Electrical Engineer’s Guide--Fuses and protection relays

1. Protection and co-ordination

Fuse and protection relays are specialized devices for ensuring the safety of personnel working with electrical systems and for preventing damage due to various types of faults. Common applications include protection against overcurrents, short-circuits, overvoltage and undervoltage.

The main hazard arising from sustained overcurrent is damage to conductors, equipment or the source of supply by overheating, possible leading to fire. A short-circuit may melt a conductor, resulting in arcing and the possibility of fire; the high electromechanical forces associated with a short-circuit also cause mechanical stresses which can result in severe damage. A heavy short-circuit may also cause an explosion.

Rapid disconnection of overcurrents and short-circuits is therefore vital. An important parameter in the design and selection of protective devices is the prospective current; this is the current which would flow at a particular point in an electrical system if a short-circuit of negligible impedance were applied. The prospective current can be determined by calculation if the system impedance or fault capacity at that point is known.

In addition, personnel working with electrical equipment and systems must be protected from electric shock. A shock hazard exists when a dangerous voltage difference is sustained between two exposed conducting surfaces which could be touched simultaneously by different parts of the body. This voltage normally arises between earth and metalwork which is unexpectedly made live and if contact is made between the two through the body, a current to earth is caused. Fuses can provide protection where there is a low-resistance path to earth, because a high current flows and blows the fuse rapidly. To detect a wide range of currents flowing to earth it is necessary to use current transformers and core-balance systems: these operate a protective device, for example a circuit breaker such as the rcd described in section 3.4 for low-voltage systems.

When designing an electrical protection system it is also necessary to consider co- ordination so that when a fault occurs, the minimum section of the system around the fault is disconnected. This is particularly important where disconnection has safety implications. for instance in a hospital. An illustration of co-ordination is shown in Fig. 1.

Fig. 1 Example of co-ordination.

Fig. 2 Time-current characteristics for a fuse and a protection relay controlling a circuit breaker

Protective devices are described by a time-current characteristic. In order to achieve co-ordination between protective devices, their time-current characteristics must be sufficiently separated, as shown in Fig. 2, so that a fault downstream of both of them operates only the device nearest to the fault. A variety of shapes of time-current characteristic for both fuses and protection relays are available for different applications.

Another consideration when designing electrical protection is back-up. In some circuits it is desirable to have a device such as an mcb which disconnects lower overcurrents and can be reset. Higher overcurrents, which the mcb cannot disconnect without being damaged, are disconnected by an upstream fuse in series with the mcb. This arrangement has the advantage that reclosure onto a severe fault is less likely, because replacement of the fuse would be necessary in this case as well as reclosure of the mcb. It is also possible for a protective device to fail, for example because of mishandling. The effects of a failure can be minimized or avoided by a back-up protective device which operates under such conditions. In Fig. 1, protective device M backs up each of the protective devices P. However, if M operates as back-up device, co-ordination is lost because all four branches of the circuit lose supply, not just the faulty branch.

8-2 Fuses

2.1 Principles of design and operation

A fuse consists of a replaceable part (the fuselink) and a fuse holder. Examples of fuse holders are shown in Fig. 3.

The simplest fuselink is a length of wire. It is mounted by screw connections in a holder which partly encloses it. When an overcurrent or short-circuit current flows, the wire starts to melt and arcing commences at various positions along it. The arc voltage causes the current to fall and once it has fallen to zero, the arcs are extinguished.

The larger the wire cross-section, the larger is the current that the fuselink will carry without operating. In the UK, fuses of this type are specified for use at voltages up to 250 V and currents up to 100 A. They are known as semi-enclosed or rewireable fuses.

The most common fuselink is the cartridge type. This consists of a barrel (usually of ceramic) containing one or more elements which are connected at each end to caps fitted over the ends of the barrel. The arrangement is shown in Figs 8.4 and 8.5. If a high current breaking capacity is required the cartridge is filled with sand of high chemical purity and controlled grain size. The entire fuselink is replaced after the fuse has operated and a fault has been disconnected. Cartridge fuses are used for a much wider range of voltages and currents than semi-enclosed fuses.

Fuselinks can be divided into current-limiting and non-current-limiting types. A sand-filled cartridge fuselink is of the current-limiting type; when it operates it limits the peak current to a value which is substantially lower than the prospective current.

