Relay Circuit Breaker Application

CHAPTER
11
RELAY AND CIRCUIT-BREAKER

Original Authors: E. L. Harder and J. C. Cunningham HE function of relays and circuit breakers in the operation of a power system is to prevent or limit damage during faults or overloads, and to minimize their effect on the remainder of the system. This is accomplished by dividing the system into protective zones separated by circuit breakers, such as are shown in Fig. 1. During a fault, the zone which includes the faulted apparatus is de-energized and disconnected from the system. In addition to its protective function, a circuit breaker is also used for circuit switching under normal conditions. The relay application problem consists of selecting a relay scheme which will recognize the existence of a fault within a given protective zone, and initiate circuit-breaker operation. This problem is considered from a system of protective point of view. The operating characteristics schemes for generators, transformers, lines, and buses are discussed in their relation to overall system performance. Reference is made to other publications, particularly Silent Sentinels, for the operating characteristics and connections of individual relays. It is proposed here only to give the general features which will determine the type of scheme to be used. The circuit-breaker application problem consists primarily of determining the interrupting requirements,
APPLICATION
Revised by: E. L. Harder and J. C. Cunningham
normal current, voltage, and other rating factors required to select the proper breaker for each location. These factors are discussed, and methods of calculating the fault current and interrupting rating are given. I. GENERAL PHILOSOPHY AND BASIC RELAY ELEMENTS As mentioned the system is divided into protective zones as shown in Figure 1, each having its protective relays for determining the existence of a fault in that zone and having circuit breakers for disconnecting that zone from the system. It is desirable to restrict the amount of system disconnected by a given fault; as for example to a single transformer, line section, machine, or bus section. However, economic considerations frequently limit the number of circuit breakers to those required for normal operation and some compromises result in the relay protection. The relays operate usually from currents and voltages derived from current and potential transformers or potential devices. A station battery usually provides the circuit breaker trip current. Successful clearing depends on the condition of the battery, the continuity of the wiring and trip coil, and the proper mechanical and electrical
Fig. lTypical
system showing protective zones.
Chapter 11
Relay and Circuit-Breaker
Application
343
operation of the circuit breaker as well as the closing of the relay trip contacts. In event of failure of one of these elements, so that the fault in a given zone is not cleared by the first line of defense, relays and circuit breakers, some form of back-up protection is ordinarily provided to do the next best thing. This means, first of all, to clear the fault automatically, if at all possible, even though this requires disconnection of a considerable portion of the system. Once cleared, the system can generally be rather quickly restored; whereas if the fault hangs on, the line may be burned down, or apparatus damaged beyond repair, or the entire system may be shut down for an extended period. The measures taken to provide back-up protection vary widely depending on the value and importance of the installation and the consequence of failure. These will be discussed in a separate section. Some utilities in measuring the performance of transmission line relay protection analyze all relay trip operations as shown in Table 1. The numbers shown are typical of a system operating 3000 miles of 110-kv line. This is only an analysis of faults for which the relay tripped or should have tripped. For each of these there were several cases where the relays should not have tripped, and did not. Thus the total number of discriminations made by the relays is possibly five to ten times
TABLE 1RELAY OPERATIONS
failure of the circuit breaker, or blown fuses in trip circuits (if used). Only a small part of the total failures occur in the protective relay itself. Thus close attention to the initial design, installation, testing, and maintenance of all of the accessory equipment, as well as of the protective relay proper, are needed to assure successful operation. The application of protective relays properly requires evaluation of several factors, namely: 1. The requirements of the power service and desired functioning of the system during fault conditions to produce this result.
2. The currents, voltages, temperatures, pressures, or other indicators at time of fault which provide the fundamental basis of discrimination. 3. The characteristics of available or standard relay elements. 4. The schemes in which they are used.
A wide variety of characteristics are now available operating in response to the prime quantities themselves, or to various functions of these prime quantities, such as power, phase angle, power factor, current comparison, power comparison, impedance, reactance, modified reactance, current ratio, or phase-sequence component. In each case the response may be instantaneous, meaning no intentional delay, or the operation may depend in a predetermined manner on the electrical quantities and time of duration.
2. Basic Relay Element87
as great as the trippings. The percentage failures are correspondingly less on this larger basis. However, Table I has been presented at this point to bring out the following factors that enter into a highly successful protective relay system : 1. Good equipment, relays and instrument transformers. 2. A system design that can be protected and correct application of relays to provide the possible protection. 3. Good maintenance primarily to assure that all the accessories are operative. The correct but undesired trippings are cases where the relays have done what should be expected from their characteristic curves and settings and the fault conditions involved. There may have been system changes since their application, or incorrect initial application, or application with foreknowledge that certain conditions would unavoidably operate the relays, but this was necessary to secure tripping in other desired cases. It is important to bear in mind that simple standard system design plans can be better protected. Distance measuring and carrier or pilot-wire types of relaying are much less subject to disqualification by system changes than are over current types. Wrong tripping and tripping failures, together with all causes of failure to clear faults, are found to stem largely from human errors, such as leaving the trip circuit open after test, or to open circuited trip coils, or mechanical
The more commonly used relay elements and their underlying principles of operation are shown in Fig. 2. The schemes in which the elements are used are much more numerous. The more common ones will be described under the application headings, such as generators, transformers, and buses. Instantaneous ElementsFor instantaneous response to current or voltage the solenoid element, Fig. 2(a), is most common, appearing individually or as the instantaneous attachment with the induction-type overcurrent relay. The beam element, (b), with spring or weight bias is used where low burden is desirable, as when setting for low ground currents with low-ratio bushing current transformers. The polar element, (j), is of far lower burden than the nonpolar types and has come into widespread use since 1935 as the receiver relay of directional-comparison carrier equipments, and is the basic element when supplied from networks or electronic devices, For example, it appears as the operating element in a pilot wire relay, in a phase-comparison carrier relay, in linear-coupler bus protection, and in supervision of pilot wires. Because of its higher pick-up to drop-out ratio and less accurate setting the clapper-type element, (c), is used less frequently for the primary protective functions but is widely used as an auxiliary relay. Induction ElementsThe induction-disk element, (d), continues to be most widely used, its reliability and inherent time characteristic giving it great flexibility for coordinating relays in series or coordinating with fuses or direct-trip devices. A variety of characteristics are available from the definite-minimum-time, which is ideal for securing definite time steps between relays, to the very inverse which provides faster tripping with the same margins when the fault current drops considerably from one
344

Fig. 2Relay
Application
Chapter 11
elements.
