Switching in Electrical Transmission and Distribution Systems

Preface

At the turn of the nineteenth century, a revolution took place in electrical engineering. In a rather short time, the transformer was invented, electric generators and motors were designed, and the step from DC to AC transmission was made. At the beginning of the twentieth century, the transmission voltages were steadily increased to reduce transmission losses. To improve operating efficiency, power systems began to be interconnected. Reserve power or spinning reserve could be then shared and capital expenditure could be reduced.

This is where power switching came in with its major task: isolating the faulted section of the system while keeping in service all healthy parts. Nowadays the power system can be regarded as one of the most complex systems ever designed, built and operated. Despite its complexity and robustness, the switching technology facilitates consumers to connect and disconnect electric loads in a rather simple and reliable way. Moreover, it protects the system from the effects of faults. However, this comes at a price since every change in the state of a system generates transients that may affect both the operating conditions of the system and its components.

With the first application of power switching in early electric systems, the development of standards for rating, testing and manufacturing high-voltage circuit-breakers began. In the United States, initiative was taken by a number of engineering and manufacturer trade organizations, such as the American Institute of Electrical Engineers (AIEE), dating back to 1884, later on merged into the Institute of Electrical and Electronics Engineers (IEEE) in 1963.

In Europe, the International Electrotechnical Commission (IEC) was founded in 1906, and the International Council on Large Electric Systems (CIGRE) started in 1921. CIGRE structured its organization by means of study committees in 1927. These study committees are responsible for the operation of working groups and task forces. Both collect field data and perform system studies, and their reports are used as input for creation and revision of IEC standards.

Over the years, many books and publications have been written on switching in electric transmission and distribution systems. A great deal of this knowledge results from the work of CIGRE and IEEE working groups, published as standards, technical brochures, reports and scientific papers.

For a utility engineer who wants to familiarize him- or herself with switching technology and the system aspects, the available literature is not easily accessible and sometimes difficult to comprehend. This book has been written to bridge the gap between the daily practice of utility engineers and the available literature.

The authors have served long periods in many working groups of CIGRE, IEC, and IEEE, and are familiar with switchgear manufacturing, power system aspects, and high power testing. In the relevant sections, contributions of the authors to CIGRE technical brochures, scientific literature, and standards are summarized.

The respective standards are referenced and run through the book like a continuous thread. This approach is intended to provide on the one hand guidance to the practical complexity of all essential switching operations and, on the other, a proper understanding of the standards and their background.

The book would not be complete without significant chapters on circuit-breakers and the switching media involved. Most of this material is gratefully taken from the book High Voltage Circuit Breakers by Professor Mirsad Kapetanović, issued by ETF – Faculty of Electrotechnical Engineering, Sarajevo (2011), commissioned by KEMA.

We acknowledge Zdeněk Matyáš M.Sc. who spent many hours harmonizing all our text, figures, and equations to be in line with the IEC standards and to achieve consistency in terminology.

Thanks are also due to Professor Viktor Kertész for the thorough checking of the mathematical sections, and we appreciate the contribution of Romain Thomas M.Sc. to the section on numerical simulation of transients.

We have dedicated our book to Professor Geert Christiaan Damstra (1930–2012), who has significantly contributed to the development of switching-, measurement-, and testing technology during his working life at the switchgear manufacturer Hazemeyer, KEMA, and the Eindhoven University of Technology. Being active and truly innovative in the high-voltage technique – literally to his last days – he has set an example to many.

René Smeets, Editor

Arnhem, Spring 2014

1

Switching in Power Systems

1.1 Introduction

As electricity comes out of AC outlets every day, and has done so for more than 100 years, it is nowadays considered a commodity. It is a versatile and clean source of energy; it is rather cheap and ‘always available’.

The purpose of a power system is to transport and distribute the electric energy generated in the power plants to the consumers in a safe and reliable way. Generators take care of the conversion of mechanical energy into electric energy, aluminium and copper conductors are used to carry the current, and transformers bring the electric energy to the appropriate voltage level. Society's dependence on this commodity has become extremely large and the social impact of a failing power system is unacceptable. The electrical power system is the backbone of modern society.

Switching operations in power systems are very common and must not jeopardize the system's reliability and safety. Switching in power systems is necessary for the following reasons and duties:

Taking into or out of service some sections of the system, certain loads, or consumers. A typical example is the switching of shunt capacitor banks or shunt reactors, de-energization of overhead lines, transformers, and so on. In industrial systems, this type of switching is by far the most common of all the switching operations.

Transferring the flow of energy from one circuit to another. Such operations occur when load current needs to be transferred without interruption from one busbar to another, for example, in a substation.

Isolating certain network components because of maintenance or replacement.