A non-current-limiting fuse, such as a semi-enclosed fuse, does not limit the current significantly.

Fig. 3 Fuse holders for miniature and compact low-voltage fuses: (a) fuse holders for miniature fuse- links; (b) Miniature fuselink (reproduced with permission from Littelfuse); (c) Fuse holders for compact low-voltage fuselinks; (d) Compact low-voltage fuselink; (e) Rewirable fuse (not to the same scale)

The element shown in Fig. 4 is a notched tape. Melting occurs first at the notches when an overcurrent flows and this results in a number of controlled arcs in series. The voltage across each arc contributes to the total voltage across the fuse, and this total voltage results in the current falling to zero. Because the number of arcs is limited, the fuselink voltage should not be high enough to cause damage elsewhere in the circuit. The characteristic development of current and voltage during the operation of a fuse is shown in Fig. 6.

The function of the sand is to absorb energy from the arcs and to assist in quenching them; when a high current is disconnected, the sand around the arcs is melted.

The element is usually of silver because of its resistance to oxidation. Oxidation of the element in service would affect the current that could be carried without melting, because the effective cross-section of the element is changed. Silver-plated copper elements are also used.

Fig. 4 Cross-section through a low-voltage cartridge fuselink

Fig. 5 High-voltage and low-voltage cartridge fuselinks (high-voltage fuse (top) and two low-voltage fuses showing a range of fuse sizes)

Many elements include an m-effect blob, which can be deposited on wire (Fig. 3(b)) or notched tape. The blob is of solder-type alloy which has a much lower melting point than the element. If a current flows which is large enough to melt only the m-effect blob, the solder diffuses into the silver. This creates a higher local resistance in the element and the fuse operates at a lower current than it would have done in the absence of the blob.

Other types include the expulsion fuse which is used at high voltage, and the Universal Modular Fuse (UMF) which is used on Printed Circuit Boards (PCBs). Fuses offer long life without deterioration in their characteristics or performance, and cartridge fuses have the particular advantage that they contain the arc products completely.

2.2 Rating principles and properties

(a) Current ratings (ZEC)

The rated current of a fuse is the maximum current that a fuselink will carry indefinitely without deterioration. In the case of ac ratings, an rms symmetrical value is given.

The current rating printed on a fuselink applies only at temperatures below a particular value. Derating may be necessary at high ambient temperatures and where fuses are mounted in hot locations such as an enclosure with other heat-generating equipment.

Fig. 6 Current and voltage during the operation of a fuse

(b) Voltage ratings (IEC)

The rated voltage of a fuse is the nominal voltage for which it was designed. Fuselinks will perform satisfactorily at lower voltages, but at much lower voltages, the reduction in current caused by the resistance of the fuselink should be considered. In the case of ac ratings, the rms symmetrical value is given, and for dc ratings the mean value, including ripple, is given.

IEC recommendations are moving towards harmonized low-voltage ac supplies of 230,400 and 690 V, but although the nominal voltage is being changed in many countries it will be possible for the voltage to remain at its previous non-harmonized level for several years. In Europe, the nominal voltage is 230 V, and the permitted variations will allow supplies to remain at 240 V and 220 V. Fuse-links marked 230 V may have been designed originally for use with higher or lower voltages, and problems may therefore arise when replacing fuselinks because a device manufactured for use at 220 V would not be safe to use on a 240 V system. A fuselink designed for 240 V could safely be used at 220 V. Similar considerations apply where the voltage is changed from 415 V to 400 V, or 660 V to 690 V. (c) Variations in rating principles The IEC rating principles are used worldwide, except in North America, where UL (Underwriters Laboratory) standards apply.

The rated current to a UL standard is the minimum current required to operate the fuse after many hours, and the current that it will carry indefinitely (the IEC rated current) is approximately 80 percent of this rating.

The voltage rating marked on a UL fuselink is the maximum voltage at which it can be used, whereas that marked on an IEC fuselink is the nominal voltage.

These differences must be considered when replacing fuselinks, particularly in the case of miniature cartridge fuselinks which are interchangeable. In general it is preferable to replace a fuselink with one of the same rating from the same manufacturer; this ensures that its characteristics are as similar as possible to those of the previous fuselink.