Chapter 11
Application
345
Fig. 2Relay elementsContinued
346
Application
Chapter 11
Fig. 2Relay elementsContinued
Chapter 11
Fig.2Relay

elements-Continued
Application
347
relay location to the nest. The more inverse characteristic also co-ordinates better I\-ith fuses. The induction disk scrvcs as a tlirwtional element, ((I), or, ivhcn used I\-ith a spring, as a \vaf tj c+mcwt, (I), the electric torque being proport ional to 131 cos 8, where 0 is the power-factor angle. IjZthcr of thaw can, of COIII~, bo current polarized instead of voltage polarized for uw in ground relaying \vhcn a txmk-nwt 1x1 curwnt is availal)lo. The relative phase of voltage and c*urwnt wn be shiftc~cl by internal or external phuw-shift ing cloviws to prtxlriw maximum torque for po\vcr factors other than unity, as in (9). For csample, the directional cbl(tmcnt for ground rclaying usually has its masimurn tor(ltic for current lagging the voltage by GO or 70 dcgrws to proviclc masim~lm torques for the faiilt contlitions. A pure \vat t char:wt(~rist ic is used with the 30 and 60 dcgrcw conncc*t ions for ~)hasc directional relays, the phase shift twing provided t)y iwing the voltage of a different phaw from the clirwnt. \li th the 90 dcgrec connection this shift is too much, ant1 the voltage is advanced about, 43 dc~grc~~+ a phase shiftclr to t)y provide a maximum-torque position for a current 45 degrees lagging. The disk is provided also n-ith an clcctromagnct, (u), producing ratio characterist its for use as a differential rclay for generators or transformers. The generator-difrerential relay is shown in (0). The transformer relay has
windings 1 and 2 tapped for different current-transformer r atios. Referring to Figure 2 (o), the differential current produces lower-pole flus which acts in the operating direction on disk currents produced by the upper pole which is transformer fed from the same differential current. The restraint torque, giving the ratio-differential characteristic, is produced by the through current in coils 2-3, which supply disk current by transformer feed to the upper pole, and by lower-pole flux produced by the same through current. The induction disk provides a tripping-time-proportional-to-impedance characteristic as shown in Figure 2(X). Multiple Electromagnets or Disks--Two electromagnets on the disk provide for balancing mechanical torques with no phase angle effects (see (p)). When these are both current electromagnets, the relay is the regulating-transformer differential relay which balances the current in the shunt exciting winding against the through line current. For example in a f 10 percent regulating transformer, it operates when the shunt current exceeds about 15 per cent of the through current. When both electromagnets are volt age energized, the voltage-differential relay results. When one is voltage (actually responsive RE2) and one is a productJ element, El cos (0+), a balance occharactercurs when I (0s (0-4) = KE. This impedance istic: is shown in current and also in an R-X plot in (m). A second disk on the same shaft provides space for two This structure is sed as the phasemore electromagnets. balance relay for motor protection, whose characteristics are shown in (p). It is used for the 3-winding-transformer differential relay, using one operating electromagnet and three restraint electromagnets, (n). With two currentinput input on each electromagnet and with two relays per phase the multirestraint bus differential relay results. Its used will the described later.
Multiple-Pole
Cylinder
or Disk
Elements-The
multiple-pole cylinder or disk element is illustrated in (f). The example shows how it would be energized to act as a single-phase directional element having torque proportional to It1 cos 8. This element also serves as a polyphasedirectional element by the connection, (q). The multiplepole element is flexible making possible a variety of other combinations.
Four-Pole
Induction-Cylinder
Elements-These
high-speed elements serve a variety of purposes as shown in Figure 2(l), (LI), (w), and (.c). The clement (u), desigbated a mho element, operates with torque E cos (0-4) restrained by torque proportional to e2. It produces the circular-impedance-tripping locus passing through the origin or relay location, the same as shown for the induct ion disk in (in). Or with either clement, the circle can be shifted from this position by current compensation, IZb, A directionalin the restraining circuit, as indicated. starting unit, (I), is obtained using current times shiftedquadrature voltage for operating and the product of two delta for restraint. This results in maximum torque for current 30 degress ahead of the quadrature volt ago or about 60 degrees lagging the unity-power-factor posit ion. The special impedance characteristic, (w), obtained, by
348
Application
Chapter
11
12 operating against EI cos (0 +20 degrees) is used to restrict the tripping area to assist other relays in differentiating heavy load swings from faults. A reactance element, (s), is obtained similarly with the phase-shift devices arranged so that the maximum-torque line is along the x (reactance) axis. Inductor-Loop ElementThe inductor loop, (h), provides a very high speed and very reliable directional element which has been used for many years now in highspeed distance measuring relays. Balance- Beam Element-The basic balance-beam impedance element is shown in (i), a balance occurring for E/1=2,,. For higher impedances than 20 (current relatively lower) the contacts remain open; whereas for lower impedances (relatively higher currents) they close quickly. Since the balance is mechanical, the phase angle between voltage and current is of minor consequence, and the tripping characteristic, plotted on an R and X diagram is substantially a circle. Modified-Impedance CharacteristicThe circular characteristic may be shifted by some circuits auxiliary to the element as shown in (k), in order to provide better discrimination between fault currents and load and swing currents on long, heavily-loaded transmission lines. The shifting imparts a directional characteristic to the relay in addition to narrowing its tripping region to more nearly just that required for faults. Networks and Auxiliary Circuits-It may be noted that in discussing fundamental relay elements certain auxiliary circuits external to the mechanical relay have been introduced: in (g), the phase shifter; in (j), the Rectox; and in (k), a full fledged network to produce in the relay element proper, the desired currents. This is a trend of which we shall certainly see more as time goes on, as static circuits are devised to produce a simple current output proportional to the desired function of the various line currents and voltages. Sequence-Segregating Networks, I,+K&The method of symmetrical components has been the key that has unlocked the door to a number of the aforementioned possibilities, some of which are illustrated in (r), (s), and (t). The positive- and zero-sequence network in (T) is commonly used in pilot-wire relaying, where it is desired to compare over the wires only one quantity, which is a good measure of the fault current irrespective of what kind of fault it may be, that is A-B, A-G&, ABC. The relay can be given almost independent and widely different settings for phase faults and ground faults, using the single relay element. For example, it may be set for one ampere of ground fault to provide the requisite sensitivity, but for ten amperes of 3-phase current to avoid operation on loads. A negative-sequence directional element is shown in (8). It is an adequate directional element for ground faults on reasonably well-grounded systems, and requires only two potential transformers rather than three as with usual residual-directional relays. Another novel application, (t), is the phase-selector relay to determine which phase is faulted. This information is necessary in single-pole tripping and reclosing schemes. It is predicated on the knowledge, from symmetricalcomponents theory, that the negative-sequence current in
