in this way, which is capable of interrupting 200–300 A at 40 kV. Kelman’s breaker   consisted of two open wooden barrels, each containing a plain-break switch. The two switches were connected in series and operated by one common handle. The wooden barrels contained a mixture of water and oil as extinguishing medium.

In the 1930s, the arcing chamber appeared on stage. The breaker, a metal explosion pot of some form, was fitted with an insulating arcing chamber through which the breaker contacts moved. The arcing chamber, filled with oil, fixes the arc, and the increase in pressure inside the arcing chamber improved the cooling effects on the arc considerably. Later, the design of the arcing chamber was further improved by pumping mechanisms, creating a cross flow of oil, giving extra cooling to the arc.

A next step in the development of oil circuit breakers was the minimum-oil circuit breaker. The contacts and arcing chamber were placed into a porcelain insulator instead of in a bulky metal tank. Bulk-oil circuit breakers with their huge metal tank containing hundreds of liters of mineral oil have been popular in the United States. Minimum-oil circuit breakers conquered the market in Europe.

Air-blast circuit breakers

Air is used as insulator in outdoor-type substations and for high-voltage transmission lines. Air can also be used as extinguishing medium for current interruption. At atmospheric pressure, the interrupting capability, however, is limited to low-voltage and medium-voltage only.

For medium-voltage applications up to 50 kV, the breakers are mainly of the magnetic air-blast type in which the arc is blown into a segmented compartment by the magnetic field generated by the fault current. In this way, the arc length, the arc voltage, and the surface of the arc column are increased. The arc voltage decreases the fault current, and the larger arc column surface improves the cooling of the arc channel.

At higher pressure, air has much more cooling power, and air-blast breakers operating with compressed air can interrupt higher currents at considerable higher voltage levels. Air-blast breakers using compressed air can be of the axial-blast or the cross-blast type.

The cross-blast type air-blast breaker operates similar to the magnetic-type breaker: compressed air blows the arc into a segmented arc-chute compartment. Because the arc voltage increases with the arc length, this is also called high-resistance interruption; it has the disadvantage that the energy dissipated during the interruption process is rather high. In the axial-blast design, the arc is cooled in axial direction by the airflow. The current is interrupted when the ionization level is brought down around current zero. Because the arc voltage hardly increases, this is called low-resistance interruption.

When operating, air-blast breakers make a lot of noise, especially when the arc is cooled in the free air, as is the case with AEG’s free-jet breaker (Freistrahlschalter) design.

Vacuum circuit breakers

Between the contacts of a vacuum circuit breaker, a vacuum arc takes care of the interruption process. The vacuum arc differs from the high-pressure arc because the arc burns in vacuum in the absence of an extinguishing medium. The behavior of the physical processes in the arc column of a vacuum arc is to be understood as a metal surface phenomenon rather than a phenomenon in an insulating medium. The first experiments with vacuum interrupters took place already in 1926, but it lasted until the 1960s when metallurgical developments made it possible to manufacture gas-free electrodes and when the first practical interrupters were built.

There are no mechanical ways to cool the vacuum arc, and the only possibility to influence the arc channel is by means of interaction with a magnetic field. The vacuum arc is the result of a metalapor/ion/electron emission phenomenon. To avoid uneven erosion of the surface of the arcing contacts (especially the surface of the cathode), the arc should be kept diffused or in a spiraling motion. The latter can be achieved by making slits in the arcing contacts or by applying horseshoe magnets as used in the vacuum interrupters by Eaton Holec.

There is generally less energy required to separate the contacts of a vacuum circuit breaker, and the design of the operating mechanism usually results in reliable and maintenance-free breakers. Vacuum breakers are produced for system voltages up to 72.5 kV, and the short-circuit current rating goes up to 31.5 kA. When the arc current goes to zero, it does so in discrete steps of a few amperes to 10 A, depending on the contact material. For the last current step to zero, this can cause a noticeable chopping of the current. This current chopping in its turn can cause high overvoltages, in particular when the vacuum breaker interrupts a small inductive current, for example, when switching unloaded transformers or stalled motors.