The IEC and UL ratings of fuse holders also differ. The IEC rating is the highest rated current of a fuselink with which it is intended to be used. A higher rating may be given in North America, this being related to the maximum current that does not cause overheating when a link of negligible resistance is used.

(d) Frequency ratings

Fuses are most commonly used in ac circuits with frequencies of 50 Hz or 60 Hz and a fuse designed for one of these frequencies will generally operate satisfactorily at the other. If the arc extinguishes at current zero, then the maximum arcing time on a symmetrical fault will be 10 ms at 50 Hz and 8 ms at 60 Hz.

Fuse manufacturers should be consulted about the suitability of fuses for other frequencies, which may include 17.67 Hz for some railway supplies, 400 Hz for aircraft and higher frequencies for some electronic circuits.

In dc circuits there is no current zero in the normal waveform and fuselinks designed for ac may not operate satisfactorily. Separate current and voltage ratings are given for fuselinks tested for use in dc circuits. DC circuits can be more inductive for a given current than ac systems, and since the energy in the inductance is dissipated in the fuse it is necessary for the dc voltage rating to be reduced as the time constant (WR) of a circuit increases.

(e) Breaking capacity

The breaking capacity of a fuse is the current which can be interrupted at the rated voltage. The required breaking capacity will depend upon the position of the fuse in the supply system. For instance, 6 kA may be suitable for domestic and commercial applications, but 80 kA is necessary at the secondary of a distribution transformer.

The power factor of a short-circuit affects the breaking capacity, and appropriate values are used when testing fuses.

(f) Time-current characteristics

The time-current characteristic of a fuse is a graph showing the dependence upon current of the time before arcing starts (the pre-arcing time); an example has been shown in Fig. 2. The total operating time of a fuse consists of the pre-arcing time and the arcing time. When pre-arcing times are longer than 100 ms and the arc is then extinguished at its first current zero (that is an arcing time of less than 10 ms on a 50 Hz supply) then the time-current characteristic can be taken to represent the total operating time.

The conventional time, the conventional fusing current and the conventional non- fusing current are often shown on time-current characteristics. These values are defined in the standards. All fuses must operate within the conventional time when carrying the conventional fusing current; when carrying the conventional non-fusing current they must not operate within the conventional time.

(g) I^2 t

I^2t is defined as the integral of the square of the current let through by a fuse over a period of time. Values are given by manufacturers for pre-arcing I^2t and total let- through I^2t.

Table 1 Example of the variation of I^2t with current rating

Table 1 shows typical values of I^2t for low-voltage cartridge fuses of selected current ratings: values differ between manufacturers.

The heat generated in a circuit during a short-circuit or fault condition before the fuse disconnects is given by the product of I^2t and the circuit resistance. As the let- through I^2t becomes constant above a particular level of fault current, the heat generated does not increase for prospective currents above this value, unless the breaking capacity is exceeded.

(h) Power dissipation

The resistance of a fuse will result in dissipation of power in the protected circuit when normal currents are flowing. This should be considered when designing the layout of a protection system.

(i) Cut-off current

A current-limiting fuse prevents a fault current from rising above a level known as the cut-off current. This is illustrated in Fig. 6. The cut-off current is approximately proportional to the cube root of the prospective current, and the maximum current is therefore very much lower than it would be if a non-current-limiting protection device were used.

Fig. 7 Examples of UMFs

Fig. 8 Blade-type automotive fuselink.

2.3 Main classes of equipment

Fuses are produced in many shapes and sizes, and various types are illustrated in Figs 5, 7, 8 and 9. The main three categories are:

-- miniature (up to 250 V)

-- low voltage (up to 1000 V ac or 1500 V dc)

-- high voltage (greater than 1000 V ac)

All three categories include current-limiting and non-current-limiting types.

(a) Miniature fuses

Cartridge fuses have in the past been the most common form of miniature fuse, but the UMF (see Fig. 7) is becoming increasingly used on PCBs. A UMF is much smaller than a cartridge fuse, and it is mounted directly on the PCB, whereas a cartridge fuse is mounted in a holder. Subminiature fuses have pins for mounting on PCBs.