the faulted phase only is in phase with the zero-sequence current. Individual overcurrent elements in the three phases could not be used for this selection as all three would pick up for a single line-to-ground fault on many solidly-grounded systems. II. Protective follows: PROTECTIVE may SCHEMES be conveniently classified as
schemes
1. Apparatus Protection 2. Bus Protection 3. Line Protection Thus, in Fig. 1, generator and transformer protection come under the Apparatus classification; generator buses, high-voltage buses, and substation buses, under the second classification; and high-voltage transmission lines and feeders under Line Protection. The relay application chart, Table 2, has been included for ready reference in determining the operating principles and application of various specific relay types referred to throughout this chapter. 3. A-C Generators Most a-c generators above 1000 kva and many smaller machines are equipped with differential protection arranged to trip if the currents at the two ends of each phase winding differ. This scheme is shown in Fig. 3. Smaller machines are sometimes operated without differential protection, but if paralleled with larger machines
Fig. 3Connections for one phase using the percentage differential relay for generator protection.
or with a system, they may be arranged to trip off on a reversed flow of power into the machine. For differential protection the Type CA normal-speed, induction, ratio-type relays are used in the large majority of cases, their speed (about 0.1 second relay time on severe faults) being adequate to prevent serious burning of the iron in nearly all cases. However, a high-speed generator-differential relay, Type HAlg, is available providing :
Chapter
11
Application
349
l-cycle protection and is being used with 100 per cent SUCcess in a number of important applications. The relay is usually arranged to trip the generator, field circuit, and neutral circuit breakers (if any) simultaneously by a manually-reset lockout relay in new installations. Frequently the relay also trips the throttle and admits For example it may be required to CO, for fire prevention. coordinate with other high speed relays or to reduce the shock to the systems. If a single-winding generator (or equivalent) is connected to a double bus through two breakers, a current transformer matching problem is introduced. The current transformers in the connections to the busses may carry large currents from one bus to the other in addition Thus, matching is not assured to the generator current. by identical current transformers as in the simpler case of Fig. 3, and consequently, the Type HA relay is preferred for this case because of its superior discriminating qualities. The Type CO relay is also used for generator differential protection. It provides straight differential protection, as contrasted with percentage differential, the diagram being the same as Fig. 3 without the restraining coils. Its setting must be considerably coarser than that of the CA relay because there are no restraining coils to desensitize it when high through-fault currents are flowing.
lower impedance grounding the differential relay protects closer to the neutral. With higher impedance grounding, the limit of protection for ground faults is farther from the neutral end, and for an ungrounded machine, the differential protection is ineffective against ground faults. The protection afforded for phase-to-phase, double-phaseto-ground, or three-phase faults is relatively unaffected by A complete discussion of the the method of grounding. methods of grounding is given in Chap. 19.
Solidly Grounded and Low Resistance or Reactance Grounded Machine-If the generator is solidly
grounded, or grounded through a reactor or resistor, drawing at least full-load current for a ground fault at a line terminal, the usual 10 percent differential relay operates for practically any short circuit within the machine and for grounds to within 10 percent of the neutral, or closer if the ground current is higher.
Ungrounded, and Potential-TransformerGrounded Generators-are those grounded only through

the natural capacitance from the metallically connected windings, buswork, and cables to ground. The potential transformer from neutral to ground, if properly applied?, serves as a measuring device only. To insure that this is so, it must be liberally designed so that under no condition will its exciting current become appreciable compared with the charging current to ground. Otherwise, ferro-resorated transnance may occur. Usually a full line-to-line former will suffice. The potential transformer and a voltage relay such as the SV (instantaneous) or CV (inverse time) may be used to initiate an alarm or optionally to trip. Or, on lower voltages, a static voltage unbalance indicator may be used connected directly to the primary circuit. Such an instrument is the Type G. These devices supplement the generator differential protection to provide indication or tripping for ground faults. Light resistance grounding as covered in the next section is generally preferred to ungrounded operation,
Double- Winding
and Multiple-
Winding
Gener-
ators-The differential protection scheme of Fig. 3 does not detect turn-to-turn short circuits within the winding because the entering and leaving currents of a phase remain equal. Double and multiple winding machines provide a means for obtaining such protection in the larger, more important generators. The currents in the parallel branches, become unequal when turns are short circuited in one branch. The differential relays, Type CA or HA, can be arranged to detect shorted turns, grounds* or phase-to-phase faults, by placing one current transformer in the neutral end of one of the parallel windings, and one of double ratio at the line end in the combined circuit. The choice of schemes depends somewhat on the facility with which leads can be brought out and the necessity of overlapping the generator breaker. With hydrogen cooling additional leads can be brought out through the necessary gas-tight bushings only with considerable difficulty, and usually there is no space for transformers inside the hydrogen compartment. Effect of the Method of Grounding-The method of grounding the generator neutral affects the protection afforded by differential relays. For example, if sufficient grounding impedance is used so that a ground fault at the generator terminals draws full load current, then for a fault at the midpoint of the winding, where the normal voltage to ground is half as great, the fault current will be approximately one-half the full load current. When a ground fault occurs 10 percent from the neutral end of the winding, the fault current, being limited largely by the neutral impedance, is about 10 percent of full load current. This corresponds to the sensitivity of a 10 percent differential relay and, therefore, represents the limit of protection for phase to ground faults with such a relay. For
*With the same limitations as for a single winding generator.