The mechanism of vacuum arc extinction

The current-interruption process in a Vacuum Circuit Breaker (VCB) is done by a metal-vapor arc, which is more commonly known as a vacuum arc [3]. This arc appears as soon as the breaker’s contacts separate, and it continues to exist until its energy input ceases. In an AC-network, the current’s value runs periodically through zero, and each current zero provides the breaker with an opportunity to quench the arc, because here, its energy input is temporarily zero. The breaker’s resistance changes rapidly from almost zero to almost infinity, and as a result, a Transient Recovery Voltage (TRV) builds up across the breaker after current zero.

Explaining the phenomena observed in electrical measurements, and modelling a VCB’s electrical behavior, requires knowledge of the physics behind the vacuum arc’s extinction. This chapter summarizes the results of this research, as found in literature. It starts with describing the aspects of the vacuum arc, and the events after its extinction. After that, it describes the control of the arc, which is required to extend a breaker’s technical life-time and to improve the interruption process. Finally, it explains the mechanisms that sometimes lead to a failure to withstand the recovery voltage, and reignite the vacuum arc, hence being unsuccessful to interrupt the current.

The vacuum arc

The vacuum arc that exists between the contacts of a VCB can generally be divided into three regions [4]. These are the cathode spot region, being the main source that provides material to the vacuum arc, the inter-electrode region, and the space- charge sheath in front of the anode (see Figure 2). Contrary to the schematic representation of the vacuum arc in Figure 1a, the cathode spot region and the anode sheath region are very small in relation to the length of the inter-electrode space. These regions have typically a constant thickness of several micrometers, whereas the rest of the arc is inter-electrode plasma.

Despite its small size, the cathode spot region covers most of the arc’s voltage uarc   (see Figure2b), and it is a typical feature of the vacuum arc that the voltage across this region remains practically constant, independent from the value of the current. This voltage depends predominantly on the type of material that is used for the breaker’s contacts. For example, for copper-based contacts, which is the main component in all commercial VCB’s, the voltage is about 16 V.

The slightly increasing voltage across the inter-electrode region is mainly due to Ohmic losses, but the voltage drop in front of the anode region is characteristic for the interaction between a plasma and a metal surface. It occurs not only at the anode, but it is also observed at other metal surface, such as the metal vapor shields.

The vacuum arc ceases to exist when its sources, the cathode spots, have disappeared. However, it takes time for the residual plasma to disappear, and the metal vapor that is still present between the contacts after the arc’s extinction. The remaining charge has still some conductance, which leads to a post-arc current when a TRV starts to build up across the gap. Although the actual arc has vanished, in addition to the vacuum arc properties, this section describes the post-arc current phenomena as well

Cathode spots

Cathode spots are observed as tiny bright spots moving across the surface. The observed light arises actually from an ionization zone in front of the cathode, see Figure 3. Between this ionization zone and the cathode, an ionic space-charge sheath is present, in which electrons from the cathode are accelerated to collide with metal vapor, and ionize it. Both the distance between the spot and the cathode, as well as the diameter of the ionization zone measure just several micrometers. Although the dimensions involved with the cathode spot region are small compared to the total vacuum arc, it takes up almost all of the arc voltage. A consequence of the small dimensions of a cathode spot is that other physical quantities, such as the current density and the electric field, are high. This turns out to be not only a consequence, but also a necessity for the spot’s survival [5, 6, 7].

A number of different processes control the electron emission from the cathode. First, there is thermal emission. When a metal is heated, an increasing number of electrons is able to escape spontaneously from the metal’s conduction band into the ambient. The current density for a metal with the temperature of a cathode spot (about 4000 K), lies in the range of 107 A/m2.

Another method for extracting electrons from a metal is by field emission, also known as Fowler-Nordheim tunneling. When an electric field is applied to a metal in vacuum, some electrons inside the metal are able to tunnel from their conduction band through the potential barrier in front of the cathode, into the surrounding space. According to this theory, the current density that results from the electric field near a cathode spot reaches a value of up to 108 A/m2 [8].