Miniature cartridge fuses and subminiature fuses are rated for use at 125 V or 250 V. UMFs have additional voltage ratings of 32 V and 63 V which make them more suitable for many types of electronic circuit. Miniature fuses are available with current ratings from 2 mA to 10 A. The maximum sustained power dissipation which is permitted in cartridge fuses ranges from 1.6 W to 4 W. Miniature fuses may have a low, intermediate or high breaking capacity. All three ranges are available for UMFs, and these are shown in Table 2.

Fig. 9 A range of low-voltage fuselink types: (2) general purpose industrial fuselinks, (3) fuselinks for domestic purposes, (4) fuselinks for protecting semiconductors, (5) fuselinks for use in UK electricity supply networks, and (6) compact fuselinks for industrial applications .

Table 2 Breaking capacity of UMFs ---*or 10 x rated current, whichever is the greater.

Cartridge fuses are available with low or high breaking capacity. Low breaking capacity types have glass barrels without sand filler and a visual check can therefore be made on whether or not the fuse has operated. High breaking capacity cartridge fuselinks are generally sand filled and have ceramic barrels; they can interrupt currents of up to 1500 A. A range of speeds of operation is available. Time-lag (surge-proof) fuses are required in circuits where there is an inrush current pulse, for instance when capacitors are charged or when motors or transformers are magnetized. The fuse must not be operated by these normal-operation surges, which must not cause deterioration of the fuse. The five categories of time-lag are medium time-lag (M), time-lag (T), long time-lag (TT), very quick-acting (FF) and quick-acting (F). They are available as cartridge fuses and the last two are used in the protection of electronic circuits. The letters shown in brackets are marked on the end caps.

UMFs are available in similar categories, which are super quick-acting (R), quick- acting (F), time-lag (T) and super time-lag (S).

The time-current characteristics of miniature fuses of the same type but with different ratings are similar in shape. Time can therefore be plotted against multiples of rated current and it is unnecessary to show separate characteristics for each current rating. Examples of time-current characteristics are shown in Fig. 10.

Fig. 10 Time-current characteristics of time-lag and quick-acting miniature fuses (reproduced with permission from ERA Technology Ltd)

Cartridge fuses have various types of elements. Fast-acting types have a straight wire, and time-delay types use wire with an m-effect blob (Fig. 3(b)), helical elements on a heat-absorbing former or short elements with springs connecting them to the end caps.

In addition to the most common types which have been described, miniature fuses are produced in a wide range of shapes and sizes. As an example of this, a blade-type automotive fuselink is shown in Fig. 8; the element in this fuse is visible through the plastic casing.

(b) Low-voltage fuses

A wide range of low-voltage fuses is available for industrial and domestic applications.

These fuses have ratings appropriate for national or international single-phase or three-phase supplies, for example 220, 230, 240,400,415, 660 and 690 V. Widely differing systems for domestic protection are used in different countries and these cannot be described separately here. As an example, in the UK current- limiting cartridge fuses are used in plugs which supply appliances, the consumer unit supplying an entire property may have current-limiting cartridge fuses, semi-enclosed fuses or miniature circuit breakers and another fuse is installed by the supply authority on the incoming supply.

Industrial fuses may have general-purpose (type 'g') fuselinks which will operate correctly at any current between 1.6 times the rated current (the conventional fusing current) and the breaking capacity. Such fuses must not be replaced by type 'a' back- up fuselinks, which have a higher minimum breaking current and do not necessarily operate safely below this current; this type 'a' back-up fuselink is used to save space.

Type 'gG' fuses are used for general application. These have a full breaking capacity range and provide protection for cables and transformers and back-up protection for circuit breakers. Specialized fuses are available for the protection of motors, semiconductors, street lighting, pole-mounted transformers and other purposes.

Reference SA provides further detail. Common applications are motor protection and semiconductor protection and these are described briefly below.

Fuselinks for motor-starter protection must be able to withstand starting pulses without deterioration. 'gM' fuselinks are designed for this purpose and they have a dual rating. A designation 100M160, for example, means that the fuselink has a continuous rating of 100 A and the general-purpose characteristics of a fuselink rated at 160 A. Fuselinks for semiconductor protection are designed to operate with an arc voltage which does not damage the semiconductor device; this voltage is therefore lower than for other types of fuselink. Arc voltages at several supply voltages are shown in Table 3 for typical semiconductor fuselinks.