Light-Resistance-Grounded
GeneratorsThis
scheme and an associated protective arrangement is illustrated in Fig. 29 of Chap. 19. Indication from a voltage relay, connected in parallel with the resistor as shown, or from a current relay, such as the Type BG, connected in series with the resistor, may be used to sound an alarm or to trip, depending on the application. Combinations of sensitive alarm and coarser trip, or of alarm and timedelay trip, have also been used. The latter gives time to transfer the load to another machine at the hazard of operating with a fault on one phase. This scheme was designed primarily for the unit station arrangement in which a generator and step-up transformer are operated as a unit without an intervening bus. However, it can also be used where an intervening bus carries the station service transformer and one or two short feeder cables. A limited amount of selectivity is possible by the use of a polarized relay, such as the CWP-1, which obtains most of its energy from a potential coil in parallel with the grounding resistor. Such a relay used in the station-service feed, for example, can detect a ground on that circuit. Field ProtectionWhile a large number of machines still operate without any protective relays to function on See also Light-ResistanceGrounded Generators.
350
Application
Chapter 11
Chapter 11
Application
351
TABLE 2RELAY APPLICATION CHART-Continued
352
Application
Chapter
11
loss of field, there is a trend to more general use of equipOn some systems where loss of ment for this purpose. field would cause serious low voltage and danger of instability but where the system is operated in a way which can tolerate the loss of one machine, automatic means are being provided to disconnect the machines on loss or partial loss of field. Fairly common is the use of a d-c under-voltage relay, a Type D-2 dArsonva1 relay in series with a resistor, connected across the slip rings for field short circuit detection; also a similar element across the field ammeter shunt for undercurrent or open field detection. These do not provide complete protection, and there are a number of installations of reactive power relays used at the generator terminals in conjunction with under voltage to trip for any field reduction which would cause serious low voltage. On many closely knit metropolitan systems loss of field has been found to be not serious if immediately corrected. The operator attempts to restore the field, the generator in the meantime operating at somewhat reduced load as If he cannot restore field within an induction generator. a few minutes, he must trip the line circuit breaker to avoid injurious rotor heating. Field Ground Detection-Some form of field ground detection is frequently provided. It is considered most important to detect the first ground because a second could short circuit part of the field winding causing unbalance and vibration which could wreck the machine. The a-c scheme provides complete coverage for solid grounds. The d-c scheme gives nearly complete coverage, complete if the main field rheostat is varied. In some instances vibration detectors26 are used if the machine is known to be operating with a ground on its field. This will trip the unit instantly in event a second ground occurs. The over-all protection scheme frequently includes armature and bearing temperature indication and sometimes alarms. Less frequently field temperature indications are provided. The voltage regulators are sometimes equipped with over- and under-voltage protection, and, of course, overspeed protection is provided. In addition to the protection described, generators can be equipped with over- or under-voltage, frequency, overspeed, and loss of field, and temperature responsive devices. 4. Transformer Protection
an emergency overload for a half hour could seriously damage the transformer without reaching an oil temperature which might be reached daily after several hours of motierate load. The thermal relay responsive to copper temperature will permit the overload to be carried, if safe, but will protect the transformer otherwise. Fault-detecting relays include percentage and straight differential schemes similar to generator-protective relays but include provisions for the magnetizing inrush current and for transformer ratio and phase shift. Also, transformers are often included with the transmission line into a single protective zone. This is particularly true of the smaller sizes such as network transformers. Many small power transformers (600-3000 kva) are provided with internal protective links that act like a single-operation breaker as the fusible element whips through the oil in the top of the tank thereby disconnecting the transformer in event of internal trouble. Others are fused to provide disconnection from the line in event of transformer failure. A typical application of the CA relay to a star-delta connected transformer bank is shown in Fig. 4. Neglecting
Power-transformer protection in general includes overload devices to protect the transformer and fault-detecting devices to protect the system and limit damage in event of fault in the transformer. In the first category is the thermal relay immersed in the transformer oil but energized from a current transformer so that it responds to the copper temperature. This relay, obtainable only on new transformers, has alarm contacts to announce the approach of dangerous temperatures, and tripping contacts that close if an unsafe temperature is reached. Oil temperature indicators perform a somewhat similar function though less effectively. For large power transformers the order of magnitudes of copper and oil timeconstants are 5 minutes and 7 hours respectively. Thus,
Fig. 4Differential protection former bank with CA relays. formers are connected star on star
of a grounded star-delta transNote that the current transthe delta side and delta on the side.
exciting current, the top phase line current on the right hand side is made up of the difference of two transformer currents or the difference of the two top phase line currents on the left hand side. Consequently, it is compared with this difference current obtained by connecting the left hand current transformers in delta. These two currents are not exactly equal, even with a perfectly sound transformer bank, because of the magnetizing current.