The individual contribution of these two processes is insufficient to account for the measured cathode spot current density, with values as high as 1013 A/m2 [9]. However, when the processes are combined, the total current density is not just the sum, but a product of the separate processes. The result of this mechanism, which is appropriately called Thermal-Field (TF) emission, corresponds well to the measured cathode spot current density.

The rise of the surface-temperature required for TF emission under a cathode spot is mainly caused by ohmic heating by the electrical current, and by ion bombardment. The latter process is the result of ions, accelerated in the electric field inside the sheath towards the cathode. These processes generate much more heat than the metal can conduct, and hence it evaporates in an explosive way, ejecting metal vapor and droplets of liquid metal into the gap.

The surface-temperature rise at this scale can only be reached with a high current density for Joule heating, and a high electric field for the accelerations of ions inside the sheath. For a constant current and voltage, the current density and electric field are simply increased by decreasing the spatial dimensions of the cathode spot to zero. However, if the spot becomes too small in diameter, the crater produces too little vapor to ionize, and destabilizes the equilibrium. This is an argument for an increasing spot size, and as a result, the cathode spot reaches a size that optimally satisfies all the requirements

Cathode spots move across the contact’s surface. This movement is strongly related to the presence of surface irregularities, such as micro-protrusions or crater rims. These irregularities enhance the electric field, resulting in an improved location for TF emission. The random distribution of irregularities across the surface is the main reason for the cathode spot’s erratic motion. However, when a cathode spot is subjected to a magnetic field B⃗, it moves in the direction of − (I⃗ × B⃗), i.e. in the opposite direction of the Lorenz force. Apparently, this force is small in relation to other processes, and although many models have been proposed in the past to explain this phenomenon (which is called retrograde motion), a final theory for this has not yet been found.

The retrograde motion determines the movement of multiple cathode spots with respect to each other as well. When the current increases, a cathode spot does not simply continue to increase its size, but it separates into two or more spots over which the total current is distributed. The current at which this happens depends mostly on the contact material, and for copper contacts, cathode spots have a maximum current in the range of 50-100 A.

The great number of models and theories on cathode spots that have been pro- posed in the past, and continue to be published nowadays, are an indication that a conclusive model has not yet been found. Finding an improved model for cathode spots is beyond the scope of this research, however basic knowledge about it might help to understand, for example, how the vacuum arc ignites, in the case of a breakdown.

Inter-electrode plasma

The majority of the ions created in the ionization zone in front of the cathode (see Figure 3) return to the cathode to bombard its surface. However, a fraction of ions is launched towards the anode, thus moving in an opposite direction of the electric current. They do this with a kinetic energy that even exceeds the corresponding arc voltage. For lower currents, their energy reaches values as high as 120 eV, while the arc voltage normally does not exceed 16 V. This effect is believed to be caused by a combination of three mechanisms [10, 11].

The first is related to a ’potential hump’ in front of the cathode. The space charge causes the potential to rise locally to a much higher value than uarc, and in the resulting electric field, ions are accelerated towards the opposite direction of the electrical current.

Another force that drives ions in the opposite direction of the current is caused by a pressure gradient. The pressure near cathode spots can rise to atmospheric values, only to decrease to a low value a little further away from the cathode. The resulting pressure gradient is strong enough to force ions to move towards the anode.

The third mechanism, which is believed to deliver the greatest contribution, is electron-ion friction. In the constricted space of the ionization zone, the kinetic energy of electrons is not only used to ionize metal vapor, but also to exchange momentum with the ions.

The density inside the inter-electrode plasma is low, which gives the charge a high mobility. Electrons cross the gap without losing much energy from colliding with ions or neutrals, and hence the plasma’s conductivity is high. As a result, the electric field remains low, and ions move towards the anode without experiencing much resistance from it. In vacuum arcs of copper-based contacts, about eight percent of the electrical current consists of ion current, which is fully compensated by the electron current.