Table 3 Peak arc voltages for semiconductor fuselinks

Semiconductor fuselinks also have lower let-through Z*r and cut-off current because semiconductors are susceptible to damage by heat and overcurrents. These fuses operate at higher temperatures than normal to achieve the necessary protection, and forced air cooling may be used to increase their current rating.

(c) High-voltage fuses

High-voltage fuses can be of current-limiting or non-current-limiting type. The latter are expulsion fuses which do not contain the arc products when they operate; they can be very noisy and are therefore normally used outdoors.

Current-limiting high-voltage fuses are enclosed (as already shown in Fig. 5)

and they may be used for the protection of motors, transformers and shunt power capacitors. The rated current of the fuselink is normally higher than the expected current. These fuses are normally used in three-phase systems and are tested at 87 percent of their rated voltage. In a three-phase earthed neutral system the voltage rating should be at least 100 percent of the line-to-line voltage, and in a single-phase system it should be at least 115 percent of the circuit voltage.

Further information can be found in references 8A, 8B and 8C.

2.4 Test methods

(a) Type tests

Before production of a type of fuselink commences, type tests are performed to ensure that preproduction fuselink samples comply with relevant national or international standards. Measurements of power dissipation, time-current characteristic, overload withstand capability, breaking capacity and resistance are included in these type tests.

(b) Production tests

Routine testing of many important fuse characteristics is not possible because tests such as breaking capacity are destructive. Extensive testing in production would also be very costly. Fuse manufacturers therefore make production fuselinks as identical as possible to the samples used for type testing.

The quality of fuselinks depends upon the quality of the components supplied to the fuse manufacturer. Key items such as barrels, filling material, element material and end caps are therefore regularly inspected and tested when received.

The dimensions and straightness of barrels are checked and their ability to withstand mechanical and thermal shock and internal pressure is tested. End cap dimensions are checked to ensure that they fit closely over the barrel. The moisture content, chemical composition and grain size of the filler are measured. The diameter or thickness of the element wire or tape is checked and its resistance per meter is measured. Where elements are produced from tape and notched, the dimensions and pitch of the notches are tightly controlled.

During assembly checks are made to ensure that the fuselink is completely filled with sand and that the element resistance is correct. After assembly the overall dimensions are checked and the resistance is once more measured. A visual check including the markings is then made.

Other tests are made in the case of specialized fuselinks. For example, the condition of the elements in a high-voltage fuselink is examined using X-ray photography.

In addition to these routine tests, manufacturers may also occasionally take sample fuselinks from production and subject them to some or all of the type tests.

(c) Site checks

Before use, every fuselink should be checked visually for cracks and tightness of end caps and the resistance should be checked. It should also be checked that the ratings, especially current, voltage, breaking capacity and time-current characteristic, are correct for the application. In the case of semi-enclosed, rewireable fuses care should be taken to use the appropriate diameter of fuse wire. Fuse holders should be checked to ensure that the clips or means of connection are secure and correctly aligned.

If a fuselink has been dropped onto a hard surface or subjected to other mechanical stress it should not be used, damage may not be visible but it could cause the fuse to malfunction with potentially serious results.

If a fault occurs and the fuselink is overloaded, it should be replaced even if it has not operated. This situation arises especially in three-phase systems where one or two of the three fuses may operate to clear the fault.

Fuselinks (as opposed to rewireable fuses) cannot be safely repaired; they must always be replaced.

2.5 Standards

Many national and international standards exist because of the number of different fuse types. Tables 4, 5 and 6 summarize the position for miniature, low-voltage and high-voltage fuses respectively. IEC recommendations are listed, together with related EN and BS standards and North American standards covering the same field.

Table 4 Comparison of international, regional and national standards for miniature fuses

In order to comply with the EMC directive, the following statement has been added to most of the UK fuse standards: Fuses within the scope of this standard are not sensitive to normal electromagnetic disturbances, and therefore no immunity tests are required. Significant electro- magnetic disturbance generated by a fuse is limited to the instant of its operation.

Provided that the maximum arc voltages during operation in the type test comply with the requirements of the clause in the standard specifying maximum arc voltage, the requirements for electromagnetic compatibility are deemed to be satisfied.