Chapter
11
Application
353
Magnetizing Inrush-While under steady operating conditions, the magnetizing current amounts to only 5 or 10 percent, it may rise to several times full load current when a transformer is first energized, and decay rather slowly from this value; that is, it may be as much as full load current even a full second after the transformer is first energized. This magnetizing inrush current is fully displaced and hence, contains a large d-c component.S The inrush current is greatest if the switch is closed at the zero point of the voltage wave. Its magnitude depends also on the residual excitation and on the leakage reactance in the supply circuit and transformer primary. Data for determining the value of the magnetizing inrush is given in Chap. 5. Ordinarily the residual flux density is low when the transformer is first energized. However, when a severe fault occurs near a transformer at a time when its flux density is maximum (voltage zero), and if the fault is interrupted an odd number of half cycles later, the residual flux at the instant of re-energizing may approach normal density. As this requires the fault to start and stop at zero voltage, it is seldom fully realized. The rate of decay of the magnetizing inrush current depends on losses and is particularly slow when a large bank is paralleled with one already operating and quite near to a large generating station.20 The d-c component, which flows at first over the supply circuit, transfers to a circulating current between the two transformer banks, and this dies out very slowly because of the high L/R ratio. For example, when the magnetizing current has dropped to 50 percent of full load in a 60-cycle transformer having 0.25 percent primary resistance, the reactance to resistance ratio is 200/.25 or 800/l. The corresponding L/R ratio or time constant which determines the rate of decay of the d-c component is 2.1 seconds. The Type CA normal-speed differential relay most commonly used for transformer protection has a 50-percent differential characteristic and 2.5-ampere minimum trip. It is prevented from operating during the magnetizing inrush by the large restraint, the inverse time characteristic, and the braking action of the direct current on the induction disc. It is found adequate in all but the extremely rare cases where one large bank is paralleled with another. When the differential relay cannot, because of its inherent characteristics, avoid tripping on the magnetizing inrush, a timing device can be used, which desensitizes the protection during the timing interval by requiring a drop in voltage in addition to operation of the differential relay to produce tripping during the inrush period. This device, known as a magnetizing-inrush tripping suppressor, is used primarily with high-speed differential relays or with pilot-wire relays when a transformer is included as a part of the line. High-speed transformer differential protection (Type HA relay) is required in certain circumstances to coordinate with other high-speed system protection, particularly where stability is critical. It must be used with the tripping suppressor as outlined above. This unit is therefore built as an integral part of the Type HA transformer relay. Three-Winding Transformers are protected in the same manner as two winding transformers except that the
Fig. 5Single line diagram showing the arrangement of circuits when using the CA-4 relay for the differential protection of a three-winding power transformer.
Type CA-4 relay for this purpose has three restraining coils to be associated with the three transformer windings as shown in Fig. 5. Regulating Transformers-Regulating transformers for voltage and phase-angle control constitute a special problem because of the change in ratio taps during operation. Figure 6 illustrates the most modern differential relay protection for such a unit. A Type CAM relay, Fig. 2(p), having one disk and two electromagnets is arranged to trip if the current in the shunt-exciting winding of the regulator greatly exceeds the proper proportion of the series-line current. For example with a + 10 percent voltage regulator, a typical relay would operate for any current in the shunt winding greater than 11.5 percent of the
Fig. 6Typical large regulating
transformer
protection.
354
Application
Chapter
11
line current. This provides sensitive protection for the shunt transformer. In addition normal two-winding transformer-differential protection is applied around the entire unit, providing overlapping protection with the bus and line protection and guarding the series transformer. It is difficult to provide complete protection to the series transformer and tapped regulating windings. Optimum use is made of ground protection, for example as shown in Fig. 6. However, the possibilities of this protection vary with the arrangement of windings and whether grounded or not.
are used in the simple differential scheme, Fig. 10(a), with low-resistance leads to minimize the current transformer requirements. The formula specifying the requirement of iron cross section, turns and lead resistance for nonsaturation18 is.. ;=4.44 IT
Remote
Trip for Transformer
FaultsBecause
of
the high record of reliability of large power transformers, a circuit breaker between the high-voltage side of the transformer and the line frequently cannot be justified, purely for protection of the transformer, and the transformer is very little hazard to the line. However, an intermediate measure costing much less than the high-voltage circuit breaker is frequently provided to trip the remote circuit breaker (or breakers) necessary to clear in the event of a transformer fault. The transformer differential relay is sometimes used to initiate a remote trip signal over a carrier or pilot wire channel, particularly if the channel is already available for some. other purpose. Another method is to close a fast spring-operated high-voltage grounding switch in response to the relay indication. This trips the ground relays at the other terminals of the line, at the expense of some added shock to the system. 5. Bus Protection The advantages of bus protection in clearing faults rapidly from a system are well recognized by the industry and the provision of relay protection for major station busses has been standard practice for a number of years. The problems involved in such protection are also quite well known. One of the principal problems is the saturation of current transformers by the d-c transient component of the short-circuit current as in Fig. 7. In severe
Fig. 7False differential current caused by d-c saturation of current transformers during the condition of an asymmetrical short circuit. The d-c time constant for the case shown is 0.14 sec. The incoming transformers are assumed not to saturate. Thus the error current of the outgoing transformer is the differential current through the relay.
cases the d-c transient component may require 100 times as much flux capacity in the transformer as is required by the a-c component to completely prevent saturation. There are a, number of successful solutions to this problem, as well as the fault-bus scheme, 11*16 which completely avoids it. One is the use of large current transformers which will not be saturated by the d-c component. These
where T is the short-circuit current d-c time constant in cycles. I, is the a-c exciting current in the secondary, selected as the threshold of saturation. It would be taken as less than the relay setting by a suitable factor of safety. E, is the required secondary rms, a-c voltage corresponding to I, on the a-c saturation curve of the current This determines the needed iron cross sectransformer. tion and turns, or the iron if turns are fixed. I is the crest value of symmetrical subtransient current, secondary amperes. R is the secondary circuit resistance in ohms, including the transformer winding and leads up to the relay (or point at which all current transformers are paralleled.) The a-c flux is neglected. It is usually relatively small as the ratio of maximum d-c flux to a-c flux is 27rT or 37.7 for a G-cycle d-c time constant. For R =0.5 ohm, I = 100 amperes, T=6 cycles, and taking I, as one ampere, in considering a 5-ampere relay setting, E, becomes 1332 volts. A current transformer which would generate this voltage at 1.0 ampere exciting current is very large. Thus, this is a bull-by-the-horns solution, and the size, weight, and cost can be afforded only in the most important installations. However, it does provide the possibility of instantaneous tripping without any time delay. A method18 has been developed for calculating with reasonable engineering accuracy, the time-to-saturate with offset currents, and the time and current settings required to prevent misoperation with time-delay overcurrent relays and usual current transformers. Induction Type Overcurrent Relays-On busses of moderate time constant, say 0.1 second or less, and with somewhat better than average current transformers, satisfactory protection can be obtained with a straight differential scheme, Fig. 8, using a fast induction element. A small ratio of maximum to minimum fault is favorable to this application. Relaying times of the order of 3 or 4 cycles for maximum faults and up to 8 or 12 cycles on minimum faults can be obtained in some cases. As men-, tioned, the performance can be predicted.* However, the time delays involved in less favorable cases are frequently so long as to point the need of a better solution. Even on substation busses having a d-c time constant as short as 0.01 second, false tripping has been experienced with 300/5 bushing current transformers and standard induction relays in the connection of Fig. 8, with 4 ampere tap, 0.5 time lever, giving 0.15 second time at 10 times tap, and with a fault current of 13 000 amperes. While this has been overcome by changing to 1200/5 current transformers, nevertheless present practice would be to install ratiodifferential relays in these cases, providing both greater sensitivity and greater safety factor.