Isolating faulted sections of the network in order to avoid damage and/or system instability. The most well-known example of this is the interruption of a short-circuit current. Faults cannot be avoided, but adequate switching devices in combination with a protection system need to limit the consequences of faults.

Figure 1.1 provides an overview in orders of magnitude of the power switched in electrical-engineering applications.

Figure 1.1 Overview of the power being switched in electrical-engineering applications.

Switching in electrical power systems re-configures the topology of an electrical network; it involves the making and breaking of circuits and causes a disturbance of steady energy flow. Therefore, transients have to be expected and are observed in the system during the change from the situation before to the situation after switching. Transients are abnormal patterns of current and voltage that have a limited duration. Attention should be paid to these phenomena because they very often exceed the values met during steady-state operation. Fundamentally in nature, any change of steady-state conditions generates transients.

The essential parameters in electrical systems are current and voltage. During switching operations, transients can be observed in both. Regarding operations related to switching-on (making or energization), the components of the system are mainly stressed by current-related transients. On the other hand, at switching-off operations (breaking or de-energization), voltage-related transients will especially stress the switching device performing the operation.

In a generalized concept, switching devices (dis)connect a source circuit to a load circuit (see Figure 1.2). Both circuits are a complicated combination of system components: lines, cables, busbars, transformers, generators, and so on. Through reduction of the complexity to relevant simple electrical elements, either lumped or distributed where necessary (refer to Section 1.3), the switching transients can be more easily understood.

Figure 1.2 General concept of a switching device located between a source- and a load-circuit.

1.2 Organization of this Book

The aim of this book is to describe and explain to technically interested and practically oriented readers the variety of switching processes and devices in electrical power systems, avoiding (as far as possible) deep physical details and formal mathematics – although both of these are in fact pillars in the resolution of problems encountered in the high-power switching technology. Numerous examples of measurements and observed effects have been selected from realistic tests of a wide variety of switchgear at the KEMA High-Power Laboratory of DNV GL – Energy, where real service conditions are simulated in powerful test-laboratories (see Section 14.2).

The book is divided roughly into two parts. The first part (Chapters 1 to 5) focuses on the switching phenomena and their description.

The second part (Chapters 6 to 14) describes the technology of the devices that must perform switching in all experienced varieties and their impact on the power system.

In Chapter 1, the necessary background on the practical aspects of switching is given. The origin and role of the two key phenomena governing the switching processes are described: the switching arc and the transient recovery voltage (TRV). Due to the nature of the transients that accompany switching, the general description must, in principle, be in terms of electromagnetic fields and travelling waves. However, at sufficiently low frequency, a simplification of the relevant circuits in terms of lumped elements can greatly facilitate mathematical formulation and calculation of the TRV characteristics in the majority of the practical cases.

Chapter 2 deals exclusively with faults in power systems. Essential transients of fault-current events are identified, together with their impact on network components. Data on fault statistics are summarized.

In Chapter 3, the switching of fault currents, correctly termed the making and breaking operations, resulting from various types of faults in power systems, is analysed. Since, in this case, the TRV plays a crucial role, the description of the TRV is given in terms of simplified RLC circuits1 – either with lumped elements or in terms of travelling waves where necessary. A systematic approach is taken to represent every case by the minimum possible number of elements, in order to facilitate understanding.

Relevant characteristics of the switching arc are considered, showing its influence on the interruption process.

Chapter 4 deals with switching of loads: overhead lines, capacitor banks and shunt reactors operated under normal condition. Although the currents to cope with are much lower than at faults, it is explained that due to the reactive nature of capacitive and inductive loads, this type of switching is sometimes an onerous challenge. The main issue here is the description of transients generated by the switching process.

In Chapter 5, the calculation of the switching transients is treated. First, a formal analysis of the analytical solution of a few simplified circuits is given. Next, the background and the basics of some state-of-the-art numerical transient simulation programs are discussed.

Chapter 6 highlights the switching processes in gaseous media, such as air, the H2 gas in oil circuit-breakers and SF6 gas, the workhorse of present-day HV insulation and switching. Issues regarding health, safety and environment related to the ‘electrical’ use of SF6 gas are treated in detail.

Chapter 7 focuses on the SF6 circuit-breaker, the present-day technology of choice in HV systems. Technology and other relevant aspects of the circuit-breaker are described.

Chapter 8 describes the switching in vacuum, as presently applied on a large scale in distribution systems and slowly emerging in high-voltage systems as well.

Chapter 9 goes into the technology of the vacuum circuit-breaker. Recent information is added on the application of vacuum circuit-breakers for HV switchgear.