3. Protection relays

3.1 Principles of design and operation

Systems incorporating protection relays can disconnect high currents in high-voltage circuits which are beyond the scope of fuse systems.

In general, relays operate in the event of a fault by closing a set of contacts or by triggering a thyristor. This results in the closure of a trip-coil circuit in the circuit breaker which then disconnects the fault. The presence of the fault is detected by current transformers, voltage transformers or bimetal strips.

Table 5 Comparison of international, regional and national standards for low-voltage fuses

Electromechanical and solid-state relays are both widely used, but the latter are becoming more widespread because of their bounce-fixe operation, long life, high switching speed and additional facilities that can be incorporated into the relay.

Additional facilities can, for instance, include measurement of circuit conditions and transmission of the data to a central control system by a microprocessor relay. This type of relay can also monitor its own function and diagnose any problems that are found. Solid-state relays can perform any of the functions of an electromechanical relay whilst occupying less space, but electromechanical relays are less susceptible to interference and transients. Electromechanical relays also have the advantage of providing complete isolation and they are generally cheaper than solid-state devices.

Table 6 Comparison of international. regional and national standards for high-voltage fuses

Contacts in electromechanical relays may have to close in the event of a fault after years of inactivity and twin sets of contacts can be used to improve reliability. The contact material must be chosen to withstand corrosive effects of a local environment because a film of corrosion would prevent effective contact being made. Dust in the atmosphere can also increase the contact resistance and result in failure. Both corrosion and dust contamination can be avoided by complete enclosure and sealing of the relay.

The contacts must also withstand arcing during bounce on closure and when opening, but this is usually less important than resisting corrosion. Because of the need for high reliability the contacts are usually made of or plated with gold, platinum, rhodium, palladium, silver or various alloys of these metals.

Solid-state relays are not affected directly by corrosion or dust, but temperature and humidity may effect them if conditions are severe enough.

Electromechanical relays operate by induction, attraction or thermally where a bimetallic strip is used to detect overcurrent. The first two types are most common and their principles are described, along with those of solid-state relays, in the following sections. Further information on protection relays can be found in references 8C, 8D and 8E.

Fig. 12 Time grading

(a) Induction relays

An induction relay has two electromagnets, labeled E1 and E2 in Fig. 11. Winding A of electromagnet E1 is fed by a current transformer which detects the current in the protected circuit. Winding E in electromagnet E1 is a secondary, and it supplies the winding on E2. The phases of the currents supplied to E1 and E2 differ and therefore the magnetic fluxes produced by the two electromagnets have different phases. This results in a torque on the disc mounted between the electromagnets, but the disc can only move when a certain torque level is reached because it is restricted by a hair spring or a stop. Normal currents in the protected circuit do not therefore cause movement of the disc.

When the disc does turn, its speed depends upon the current supplied by the current transformer and the eddy current braking effected by a permanent magnet located near the edge of the disc. When the disc rotates through a certain angle, the relay contacts close and the time for this to occur can be adjusted by the position of the stop or the angle through which the disc has to rotate. This adjustment allows protection co-ordination to be achieved by means of 'time grading'. For example, in the radial feeder shown in Fig. 12, the minimum rime taken for the protection relay to operate would be set higher at points closer to the supply. A fault at point X, a considerable distance from the supply, would cause operation of the relay set at a minimum time of 0.4 s and the fault would be disconnected before it caused operation of relays nearer the supply, thus preventing the unnecessary tripping of healthy circuits.

Fig. 11 Principle of operation of an induction relay

Fig. 13 Principle of operation of an attracted-armature relay

Fig. 14 Components of a solid-state time-delayed overcurrent relay

Fig. 15 Comparison of time-current characteristics of electromechanical and solid-state relays

Induction relays have Inverse Definite Minimum Time (IDMT) time-current characteristics in which the time varies inversely with current at lower fault currents, but attains a constant minimum value at higher currents. This constant minimum value depends upon the adjustments previously described.

Further adjustment is possible by means of tappings on the relay winding A in Fig. 11. For example, if a current transformer has a secondary winding rated at 1 A, tappings could be provided in the range 50 percent to 200 percent in 25 percent steps, corresponding to currents of 0.5 A to 2 A in 0.25 A steps. If the circuit is up-rated, it may then be possible to adjust the relay rather than replace it. For example, if the 100 percent setting is used when the maximum current expected in the protected circuit is 400 A, the 150 percent setting could be used if the maximum current is increased to 600 A.