Chapter 11
Application
355
primary currents, a ratio of five volts per 1000 amperes being commonly used. These voltages, which add up to zero for through faults:and to a value proportional to the fault current for internal faults are joined in a series loop to the relay as shown in Fig. 10. The rt 1 percent tolerance
Fig. 8Bus-differential-current
relay scheme.
Multirestraint
Ratio Dierential
Relays-Relays
have been developed which will not operate falsely even when used with normal sized current transformers which saturate due to the d-c component of current. These are multirestraint relays, connected as in Fig. 9; however, their success is due also to exploitation of variable percentage characteristics, and the tendency of the d-c component to brake, rather than drive, the induction disk. No small part of this development is the reduction of the operating limits to a few simple rules which insure safe application. This relay scheme provides operation generally in from three to six cycles and can be set as low as one percent of the maximum through-fault current. Linear Coupler SchemeThe multirestraint relays just described may, of course, be used when the setting does not need to be as low as one percent of the maximum-through-fault current. However, on busses where a setting of four percent or more of the maximum throughfault provides the requisite sensitivity, a simpler and faster scheme (one cycle) can be used known as linear-coupler bus protection.21,27 The linear coupler is an air-core mutual inductance used directly in the primary circuit in the same manner as a current transformer except that the secondaries are usually connected in series, as shown in Fig. 10, instead of in parallel as are current transformers, Fig. 8. The secondary induced voltages are proportional to the
Fig. 10Linear-coupler
bus-protection
scheme.
Fig. 9Protection
of a six-circuit bus using two multi-restraint relays per phase.
within which the mutual inductances are held commercially, limits the maximum possible false differential to 2 percent. Thus the minimum setting of 4 percent allows a 2:l factor of safety. Solenoid elements, Fig. 2(a), are used for settings down to about 1500 amperes on a 6-circuit bus, and polar elements, Fig. 2(j), with saturating transformer and Rectox for lower settings. Impedance Schemes-Busses having reactors in a majority of the feeders and possibly in the bus-tie circuits provide the possibility of protection by impedance reor modified-impedance elements can lays. l3 Impedance be used, see Fig. 2, (i) and (k). For a fault on the bus, the maximum impedance measured is that of the arc which is taken as about 300 to 500 volts per foot for current above 500 amperes. Considering the possible arc length during the first few cycles of fault, a maximum arc voltage can be computed and this, divided by the minimum fault current gives the greatest fault impedance encountered for internal faults, that is, for faults on the protected bus. Provided this impedance is smaller than the impedance from the relay to a fault anywhere beyond the reactors, a basis of discrimination exists and the impedance scheme can be used. Two relay arrangements are used. When the bus tie circuits include reactors, separate impedance relays can be used on each generator or transformer feed to the bus, operation of any of which will trip the bus. This arrangement is most feasible when the generators and transformers are matched, acting as a unit, and the generators on the bus are either high and low pressure units of a single combination or are treated as a single generator. The other arrangement requires totalizing all of the main feeds to the bus and the use of a single set of impedance relays. The grouped main sources provide the possibility of a large false differential current for through faults on one of these main circuits. The voltage is also low under this condition
366
Application
Chapter
11
if the source circuits are not equipped with reactors. Usually a fast induction relay, Fig. 2(d), is used together with adequate current transformers in the main feeds so that it will not operate for through faults on those feeders. Combined impedance and fast-induction-element operation is then required to trip the bus. Directional relays can be used in a variety of ways, either as the basic protection, tripping when fault current flows into and not out of the bus,24 or as an adjunct for determining which of several bus sections included within a common differential protective zone is in trouble.13 Fault Bus-For new and certain existing segregated phase busses, the fault-bus scheme11,16 provides a distinctly different mode of attack to the bus protection problem. All metal parts to which the bus may flash are connected together and grounded through a current transformer and relay. This construction lends itself particularly well to metal-clad switch-gear. The entire cubicle or switchhouse is insulated from ground except for the ground connection through the current transformer. The simplicity of this scheme is strongly in its favor where the construction permits its use. However, it is sometimes difficult to secure overlapping protection with the adjacent system elements. Summary of Bus Protection-While personal preference, experience, and factors peculiar to a particular installation play a large part, some of the general factors that lead to the selection of one or the other of the several schemes are described below. One-cycle operation, simplicity, and savings in cable costs, as compared with the multirestraint scheme, are favorable to linear couplers. Quite adequate speed (3-6 cycles), the use of existing current transformers, the use of current transformers which can be used for certain other purposes also (such as back-up protection), simple application rules, and ability to set for minimum faults one percent of maximumthrough-fault sometimes eliminating need of separate ground relay, are all favorable to the multirestraint system. Existence of reactors and the cost or difficulty which would otherwise be involved of installing current transformers on all feeders, favors impedance schemes. The fault-bus scheme is limited to cases where the structure can be insulated from ground, but in these cases its simplicity is favorable. Simple time-delay over current frequently involves excessive delay, but if used with ordinary current transformers, it may be lowest in cost. The directional schemes are used to good advantage by some and have the advantage of securing fast operation with ordinary current transformers, but are considered less favorably by others because of the number of contacts to be coordinated for correct operation. In most cases, spring-operated, manually-reset auxiliary tripping relays are used, unlatched electrically by the main differential relays. These trip the necessary circuit breakers and provide lockout. 6. Transmission Line Protection