In Chapter 10, a variety of special switching situations and the technology of the appropriate devices are highlighted. A number of switching situations pose specific challenges to systems and switching devices. These are: generator circuit-breakers, switching with GIS- and air-break disconnectors, loop switching, switching in HV cable systems, the application of special by-pass switches in a series compensation capacitor bank, switching in the vicinity of shunt capacitor banks and switching in ultra-high-voltage (UHV) systems, high-speed earthing switches, direct current (DC) circuit-breakers and fuses.

Chapter 11 is devoted to switching overvoltages in systems. Practical values are given. Methods are discussed that enable reduction of the voltage stresses in particular situations. This is followed by an overview of controlled switching strategies that provide mitigation of unwanted transients.

Chapter 12 discusses the various investigations on reliability of circuit-breakers that have been conducted in the past 30 years. In addition, experiences regarding the endurance of switchgear against electrical and mechanical stresses are highlighted.

Chapter 13 deals with standardization and specification of switchgear. An explanation is given of the standardization framework for circuit-breakers that has been developed during the last half century in order to facilitate a system of quality assurance. In this chapter, as well as in other parts of the book, the worldwide accepted standardization system of the IEC (International Electrotechnical Commission)2 will be followed mostly.

Chapter 14 describes the backgrounds of testing methods for circuit-breakers. A detailed analysis is given of the various possibilities and practices of testing of the switching and breaking capabilities of circuit-breakers.

Throughout the book, extensive reference is made to documents of the CIGRE (Conférence International des Grands Réseaux Électriques or International Council on Large Electric Systems)3, a non-profit association for promoting collaboration with experts from all around the world by sharing knowledge and joining forces to improve the electric power systems of today and tomorrow. CIGRE has over 12 000 members from the electrical energy industry. More than 3500 experts from all around the world work actively together in structured Working Groups (WGs) coordinated by the CIGRE Study Committees (SCs). Their main objectives are to design and deploy the power system for the future, optimize existing equipment and power systems, respect the environment and facilitate access to information. For switching in power systems and its impact, the relevant study committees are A3 (High-Voltage Equipment), B4 (HVDC and Power Electronics), B5 (Protection and Automation) and C4 (System Technical Performance).

CIGRE documents can be accessed through www.e-cigre.org. A large volume of information is laid down in Technical Brochures (TB), the output documents of working groups, and in CIGRE meeting papers.

Another major source of reference is the IEEE (Institute of Electrical and Electronics Engineers). With more than 425 000 members IEEE is the world's largest professional association dedicated to advancing technological innovation.

IEEE has a number of societies. The one that is most closely related to the scope of this book is the PES (Power and Energy Society) that covers planning, R&D, design, construction and operation of facilities systems for generation, transmission and distribution of electric energy. PES comprises 30 000 industry professionals, academics and students with a common interest in the electric power industry. It provides the world's largest forum for sharing the latest in technological developments in the electric power industry, for developing standards and for education.

Other societies related to switching technology are: the IAS (Industry Applications Society); the DEIS (Dielectrics and Insulation Society) that deals with insulation materials and systems, dielectric phenomena and discharges in vacuum, gaseous, liquid and solid electrical materials; the IEEE PELS (Power Electronics Society), amongst others, on the development and practical application of power electronics technology; the NPSS (Nuclear and Plasma Sciences Society) that covers, amongst others, discharges in power switching devices.

IEEE documents are collected in the IEEE Xplore® Digital Library containing more than 3 million documents from IEEE and IEEE journals, transactions, magazines, letters, conference proceedings and active IEEE standards.

Within IEEE, the IEEE-SA (IEEE Standards Association) is developing and nurturing standards (see Chapter 13).

1.3 Power-System Analysis

Power-system analysis is a broad subject, too broad to be covered by a single textbook. Textbooks about the fundamentals of power-system analysis [1–4] give an overview of the structure of power systems (from the generation of electric energy, to the transmission and distribution to the customers) and take only the systems steady-state behaviour into account. This means that only the power-frequency phenomena are considered.

An interesting aspect of power systems is that the modelling of the system depends on the time scale that is being considered. In general, the time scales of interest are:

Years, months, weeks, days, hours, minutes for steady-state analysis at power frequency (50 or 60) Hz. This is the time scale on which textbooks on the fundamentals of power-system analysis focus. The steady-state analysis covers a variety of topics, such as planning, design, economic optimisation, load flow / power flow computation, fault calculation, state estimation, protection, stability and steady-state control.

Seconds for dynamic-behaviour analysis. The dynamic behaviour of electrical networks and their components is important in order to predict whether the system, or a major part of it, remains in a stable state after a disturbance, for example, after a switching operation occurring at initiation or removal of a fault. The stability of power systems depends particularly on the ability of the installed control equipment to damp the electromechanical disturbances of the synchronous generators.

Tens of microseconds to milliseconds for transients related to switching (kilohertz to tens of kilohertz). The insight into the transient behaviour of systems is important for understanding the effects of switching events (i.e. connection/disconnection of loads or fault clearing).