(b) Attracted-armature relays

The basis of operation of an attracted-armature relay is shown in Fig. 13. The electromagnet pulls in the armature when the coil current exceeds a certain value and the armature is linked to the contacts and when it moves it opens normally closed contacts and closes normally open contacts. The time required for operation is only a few seconds, and it depends upon the size of the current flowing in the coil.

These devices are called instantaneous relays. They have a range of current settings which are provided by changing the tapping of the coil or by varying the air gap between the electromagnet and the armature.

(c) Solid-state relays

The first solid-state relays were based on transistors and performed straightforward switching, but now they can often perform much more complicated functions by means of digital logic circuits, microprocessors and memories.

Currents and voltages are measured by sampling incoming analogue signals, and the results are stored in digital form. Logic operations such as comparison are then performed on the data to determine whether the relay should operate to give an alarm or trip a circuit breaker. Digital data can also be stored on computer €or subsequent analysis, for instance after a fault has occurred.

Solid-state relays may incorporate a variety of additional circuits. Figure 14 shows, for example, a circuit which imposes a time delay on the output signal from a relay.

Such circuits may also be used to shape the time-current characteristic of a protection relay in various ways. A comparison of typical time-current characteristics from electromechanical and solid-state relays is shown in Fig. 15. There are many other possible functions such as power supervision to minimize power use, an arc sensor to override time delays and measurement of true rms values in the presence of harmonics.

Solid-state relays my require shielding against electromagnetic interference arising from electrostatic discharges or high-voltage switching. Optical transmission of signals is sometimes used to reduce the effects of this interference.

The electronics in solid-state relays can be damaged by moisture, and the relays are usually encapsulated to prevent this.

3.2 Rating principles and properties

All components of the protection system including current transformers, relays and circuit breakers must have the correct current, voltage and frequency rating, and the I^2t let-through and interrupting capacity of the entire system depends upon the circuit breaker. A protection relay must have a minimum operating current which is greater than the rated current of the protected circuit, and other properties of the relay must be chosen correctly in order for it to operate the circuit breaker as required by the application.

Manufacturers publish information regarding the selection, installation and use of relays. The following points in particular will need to be considered.

Heat is generated within a relay in use and if several relays are grouped together in an enclosed space provision should be made to ensure that temperature rises are not excessive.

Protection levels, time delays and other characteristics of both electromechanical and solid-state relays can be changed on site. For example, overvoltage protection may be set to operate at levels between 110 percent and 130 percent, and Fig. 16 shows the effect of adjusting the operating time of an overcurrent relay.

In general, electromechanical relays can be adjusted continuously and solid-state relays are adjusted in steps. Many other adjustments can be made. For example, DIP switches in solid-state relays can be used to set the rated voltage and frequency and to enable phase-sequence supervision.

The adjustment of operating times is used to provide time grading and co-ordination, which have already been explained with reference to Fig. 12. An alternative system uses protection relays which operate only when a fault is in a clearly defined zone; this is illustrated in Fig. 17. To achieve this, a relay compares quantities at the boundaries of a zone. Protection based on this principle can be quicker than with time-graded systems because no time delays are required.

Fig. 16 Adjustment of time-current characteristics for an overcurrent relay.

Fig. 17 Overlapping zones of protection. The protection zone marked -%-x-% overlaps the zones marked -. X marks the positions of circuit breakers with their associated protection relays

3.3 Main classes of relay

Protection relays may be 'all-or-nothing' types, such as overcurrent tripping relays, or they may be measuring types which compare one quantity with another. An example of the latter is in synchronization, when connecting together two sources of power.

A protection relay may be classified according to its function. Various functions are noted below, and details for a wide range of functions are give in reference 8F.

Common applications are:

--undervoltage and overvoltage detection

--overcurrent detection

--overfrequency and underfrequency detection

These functions may be combined in a single relay. For example, the relay shown in Fig. 18 is used in power stations and provides overvoltage and undervoltage, overfrequency and underfrequency protection, and it rapidly disconnects the generator in the case of a failure in the connected power system.