Anode sheath

Particles in a vapor move with a random velocity determined by their temperature. This creates a pressure that exhibits a force to the walls of the vapor’s container. Most of the particles that collide with a metal object, such as the contacts or the vapor shield of a vacuum tube, are removed from the plasma, as they are either absorbed by the metal (electrons), or neutralized by electrons from the metal (ions). In that way, the metal acts as a sink for plasma.

Since their thermal energy is higher and their mass is lower, electrons have a higher thermal velocity than ions. Because of this, the flux of electrons at a metal boundary would be larger than the flux of ions, which results in an electrical current. An electric field in front of the electrodes repels the surplus of electrons to maintain a net charge flux of zero. This explains the electric potential difference between the plasma and the anode, which is mainly distributed across the small ionic space- charge sheath in front of the anode (see Figure 2).

The electric field of a singly charged particle in space stretches to infinity, but in the presence of particles with charge of opposite polarity, the spatial influence of its electric field is finite. The distance of the electric field’s influence is expressed as the Debye length, and is slightly longer than the average distance between particles. In general, a plasma is called neutral when its size is several orders larger than its Debye length. The neutralizing effect of the space charge on the electric field also affects the size of the sheath in front of a metal object, and the anodic space-charge sheath thickness is therefore only several Debye lengths.

Arc control

When the arc current surpasses a certain threshold, the arc constricts towards the anode because of the electromagnetic forces. As a result, the current density at the anode’s surface concentrates in a single spot, which is called the anode spot. The energy involved can cause this spot to melt and produce metal vapour, which in turn is partly ionized by incident electrons. At that point, the anode has changed from a passive charge collector to a new source of charge.

An anode spot differs from a cathode spot in a sense that all the current is concentrated in this single, stationary spot, and that it takes more time to cool down after arc extinction. After current zero, the former anode is bombarded by ions from the residual plasma, which are accelerated in the electric field of the TRV. The former anode becomes the new cathode, and with its increased surface temperature and an increased amount of vapor in

front of it, the conditions for cathode spot formation, and hence the creation of a new vacuum arc, are severely improved. Therefore, the anode spot is not only destructive for the contacts, thereby limiting the breaker’s technical life-time, but it also increases the probability of a reignition.

Eventually, the TRV causes the reignition of a VCB, but anode spots enhance the conditions for it, and they mainly determine the current interruption limit of a breaker. The most useful techniques that are used by manufacturers to increase the current interruption limit are described below.

Contact diameter

As a result of their retrograde cathode spots move towards the edge of the cathode, thereby maximizing the area of charge production. The charge has to travel towards the center of the anode for arc constriction. By increasing the contact’s size, the distance between the cathode’s edge and the anode’s center increases, and hence, arc constriction takes place at a higher current level.

Increasing the contact diameter to obtain higher rated short-circuit current ratings for VCB’s is favorable. However, there are some disadvantages that limit the use of larger contacts. One of them is that it requires larger ceramic envelopes, which makes them more expensive, and another disadvantage is that larger contacts increase the probability of having a surface irregularity that enhances the electric field when a voltage is applied, which can lead to a re-strike. The latter effect is called the surface effect.

Magnetic field

Nowadays, almost all the vacuum interrupters have a mechanism that generates a special magnetic field between their contacts. The short-circuit current itself generates this field, as the contacts are constructed in a way that the current spirals through them. Depending on the contact configuration, the current generates a magnetic field that is either parallel to the interrupter’s axis (Axial Magnetic Field, or AMF [12, 13]) or radial to it (Radial Magnetic Field, or RMF [14]). Figure 4 depicts two examples of such contact types.

Both types of magnetic fields are intended to relieve the thermal stress on the anode. In an AMF, the arc maintains its diffuse state at higher currents, while an RMF allows the formation of an anode spot, but forces it to rotate quickly across the anode, thereby limiting the average thermal stress on it.

In general, breakers with AMF type contacts allow for higher currents than breakers with RMF type contacts, since the absence of an anode spot results in less vapor release. However, AMF type contacts are more complicated to manufacture than RMF type contacts, which as a consequence makes their production more expensive.