Microseconds or less for disturbances having a disruptive origin (tens of kilohertz to several megahertz) like the effects of atmospheric disturbances (lightning strokes), breakdown phenomena causing excessive voltages and currents. Physically, the impact of these fast transients is mostly limited to the part of the system where the disturbance originates, that is, most often the affected part or component of the system itself and its immediate vicinity.

Although the power system itself remains unchanged when different time scales are considered, the components in the power system should be modelled in accordance with the appropriate time frame.

An illustrating example is the modelling of an overhead transmission line. For steady-state consideration at the power frequency of 50 Hz, the wavelength of the sinusoidal voltages and currents is 6000 km:

(1.1)

Thus, the transmission line is essentially of electrically small dimensions compared with the wavelength of the voltage or current. The generally valid Maxwell equations can therefore be approximated considering a quasi-static approach, and the transmission line can be rather accurately modelled by lumped elements. Kirchhoff's laws can be fruitfully used to compute the voltages and currents.

In contrast, when the effects of a lightning stroke have to be analysed, frequencies of 1 MHz and higher occur, and the typical wavelength of the voltage and current wave is 300 m or less. In this case, the transmission line is far from being electrically small, and it is no longer justified to use the lumped-element approximation. The distributed nature of the parameters of the transmission lines has to be taken into account, and has to be dealt with by travelling waves [5, 6].

Despite the fact that mainly lumped-element models are used in modelling, it is important to realize that the energy is mainly stored in the electromagnetic field surrounding the conductors with almost none in the conductors themselves. The Poynting vector, being the vector product of the electric field intensity vector E and the magnetic field intensity vector H, indicates the direction and intensity of the electromagnetic power flow.

(1.2) numbered Display Equation

Due to the finite conductivity of the conductor material and the finite permeability of the transformer-core material, a small electric field component is present inside the conductor and a small magnetic field component results in the transformer core:

(1.3) numbered Display Equation

(1.4) numbered Display Equation

When one speaks of electricity, one thinks of current flowing through the conductors from generator to load. This approach is valid because the physical dimensions of the power systems are large compared with the wavelength of the currents and voltages. For steady-state analysis of the power flow at the power frequency 50 or 60 Hz, complex calculus with phasors representing voltages and currents can be used successfully. Switching transients, however, involve much higher frequencies, up to kilohertz and megahertz, and the complex calculus can no longer be applied. Now the differential equations describing the system phenomena have to be solved. In addition, lumped-element modelling of the system components has to be done with care if Kirchhoff's voltage and current laws are used.

In the case of a power transformer under normal power-frequency conditions, the transformer ratio is given by the ratio of the number of turns of the primary and the secondary winding. However, for a lightning-induced voltage wave or a fast switching transient, the stray capacitance of the windings and the stray capacitance between the primary and secondary coil determine the transformer ratio. In these two situations, the power transformer has to be modelled differently. When one cannot get away with a lumped-element representation, wherein the inductance represents the magnetic field, the capacitance represents the electric field, and the resistance represents the losses, travelling wave analysis must be used. The correct ‘translation’ of the physical power system and its components into suitable models for the analysis and calculation of power-system transients requires insight into the basic physical phenomena. Therefore, it requires careful consideration and is not easy [7].

A transient occurs in the power system when the network changes from one steady state into another. This can be, for instance, the case when lightning hits the earth in the vicinity of a HV transmission line or when lightning hits a substation directly.

The majority of power-system transients are, however, the result of switching actions. Load-break switches and disconnectors switch off and on parts of the network under load and no-load conditions, respectively. Fuses and circuit-breakers interrupt higher currents and clear short-circuit currents in faulted parts of the system. The time period when transient voltage and current oscillations occur is in the range of microseconds to milliseconds. On this time scale, the presence of a short-circuit current during a system fault can be regarded as a steady-state situation, wherein the energy is mainly in the magnetic fields, and, after the fault current interruption, the system is transferred into another steady-state situation, wherein the energy is predominantly in the electric fields.

1.4 Purpose of Switching

1.4.1 Isolation and Earthing

Isolation of components from energized sections of the system is the simplest (no-load) switching operation. Isolation is usually necessary for safe maintenance, repair, and replacement of power-system components. Only after isolation and earthing, can personnel approach the equipment. In many countries, a visible break between live and workable parts is required.

To reduce the probability of breakdown to the absolute minimum, a very large contact distance, to be achieved with the switching device, is necessary. Such switching devices are commonly called disconnectors or disconnecting switches. These devices can operate in open air (such as in outdoor substations) or in an SF6 (sulfur hexafluoride) environment, such as in gas-insulated switchgear (GIS) where the conductors and switchgear are insulated by pressurized SF6 gas contained in metal tubes.