on the protective relay has become increasingly more difficult. However, developments in the protective relaying art have kept pace with the requirements. Through the introduction of improved relaying principles and better use of the old principles, high-speed action can be obtained on the complicated systems of today with better overall results than that previously possible on the simple radial systems. Starting with the induction-type overload and reverseload relay in about 1901, which used power for discrimination, the directional overcurrent relay with inverse-time characteristics was introduced in 1910. Later, in 1914, the definite minimum time characteristic was added. This simplified the relay coordination problem and is still used in the greater proportion of overcurrent relays today. The first impedance-, or distance-measuring relay, the type CZ, was introduced in 1922. Shortly after this the importance of speed in fault clearing, particularly with inter-connected systems, was beginning to receive merited attention and in 1929 the high speed impedance relay, type HZ, operating in one cycle and using the balance beam principle, was introduced31. At about the same time, circuit breaker operating times were lowered from about 24 cycles to 8 cycles. The reductions of overall fault clearing times that could be realized by these progressive changes in the arts2 are shown in Fig, 11. 64Starting with times up to 2 seconds for the slow-speed relays and 24 cycle breakers, the change to high-speed relays brought the time down to about 27 cycles with about one second in the end zones. Decreasing the circuit-breaker operating time to 8 cycles further lowered the overall clearing time to 8 to 10 cycles for about 80 percent of the line length but left times of about 27 cycles in the end zones. In about 1935 carrier current relaying passed out of the experimental stage511525354 and reached general acceptance, making available uniform high speed action throughout the entire section. Shortly later, in 1938, the Type HCB relay 62 based on symmetrical component principles made one-cycle operation practical over two a-c pilot wires. Summarizing, and referring to Fig 11, with 8 cycle breakers the total clearing time is under 0.2seconds for pilotwire or carrier current relaying, 0.2seconds (with 0.5seconds for the end zones) for high speed distance relaying and 0.2 to 2.0 seconds or longer for overcurrent protection, depending on the layout.
As systems have grown in extent and complication, from the simple radial systems of the early 1900s to the looped and interconnected systems of the present, the task imposed
Fig. 11Reduction of fault clearing time obtainable through the use of higher speed circuit breakers and relays.
Chapter
11
Application
Fig. 12Composite
power system illustrating
typical protective problems
and their solution.
A cross section view of the industry today shows all of these relay and circuit breaker types and speeds to be in current operation. Although speeds of new oil circuit breakers appear to have stabilized at 8 cycles except for special cases, there appear to be distinct fields for both normal-speed and high-speed relays. 7. Protection of a Typical System A composite diagram showing the typical transmission line conditions in many of the large systems is given in Fig. 12. For example the main transmission lines are shown as 220-kv, although they may be from 66 to 287-kv, bringing in power from a remote hydro plant or interconnecting with an adjacent system. These lines are equipped with the finest and fastest protection, high-speed distance, or carrier-pilot relays .54 Balanced-line protection36 may also be used if the lines are paired although high-speed distance or carrier relays are required to secure fast operation with one line out of service. The 33-kv circuits, often looped or interconnected, carry bulk power out through the territory served by the partic-
ular utility, to substations in the various towns and comA looped 33-kv circuit may use directional overmunities. current relays, Type CR, set with selectively higher time settings each way around the loop, as shown in Fig. 12. Impedance relays, preferably the step-type HZ, or alternatively the normal-speed impedance relays, Type CZ, may be used as in the 33-kv loop on the right. Impedance relays are particularly desirable if interconnections are contemplated as shown dotted. Loops involving short lines of 33- or 13.8-kv lend themselves well to pilot wire protection62163as in the center right. Induction-type overcurrent relays, usually with instantaneous trip attachments for operation at the higher currents, will be found on a majority of the radial feeders and Network feeders are cleared at 4-kv or 2.3-kv primaries. the load end by the network relays, essentially a reverse power form of protection. 8. Relay Symbols Relay symbols are useful in illustrating tection used for each element of a system. the form of proWith modifying
358
Application
Chapter
11
notations as to relay types and settings, these symbols compress the otherwise complicated picture of complete system protection into a form that can be readily visualized. The standard symbols are given in Table 3. Their use has been illustrated in Fig. 12.
TABLE 3 -RELAY SYMBOLS
Relative Number of different kinds of faultsThe relative numbers of different types of faults vary widely with such factors as relative insulation to ground and between phases, circuit configuration, the use of ground wires, volt age class, method of grounding, speed of fault conditions, qualclearing, isokeraunic level*, atmospheric ity of construction and local conditions. Thus the figures given below serve merely to indicate the order of prevalence and emphasize that there are usually a great many more line-to-ground faults than faults of other types. 5 percent Three-Phase Faults 10 percent Two-Line-to-Ground Faults 15 percent Line-to-Line Faults 70 percent Line-to-Ground Faults
Total 10. Overcurrent Protection The general plan of coordination with overcurrent relays on a radial system is shown in Fig. 13. The time shown in each case is the fastest operating time for a fault at the location of the next device in sequence. At lighter generating capacity the fault currents are reduced and all operat100 percent
Fig. 13Coordination
of overcurrent protection power system.
on a radial
9. Fault Frequency and Distribution About 300 disturbances (or one per ten miles) occurred per year in a typical system operating 3000 miles of 110-kv circuit. This system used mostly overcurrent and directional relays, and in a 4-year period experienced 2800 relay operations of which 92.2 5.3 2.1 0.4 The faults percent percent percent percent were were were were correct and desired correct but undesired wrong tripping operations failure to trip
were as follows: Conductors 56 11 11 11 11 percent percent percent percent percent
Lightning Sleet, Wind, Jumping Apparatus Failure Close-in on Fault Miscellaneous
ing times increase, but because of the inverse time characteristics of the relay the margins between successive relays also increase. Relays used with feeder circuit breakers must be coordinated with fuses of distribution transformers and with the main and branch line sectionalizing fuses.83 Several characteristic curve shapes are available in different designs of the induction-type overcurrent relays as illustrated in Fig. 11. These provide latitude in selecting the relay that coordinates best with the fuse curves at the current involved. The definite minimum time characteristic provides a ready means for coordinating several relays in series with only an approximate knowledge of the maximum current, and results in relatively small increase in the relay time as the fault current is lowered. It is used in the majority of overcurrent relay applications. The inverse and very inverse characteristics are sometimes more favorable where close coordination with fuses is required. They also make it possible to take advantage of the reduction of maximum fault current as distance from the power source increases. Several relays in series can be set for the same time for
*Number of storm-days per year,
Chapter 11
Application
Fig. 14Characteristics
of various induction
type overcurrent
relays.