Vapor shield

Vapor shields do not actually control the arc, but they increase the technical life- time of vacuum interrupters. They envelope the contacts to prevent metal vapor, released during arcing to attach to the ceramic envelope. In this way, it prevents the formation of a conducting path along the inner side of the ceramic enclosure that eventually may short out the breaker, which would render the interrupting device useless.

The metal vapor shield acts also as a sink for charge. As a result, the shield’s configuration has an influence on the breaker’s electrical behavior. This is most clearly seen in the post-arc current. Shields with a small diameter allow less charge to be present, and drain it more quickly than shields with a larger diameter. As a result, the post-arc current’s magnitude and duration are proportional to the shield’s diameter [15, 16]. Therefore, it is beneficial for the breaker’s recovery to have shields with a small diameter, but a minimum value is also required to prevent a breakdown via the shield.

Shields are also applied to protect other vulnerable parts of the breaker, such as the metal bellows, but their influence on the breaker’s electrical behavior is negligible.

The post-arc current and recovery voltage

When the arc current approaches zero, the number of cathode spots reduces until only one is left. This spot continues to supply charge to the plasma, until finally, the current reaches zero. At current zero, the inter-electrode space still contains a certain amount of conductive charge. As the current reverses polarity, the old anode becomes the new cathode, but in the absence of cathode spots, the overall breaker’s conductance has dropped, which allows the rise of a TRV across the VCB.

The combination of the residual plasma’s conductance and the TRV gives rise to a post-arc current, which, depending on the arcing conditions and on the TRV can reach a peak value of several milliamperes to several tens of amperes. The post-arc current has been subject to investigation for many years, since it is one of the most distinctive electrical features of short-circuit current interruption with VCB’s [14, 17, 18, 19, 20, 21]. Because the post-arc current shows a clear dependence on the arcing conditions, it is a reasonable assumption that it reflects the conditions inside the breaker immediately after current zero. This would provide researchers with a tool to investigate the interruption performance without having to damage the VCB to look inside. However, the post-arc current unfortunately contains a considerable scatter that disturbs the relationship between the arcing conditions and the post-arc current. This has to do with the final position of the last cathode spot which, till now, can only be determined by looking inside the breaker.

When the final position of the last cathode spot is near the edge of the cathode, a significant amount of charge is ejected away from the contacts, and disappears, e.g. by recombination at the breaker’s vapor shield. If this is the case, less charge is returned to the external electrical circuit by means of a post-arc current, compared to the situation in which the final cathode spot’s position is close to the center of the cathode (see Figure 5). Since a cathode spot moves randomly across the cathode surface (but is biased by an external magnetic field), its final position is unknown. As a result, the post-arc plasma conditions are different for each measurement, which gives the post-arc current its random nature.

Nevertheless, some general conclusions about the shape and intensity of the post- arc current can be drawn. For instance, that its peak value and duration increase with an increasing value of the short-circuit current and with an increasing arcing time.With regard to its mechanism, the generally accepted theory divides the post-arc current into three Phases, which are described below.

During arcing (before current zero), ions are launched from the cathode towards the anode. At current zero, the ions that have just been produced continue to move towards the anode as a result of their inertia. Electrons are much lighter than ions, and it can be readily assumed that they adapt their speed immediately to a change of the electric field. As a result, the electrons match their velocity with the ion velocity to compensate for the ion current, and this makes the total electrical current zero.

We now enter phase 1. Immediately after current zero, the electrons reduce their velocity, and a net flux of positive charge arrives at the post-arc cathode. This process continues until the electrons reverse their direction, and until this moment, the net charge inside the gap is zero. With no charge and a high conductivity of the neutral plasma, the voltage across the gap remains zero in this phase.

As soon as the electrons reverse their direction, the post-arc current enters its second phase. In this phase, the electrons move away from the cathode, leaving an ionic space charge sheath behind, see Figure 2.6. Now, the gap between the electrodes is not neutral any more, and the circuit forces a TRV across it. This potential difference stands almost completely across the sheath, which, contrary to the plasma, is not (charge) neutral.