The no-load switching, that is, only isolation from energized sections, might seem to be an easy, straightforward operation. Nevertheless, due to the stray capacitance of the power system, a very small current always flows in energized systems. Because of this, disconnector switching is also associated with the extinction of an electric arc (see Section 1.5). Disconnector switching is discussed in detail in Section 10.3.2.

Earthing is the switching operation that connects a previously live part of the system to earth. In normal earthing operation, the section to be earthed is de-energized. In a faulty situation, when earthing is performed with energized sections or components of the power system, large currents can result, depending on earthing of the neutral of the power system. In any case, earthing switches must be capable of conducting the fault current, while special fast- and high-speed earthing switches have to perform the earthing operation under all (including faulty) conditions, see Section 10.4.2.

1.4.2 Busbar-Transfer Switching

For reliable operation of power systems many components and connections are installed in a redundant way. Busbars in substations are usually in a double arrangement. In cases when the flow of current has to be maintained but diverted (or commutated) from one busbar to another, switching devices, such as disconnectors are used to transfer the load current to the parallel busbar. Thus, the net load current will continue to flow uninterrupted. Because of the presence of the parallel busbar, current transfer up to a significant load current is relatively straightforward.

Bus transfer switching is treated in detail in Section 10.3.3.

1.4.3 Load Switching

Loads are regularly switched in power systems. For industrial systems, contactors are designed to switch normal loads, such as motors, pumps, furnaces, and so on, very frequently. In utility power systems, load-break switches and circuit switchers are the devices that can interrupt the load current – but not the (full) fault current. The frequency of normal-load switching in utility systems is usually very low. This is not the case for reactive-power installations, such as shunt capacitors and shunt reactors that are switched very frequently, often twice daily.

Unlike normal loads that mostly have a power factor close to one, shunt-reactor currents and capacitive currents have a phase angle of 90 electrical degrees between current and voltage. This has severe implication for the switching of these devices, as will be explained in Sections 4.2 and 4.3. Reactors can store energy in their magnetic field and capacitors store electric charge, the energy of which is released when the de-energization operation fails. The release of this energy in the system may have detrimental effects for the installed switchgear and other components.

1.4.4 Fault-Current Interruption

When a fault occurs in the power system, the associated short-circuit current is detected by protective relays which initiate circuit-breaker operation in order to interrupt the fault current (see also Section 2.1). The event is also known as fault clearing. The protective relays continuously monitor currents and voltages, collecting the information from the instrument transformers, that is, current and voltage transformers.

The time between the occurrence of a fault and its detection by the protection system, the relay time, is usually of the order of one to three half-cycles of the power-frequency of 50 or 60 Hz. The protection system issues a tripping command to the circuit-breaker(s) that should isolate the faulted section from the rest of the network. The tripping command activates the operating mechanism and through its kinematic chain makes the contacts in the circuit-breaker separate. After a certain opening time, the circuit-breaker arcing contacts will open in all three poles; this is usually referred to as contact parting or contact separation.

The pole of a circuit-breaker, or more generally of a switching device, is the part of the device that is located in one of the phases of the network, so there are three poles in a three-phase device.4 A switching device is called single-pole if it has only one pole. If it has more than one pole, it may be called multi-pole (two-pole, three-pole, etc.) provided the poles are or can be coupled in such a manner as to operate together. The part of the pole in which the actual current is to be interrupted is generally called the interrupter or interruption chamber. It consists of contact system(s), a mechanical device supporting the arc-extinction process, and insulation. Depending on the rated voltage, a pole can consist of two or more interrupters placed in series in order to share the voltage. Grading capacitors across each interrupter have to take care of an equal voltage distribution across each interrupter.

So, a circuit-breaker may be designed as three single-pole switching devices or as a three-pole switching device and each pole will contain one or more interrupters. A three-pole device will be equipped with a single operating mechanism while single-pole devices will have one operating mechanism per pole or, at the highest system voltages, even several operating mechanisms per pole.

The electric arc in a circuit-breaker plays a key role in the interruption process and is therefore often called a switching arc. Upon contact separation, an arc is formed in the interrupter(s) of each pole. Actual interruption must wait for a zero crossing of the current. The arc is in essence resistive and therefore the arc voltage and the current reach the zero crossing at the same instant. Around current zero (see Section 1.5), the energy input in the arc channel is rather low (at current zero there is even no energy input), and if the circuit-breaker design is such that the cooling by the extinction medium is adequate, the current can be interrupted. Depending on the type of circuit-breaker, the device may not be ready to interrupt at the first occurring current zero after contact separation. It takes a certain minimum arcing time before the circuit-breaker can actually interrupt the current, because sufficient cooling pressure of the extinction medium must be available and/or sufficient contact distance must be reached.