Fig. 18 Mains & coupling relay showing adjustments for voltage and frequency protection (reproduced with permission from SEG)

Another application in power systems is the protection of transmission lines by distance relays. Current and voltage inputs to a distance relay allow detection of a fault within a predetermined distance from the relay and within a defined zone. The fault impedance is measured and if it is less than a particular value, then the fault is within a particular distance. This is illustrated in the R-X diagrams shown in case (a) of Fig. 19; if the R and X values derived from the measured fault impedance R + iX result in a point within the circle, then the relay operates. If directional fault detection is required, the area of operation is moved, as shown in case (b).

Fig. 19 Operating areas for distance protection relays

Another form of protection for lengths of conductor is the pilot wire system.

Current transformers are placed at each end of a conductor and are connected by pilot wires. Relays determine whether the currents at the two ends of the conductor are the same, and they operate if there is an excessive difference. In the balanced voltage system shown in Fig. 20(a), no current flows in the pilot wire unless there is a fault.

In a balanced current system, current does flow in the pilot wire in normal conditions, and faults are detected by differences in voltage at the relays which are connected between the pilot wires; this is shown in Fig. 20(b).

CT Protected conductor CT

Fig. 20 Pilot wire protection systems

Other applications involving relays include: checking phase balance the protection of motor starters against overload the protection of generators from loss of field the supervision of electrical conditions in circuits

3.4 Test methods

(a) Production tests

Manufacturers of electromechanical relays often produce their own components and inspect them before assembly. Components for solid-state relays are generally bought in from specialist manufacturers and there is an incoming check to eliminate those that would be subject to early failure or excessive drift. Other characteristics such as memory are checked as appropriate.

After assembly, manufacturers test all protection relays to ensure that they comply with relevant national and international standards. The calibration of adjustable settings is checked.

Fault conditions that the relays are designed to protect against can be simulated, and typical inputs to relays from current or voltage transformers can be duplicated by test sets; these can be used to check, for example, the correct functioning of relays for phase comparison and the proper disconnection of overcurrent and earth faults of various impedances. Such tests can be performed by setting up a test circuit or by means of a computer-based power system simulator which controls the inputs to the relays. The latter method allows the effect of high-frequency transients and generator faults to be investigated, and it is independent of an actual power supply; it may, for example, be used for the testing of selectable protection schemes in distance relays.

Other tests include:

-- environmental tests are performed to check, for example, the effects on performance of temperature and humidity

-- impact, vibration and seismic tests are performed on both solid-state and electromechanical relays, although the latter are more prone to damage from such effects

-- voltage transients are potentially damaging to solid-state relays, and the relays are tested to ensure they can withstand a peak voltage of 5 kV with a rise time of 1.2 ps and a decay time of 50 ps.

(b) Site tests

Primary injection tests are applied during initial commissioning and these should show up any malfunction associated with protection relays. Secondary injection tests should be performed if there is a maloperation which may be related to the protection relay. Details of these injections tests are given in reference 8E. Periodic inspection and testing is necessary throughout the lifetime of a protection relay and computerized equipment is available for this purpose. Some relatively complicated protection schemes incorporate automatic checking systems which send test signals at regular intervals; the test signals can also be sent manually. Digital relays may include continuous self-checking facilities.

4. Standards

There are many standards covering various types of relay and aspects of their use.

Some standards which apply to relays in general are also relevant to protection relays and some apply specifically to protection relays. The key IEC recommendations together with equivalent BS and EN standards and related North American standards are summarized in Table 7.

Table 7 Comparison of international. regional and national standards for protection relays

References

Williams, D.J.A., Turner, H.W. and Turner, C., User's Guide m Fuses, 2nd edn, ERATechnology Ltd, UK, 1993.

Wright, A. and Newbery, P.G., Eleclric Fuses, Peter Peregrinus Ltd, London & New York, 1982.

Wright, A. and Christopoulos, C., Electric Power System Protection, Chapman & Hall, London, 1993.

GEC Measurements, 'Protective relays -Applications guide', 3rd edn. The General Electric Co., UK, 1987.

Jones, G.R., Laughton, M.A. and Say. M.G., Electrical Engineer's Reference Book, Butterworth- Heinemann, Oxford, UK, 1993.

Recommended practice for protection and co-ordination of industrial and commercial power systems, ZEEE, Wiley Interscience, 1986.