Initially, the plasma connects the vapors shield surrounding the contacts electrically to the cathode, as in Figure 7a [19, 22]. As a result, the distribution of the electric field in the sheath does not change much as the sheath continues to grow, since the vapor shield maintains the cathode’s potential, see Figure 7b. However, after a while, the metal vapor shield becomes disconnected from the electrical circuit as the sheath progresses towards the anode (see Figure 7c). This process changes the electrical configuration of the vacuum chamber drastically, and it is frequently observed that at this moment, the post-arc current shows a distinctive drop towards zero. Measurements performed by others on the vapor shield’s potential in the post-arc phase confirm this theory [23].

The sheath continues to expand into the inter-electrode gap until it reaches the new anode. At that moment, the post-arc current reaches its third phase. The electrical current drops, since all electrons have been removed from the gap. The electric field between the contacts moves the remaining ions towards the cathode, but the current that results from this process is negligible.

Failure mechanisms

A failure occurs when a breaker is unable to withstand a voltage after current interruption, and a new arc is formed, through which the short-circuit current continues to flow. Knowing the nature of a failure makes it easier for developers to prevent it from happening, and nowadays, commercially available VCB’s are well capable of reliably interrupting currents according to their current and voltage rating.

If nevertheless the breaker does not interrupt the current at the first current zero after contact separation, it most likely interrupts the current at the next current zero. The reason for this is that the conditions for a successful current interruption at the following current zero have improved. Although it is difficult to investigate the conditions in a hermetically sealed breaker, two examples can be given that make this assumption probable.

The first has to do with the arcing time. The average contact separation speed of a VCB is about 1 m/s. This means that when the contacts start to separate just an instant before the first current zero, the contacts have not reached their maximum separation at current zero. Since the breakdown voltage increases proportionally with the contact separation, a breakdown at a voltage below the breaker’s rated voltage can occur. In the next current loop, the contacts continue to separate to reach its maximum at the following current zero and interrupt the current correctly.

Another example has to do with the electrical properties of the external circuit. A short-circuit network has a dominant inductive nature. That means that when the short-circuit occurs at an instant other than the instant of a maximum voltage, a DC component adds to the short-circuit current. This DC component decays due to resistive elements in the circuit, and as a result, the short-circuit current can be divided in successive major and minor loops (see Figure 9). The breaker’s current- interruption conditions after a major loop are worse than after a minor loop. Hence, it is likely that if a breaker fails to interrupt the current after a major loop, it can still break the current at the next current zero, after the minor loop.

In general, events of continuation of arcing after initially successful interruption are divided into two different types, depending on the time after current zero at which they occur. One is called dielectric breakdown, which happens sometime after current zero, and the other is called thermal breakdown, which is the breakdown type that occurs almost immediately after current zero, when residual charge and vapour are still present in abundance between the contacts [24, 25].

Dielectric re-strike

The Fowler Nordheim tunneling is the mechanism that draws electrons from a metal surface by means of an electric field. The resulting current density causes locally Joule heating of a contact, increasing its temperature locally, which may eventually lead to the formation of a cathode spot, and subsequently to a re-strike. Since such a failure is initiated by an electric field, it is called (dielectric) re-strike. This term is used for describing the failure a VCB, at an instant when the presence of residual vapor from a vacuum arc is unlikely. Such a failure occurs, for example, several milliseconds after current zero.

For perfectly smooth contact surfaces, the electric field in a VCB under normal operating conditions is generally too low to cause dielectric re-strike. However, irregularities are widely present on the contact surface, which locally enhance the electric field (see Figure 10a). This concentrates the current density to smaller surface areas, which increases the ohmic heating locally, and thus creates the conditions for a re-strike.

Each time that a protrusion melts due to Joule heating, surface tensions of the liquid metal smooths this protrusion out, thereby improving the contact’s dielectric property. Because of the great number of microscopic protrusions on new contacts, a VCB initially starts with a lower breakdown voltage, which increases to an asymptotic maximum value after a series of re-ignitions, because with each breakdown, one or more protrusions are removed. Manufacturers make use of this principle, and increase a VCB’s breakdown strength by applying an AC low-current arc for some time to new VCB’s. This technique is known as surface conditioning.