Then, after the minimum arcing time has elapsed, the current can be interrupted at the first following current zero. Current interruption will take place at the respective current zero of each of the three phases. When all the poles have interrupted, the fault has been cleared. The time between the instant of energizing the trip coil of the circuit-breaker and the current interruption in all phases is called the break time.

All relevant times are displayed in Figure 1.3 showing a three-phase-fault interruption sequence in an effectively earthed neutral system (see Section 3.3.2).

Figure 1.3 Circuit-breaker opening in a three-phase circuit and the IEC definitions of relevant times.

In the standard IEEE C37.04 [8], the rated interrupting time (i.e. the time between energization of the trip circuit and interruption in all phases) is expressed by the number of power-frequency cycles. A three-cycle breaker thus needs three power-frequency cycles to clear a fault.

1.5 The Switching Arc

When current flows through a circuit-breaker and the contacts of the breaker separate, the current continues to flow through the arc that starts at contact separation. Just before contact separation, the circuit-breaker contacts touch each other at a very small surface area, the contact bridge, and the resulting high current density makes the contact material melt. The melting contact material virtually explodes and this leads to a gas discharge in the surrounding medium, that is, air, oil, or SF6.

The matter changes from a solid state to a liquid state. When more energy is added and the temperature increases, the matter changes from a liquid state to a gaseous state. A further increase in temperature gives the individual molecules so much energy that they dissociate into separate atoms, and if the thermal energy level is increased even further, the electrons in the outer shell(s) of the atoms acquire sufficient energy to become free electrons, leaving positive ions behind. The mixture of free electrons and ions is called the plasma state: a state of matter in which a certain portion of the particles is ionized. Because of the presence of free electrons and the heavier positive ions in the high-temperature plasma channel, the plasma channel is highly conductive and the current continues to flow through the arc plasma after contact separation. The electric arc is the plasma channel between the circuit-breaker contacts, a high-current electrical discharge in the extinction medium.

When considering current interruption, it is important to realize that an electric arc is always drawn at contact separation, and it appears immediately, automatically, and inevitably.

The electric arc is the only known element, apart from power semiconductors, that is able to change from a conducting to a non-conducting state in a short period. In HV circuit-breakers, the electric arc is a high-pressure arc burning in oil, air, or SF6 (see Chapter 6). In medium-voltage (MV) circuit-breakers, the arc exists in vacuum or, more correctly, in the metal vapour released from the contacts (see Chapter 8).

Current interruption is performed by cooling the arc plasma so that it disappears at its most critical period of existence around the current zero.

Interruption of a short-circuit current is a very important function of a circuit-breaker. This function is verified in an extensive system of test-duties, set up by standardizing bodies, such as IEC and IEEE (see Section 13.1).

To understand the inevitability but also the advantage of an electric arc, consider a simple 50 Hz circuit with the r.m.s. inductive current I = 100 A at line-to-line voltage Ur = 10 kV (Figure 1.4).

Figure 1.4 Equivalent circuit diagram for inductive-current interruption.

Assume the hypothetical case that this current is interrupted immediately at contact separation without an arc at an instantaneous ‘chopped’ value of ich = 100 A as depicted in Figure 1.5.

Figure 1.5 Current and voltage at current interruption without arc.

The value of the inductance L can be calculated straightforwardly from the r.m.s. values of the current and voltage:

(1.5) numbered Display Equation

At the moment of interruption, the instantaneous voltage across the load uC is given by:

(1.6)

In practice, the load reactance will always possess a certain stray capacitance Cs, which is taken in this example as a lumped stray capacitance Cs = 5 nF. The stray capacitance represents in fact the energy storage in the electric field caused by the voltage at the terminals of the load. At current interruption, this capacitance remains charged at the voltage uC. Now the situation is that the reactor stores magnetic energy because of the current ich, and the capacitor stores electric energy because of its voltage uC. In terms of energy, the magnetic energy stored in the disconnected load reactor is:

(1.7) numbered Display Equation

whereas the electric energy stored in the capacitor is:

(1.8) numbered Display Equation

Since the circuit-breaker is assumed to have disconnected the load from the source, the stored energy cannot transfer back to the source and remains in the load. The energy in the reactor and in the capacitor varies widely, and there is an exchange of the energy between the two components tending to reach a state of balance. The exchange results in an oscillation. As a consequence, the total energy Em + Ee at a certain moment will be present only in the capacitor with zero current and energy in the reactor. At that moment, the voltage across the capacitor uC,max can be calculated as:

(1.9) numbered Display Equation

from which it results that uC,max = 607 kV.

This is of course a rather unrealistic situation since 607 kV is a voltage unimaginable in a 10 kV system. The voltage cannot reach this value because a breakdown will occur at the dielectrically weakest location in the circuit. This location is the contact gap of the switching device.

The arc (electric arc discharge) will start because conducting plasma is created by the melting, evaporation and ionization of the metal originating from the last contact bridge through which all current is passing (see Section 1.4.4). Every time when the arc has a tendency to extinguish, for example when the current is too small to maintain the arc, there is still substantial energy trapped in the reactor and, consequently, the voltage appearing immediately across the contact gap will re-ignite the arc and re-create the conducting arc channel.

Therefore, in an inductive circuit, the arc is a continuously surviving discharge, lasting until the magnetic energy stored in the load reactor is released back to the source. Only at the instant of current zero (i = 0) is there no magnetic energy in the reactor and the arc disappears.

This shows the big advantage of the arc interruption over a sudden interruption: the arc allows a natural transfer of load energy back to the source, thus avoiding excessive overvoltages.

In AC systems, current zero crossings occur every half-cycle and all HV AC power switching devices interrupt the current at one of those current zeros.

Switching arcs are normally not visible in HV switching devices because they appear in a hermetically sealed interrupter. In simpler switching devices, however, the arc and its consequences can be observed. Figure 1.6 shows a switching arc in a load-break switch (see Section 1.4.3). In this example, a current of 700 A is interrupted in a 12 kV test circuit, which is a normal action at load-current switching. The impact of the arc can be seen: the pressure rise in the interruption chamber between open contacts in the atmospheric air causes a plasma jet to eject ionized gas to the outside, together with the debris from molten material of the contacts and the interruption-chamber walls. The interruption principle of confining the arc into a chamber with walls that release evaporation material that contributes to the cooling of the arc is sometimes referred to as the deion principle (see Section 6.2.3).

Figure 1.6 Switching arc of a load-break switch interrupting 700 A in a 12 kV test circuit.

From this example it is clear that arcs created in the largest single-interrupter circuit-breakers capable of breaking 63 kA in 550 kV systems, that is, having 4125 times greater apparent switching power than the load-break switch, are hugely violent phenomena.

Although current zero is the only opportunity for a switching device to interrupt a current, this does not imply that every current interruption is finally successful. The arc being present between the contacts may have disappeared, but the hot remnants, for example, ionized gases in SF6 circuit-breakers and metal vapour in vacuum breakers, will reduce the dielectric strength, thus influencing the ability of the circuit-breaker to withstand the transient recovery voltage (TRV). The transient recovery voltage is the voltage that appears across the gap immediately following current interruption, as a reaction of the network to the new situation. A re-ignition may occur followed by another loop of power-frequency current. Eventually, after several unsuccessful attempts, the device may not be capable of interrupting the current and will explode, causing a short-circuit by itself.

1.6 Transient Recovery Voltage (TRV)

1.6.1 TRV Description

The TRV is the voltage across the open circuit-breaker contacts immediately after current interruption. The TRV, that is, uab, appears as the difference between the voltage-to-earth at the source side and the voltage ubn at the load side (see Figure 1.9):

(1.10) numbered Display Equation

Thus, a TRV always consists of two components: a source-side component uan and a load-side component ubn. In all cases, the TRV starts from zero at current zero, makes an excursion to the momentary power-frequency voltage, overshoots in a damped oscillatory manner, and continues to oscillate until a steady-state condition is reached. This steady-state situation is a power-frequency voltage, called the recovery voltage (RV).

Figure 1.7 Current-interruption in a purely inductive AC circuit.

Figure 1.8 Current zero and transient recovery voltage in an inductive AC circuit.

Figure 1.9 Equivalent circuit diagram for multi-frequency TRV.

The complete interruption process is outlined schematically in Figure 1.7. In this case the RV is equal to the source voltage.

The frequency of the TRV is determined by the relevant inductance and capacitance. The peak value in the undamped case is two times the peak source voltage; in practice, the peak value of the TRV is lower due to damping.

The TRV affects the interruption for two reasons:

Determined by the oscillation frequency, the rate-of-rise of recovery voltage (RRRV) can be very high. This implies that very shortly after extinction of the arc, a high voltage appears across the contact gap. If there are still ionized, hot residues of the arc remaining to a certain degree, the arc will be re-established (will re-ignite) due to the impact of the TRV. In Section 13.1.2 it will be described that the standardized RRRV corresponds to the slope of the tangential line of the TRV wave shape, a value not necessarily equal to the highest value of the derivative of the TRV (du/dt), see Figure 1.8.

The peak value of the TRV can be very high. In testing and standardization, the damping is expressed by the amplitude factorkaf, defined as the ratio between the transient peak value and the steady-state value; in Figure 1.8 the steady state voltage is the peak of the power-frequency voltage. The value of kaf is in the range 1 < kaf ≤ 2.