The re-strike is not only enhanced by irregularities on the contact’s surface, but also by microscopic particles in the vacuum [26, 27]. Although manufacturers spend much effort on cleaning the interior of a VCB, the presence of these particles is inevitable. They originate, for example, from protrusions on the contact surface, which are drawn from it under the influence of an electric field, but they can also be left-overs from a vacuum arc, which are not properly fused to the contacts or the shield after the arc’s extinction.

There are several ways how a micro particle can contribute to a re-strike. For example, when in the vicinity of a contact, it enhances the electric field as indicated in Figure 10b, which may lead to a similar current-density concentration as described earlier for a surface irregularity. Another way is that the incident electron current on the particle increases its temperature, and eventually vaporizes it, causing an improved scenario for a re-strike.

If a particle is charged, it accelerates in the electric field and collides with a contact. The transfer of kinetic energy can cause the release of vapor, or even charge. With the surface deformation from the impact, new protrusions are formed that enhance the electric field (see Figure 10c). All these mechanisms contribute to a higher probability for a re-strike.Field emission of electrons, the release of vapor and charge and other processes are enhanced when the temperature of the contacts increases. After current zero, the residual charge and vapor decays within microseconds and milliseconds, respectively, but cooling of the contacts takes a considerably longer time. Especially in the case that an anode spot has been active during arcing, a pool of liquid metal on the (former) anode is likely to remain present for a considerable time after current zero, and it thereby enhances the conditions for a dielectric re-strike.

Reignition

A thermal reignition occurs when a VCB fails in the period immediately following current zero. This term is originally used for gas circuit breakers, where the probability of having a breakdown depends on the balance between forced cooling and Joule heating of the residual charge in the hot gas between the contacts. This process differs strongly from thermal reignition in vacuum, where charge and vapor densities are much lower than in gas breakers.

As described before, immediately after current zero, the gap contains charge and vapor from the arc, and the contacts are still hot, and can also have pools of hot liquid metal on their surface. It takes several microseconds for the charge to remove (by diffusion and by the post-arc current), but it takes several milliseconds for the vapor to diffuse, and the pools to cool down [25, 28]. When a failure occurs when vapor is still present, but charge has already decayed, this is called dielectric reignition. With the increased contact temperature, the conditions for a failure of the type are improved, but the increased pressure might also cause another process, called Townsend breakdown [29, 30, 31].

When an electron, accelerated in the electric field, hits a neutral vapor particle with sufficient momentum, it knocks out an electron from the neutral. This process reduces the kinetic energy of the first electron, but from here, both electrons accelerate in the electric field, hit other neutral particles and cause an avalanche of electrons in the gap, which eventually causes reignition.

This process of charge multiplication enhances when the probability of an electron hitting a vapor particle increases. This can be achieved by either increasing the vapor pressure, or by increasing the gap length. Both methods reduce the reignition voltage, but at some point, electrons collide with particles before reaching the appropriate ionization energy. As a result, after reaching a minimum value, the reignition voltage eventually rises with increasing vapor pressure or gap length. Figure 11 depicts the relation between the vapor pressure and the breakdown voltage at constant gap length. Such a graph is called a Paschen Curve.

The Townsend breakdown theory is based on a stationary vapor, in which the electric field is more or less equally distributed. This differs strongly with the situation between VCB contacts immediately after current zero. Here, the charge distribution is definitely not equal, and in some regions of the gap, ions still have a considerable drift velocity. This makes the determination whether or not the observed reignition of the VCB resulted from Townsend breakdown particularly difficult.

In addition to Townsend breakdown, the increased anode temperature improves the conditions for extracting electrons from it. Moreover, the TRV not only extracts electrons from the anode, it also launches ions towards it, a process which further increases the anode’s temperature [32]. This might eventually lead to a failure that is similar to dielectric re-strike, but since it occurs during the post-arc current, it is still called thermal reignition.