Representative values of kaf are very difficult to calculate, because resistances depend strongly on frequency. Due to the skin-effect, the effective conduction takes place in a surface layer, the thickness of which is smaller at higher frequencies, leading to a rapid increase in damping resistance at higher frequencies.

The technology of circuit-breakers must be able to:

Withstand the high thermal energy of the arc before current zero.

Rapidly remove the remnants of the arc after current zero in order to withstand the TRV. In gas/oil breakers, this is achieved by a forced gas flow through the former arc path, removing the ionized medium. In vacuum interrupters, the natural diffusion of the metal vapour plasma towards the very low background pressure enables fast recovery of the gap.

Background information on the awareness of TRV over the years and its standardized description can be found in Section 13.1.1.

Circuits subjected to a short-circuit are mainly inductive. This is because the current value is limited by the reactance (X = ωL), rather than by the resistance R. In other words, R ≪ ωL. This causes the short-circuit current to lag approximately 90° with respect to voltage. In 50 Hz circuits the ratio is standardized: X/R = 14.14 and the phase lag of the current with respect to the voltage is 85.9° (electrical degrees). In medium voltage cable networks the phase lag is smaller, which results in a less onerous interruption regarding TRV.

1.6.2 TRV Composed of Load- and Source-Side Contributions

In practical cases, there will be more parts of the network involved in the TRV wave shape than just the single LC elements of the circuit described in the example above. Consequently, the TRV often comprises multiple-frequency components decisive for its RRRV and peak value.

As an example, the case is considered of a fault some distance away from the circuit-breaker. A simplified single-phase equivalent circuit is shown in Figure 1.9. It comprises two separate LC circuits, one with CS and LS at the source side and one with CL and LL at the load side. After interruption of the current – which is lower than the terminal-fault current because of the load-side impedance – the two parts of the circuit are disconnected completely and have no electrical interaction.

The transients in both circuit parts behave independently and, in the construction of TRV, understanding of the initial, intermediate, and final conditions is very helpful:

Initial condition (at current zero):

Both transients start at the same voltage level:

(1.11) numbered Display Equation

This is the voltage to earth at current zero; in the simple-circuit case it is the voltage across the capacitors CS and CL that are charged to equal voltage at current zero.

Oscillation interval from current zero to the decay of the transients after which only power-frequency recovery voltage remains:

The TRV component at the source side uan has an amplitude

(1.12)

The TRV component at the load side ubn has an amplitude

(1.13)

Final condition after decay of transient components:

The TRV component at the source side uan(t) will oscillate from the initial voltage uan(0) to the power-frequency voltage of the source Û·cos(ωt), ignoring the voltage drop across the capacitance, because 1/(ωCS) > ωLS.

The TRV component at the load side ubn(t) will oscillate from the initial voltage ubn(0) to zero (in the absence of an active source and/or a charge-storing element).

Keeping these simple and clear rules in mind, understanding TRV in various situations is straightforward.

Considering the fact that in practical cases the frequency of the source- and load-side TRV component is much higher than the power frequency, the transient components can be treated independently from the power-frequency voltage.

In practical cases, in circuits with power factor close to zero, the equation of the TRV reads as follows:

(1.14)

with βS and βL being the damping in the source- and load-side circuit:

(1.15) numbered Display Equation

and ω0S = 2π fS and ω0L = 2π fL the respective angular frequencies.

The formal mathematical derivation of this generalized TRV equation can be found in Section 5.1.3.

Because of the presence of two frequencies, there is now a first local maximum value u1 and a global maximum uc.

The effect of the double-frequency nature of the TRV is twofold:

The peak value uc is reduced (with respect to a single-frequency TRV) because the amplitude of each of the oscillations is smaller than Û;

The rate-of-rise of TRV (RRRV) may increase because of a second component having a higher frequency.

In Figure 1.10, the transients at the load and source sides are given for different load and source impedances (LS = 0.5 LL). Both transients contribute to the TRV. As can be seen in this example, the initial rate-of-rise of TRV (RRRV) is determined by the TRV component of the highest frequency (here the source side), whereas its peak value will usually be determined by the oscillation having the largest amplitude (the load side in Figure 1.9). This is normally the circuit part experiencing the largest voltage drop.

Figure 1.10 Double-frequency TRV with the source-side and load-side components.

A clear distinction can be made between a fault- and load-current interruption as detailed in Chapters 3 and 4 respectively:

In fault-current breaking, the current to be interrupted is mainly determined by the source impedance: XS ≫ XL, as shown in Figure 1.11. Then, the TRV amplitude is mainly contributed by the source-side circuit (u0,S ≫ u0,L).

In load switching, the current is clearly determined by the load impedance, thus XS ≪ XL and u0,S ≪ u0,L, leading to the situation in Figure 1.12 where the load oscillation is