(a) Type COH. (b) Very inverse-low energy relay, Type CO. faults immediately beyond the relay and still provide the requisite 0.25 second or more margin for fault beyond the next relay because of the lower current value for fault in that location. For example the timing on curve (b), Fig. 14, doubles when the current is reduced from 700 percent to 400 percent of pick-up value. Several settings of 0.3 second at 700 percent could be used in series, while still having 0.3 second margin between successive relays if the fault current dropped in the ratio 7 to 4 between successive locations. The choice of relays is also influenced in certain cases by the lower burden of the low energy and very inverse types.
(c) Inverse-low energy relay, Type CO. (d) Standard, definite minimum time, Type CO relay.
11. Normal-Speed
Impedance
Relay*
The time-distance tripping characteristic of the Type CZ normal-speed directional distance relay is illustrated in Fig. 15, which shows a number of line sections in series. This may equally well be a loop, the two ends of the section shown being at the same supply point. The tripping time of the relay increases in direct proportion to the distance from the relay to the fault, except that the minimum time is about x second for a fault at the relay. Each relay is
*The trend is toward the high-speed impedance relay described in Sec. 12 even for intermediate voltage transmission lines.
Fig. 15Time-distance curves of the Type CZ relay. The slope of the curve is changed by varying the resistance in series with operating time with zero the potential coil. The minimum voltage on the relay is about l/4 sec.
adjusted to trip in approximately 3/4 second for a fault at the next bus, except as will be noted. It is essential that for a fault near bus 4, breaker NO. 3 be tripped in preference to breaker No. 1. Thus the operating time of relay No. 1 must exceed that of relay No. 3 for fault at location No. 4 by one circuit breaker operating time plus margin. For 8-cycle breakers a reasonable breaker time plus margin is 0.4 second.
360
Application
Chapter 11
The operating time for faults anywhere on the system can be readily determined by drawing the straight lines representing the relay times, using whichever criterion rules in each case; that is, 0.75 second at the next bus or 0.4 second above the next relay at the second bus. The particular time values mentioned are typical only. The relay tripping time is independent of current magnitude once the overcurrent setting has been exceeded and timing thereby initiated. Thus variations in the amount and location of connected generating capacity, or switching lines out, does not materially affect the coordination of the distance type relays over the remainder of the system. The normal-speed type CZ relay is not usually employed on lines shorter than those in which at least 5 volts secondary result at the relay for a fault at the other end of the line. As the relay is normally subjected to full voltage and must discriminate on values between zero and that for a fault at the other end of the line, the operating forces approach the frictional forces below this limit.
12. High-Speed Impedance
Relays, HZ and HZM
The high-speed distance type relay has the step type time-distance characteristic illustrated in Fig. 16, obtained
Fig. 16Time-distance curves of the Type HZ step type, high speed distance relay. When carrier current is added the time is reduced to that shown dotted.
relay. The third-zone impedance element, which operates when either of the other two elements operate, is used to start the timer that closes first a second-zone timing contact T2, and later a third-zone timing contact, Ts. Thus for a fault in the first 90 percent of the section, known as zone 1, the contacts D and 21 operate, giving immediate high-speed tripping in one to three cycles, as indicated in the timing chart of Fig. 16. While the other elements also operate their action in zone 1 is unimportant because the circuit breaker has already been tripped. Thus, in zone 1 the tripping time is that of elements 21 and D. For the second zone, which extends approximately to the middle of the next section, contacts D, 22 and Tz in series do the tripping, provided the fault lasts for the time setting T2. If the fault is in the next section it will be cleared by the proper breaker in advance of 72 operation, although back-up protection is provided by the second zone setting extending into the next section. This also provides operation for bus faults if they are not previously cleared by busprotective relays. The third zone, corresponding to tripping through the contacts of elements D, 23, and Ys,completely overlaps the next section, providing complete back-up protection. It must of course be timed selectively with the Tz timing of the next section. The flexibility of this arrangement in molding its characteristic to various section lengths and breaker and relay times is apparent. The highly successful operation of several thousand such relays in service indicates that for practical systems, which of course depart in many ways from the simple ideal case represented in Fig. 16, the flexibility is sufficient to secure in general the operation outlined. For good operation the line should be electrically long enough so that there will be at least 5 percent voltage at the relay for a fault at the next bus, although in special cases successful operation can be obtained somewhat below this limit. In some cases the CZ characteristic lends itself better to coordination with other back-up protection, but the high speed of the Type HZ first-zone element is desired. For this purpose these two elements have been combined and make available the time-distance curve shown in Fig. 18.
Fig. 17Contact circuit for producing the stepped time-distance characteristic shown in Fig. 14. Z1-First Zone Impedance Element Contact. &--Second Zone Impedance Element Contact. Za-Third Zone Impedance Element Contact. Tz, Tt, timer contacts having separately adjustable time settings. &--Directional element contact.
by separate directional, impedance, and timing elements with contacts connected as shown in Fig. 17. There are a total of three balance-beam type impedance elements, each arranged with a current operating winding at one end of the beam and a voltage restraining winding at the other. When the ratio of voltage to current falls below the impedance setting of the relay high-speed action closes the contacts. The impedance elements 21, &, and 23 are set for successively greater distances. The directional element closes only for faults in the desired tripping direction from the
Fig. 18Typical time-distance characteristics of the HCZ relay. Note the slope of the CZ element necessary for different length sections to secure selectivity.
Modified Impedance Relay, Type HZM.41 The operating characteristic of the standard type HZ impedance relay is nearly independent of the phase angle between current and voltage. That is its reach-vs.-angle characteristic is a circle centered at the origin as shown in Fig. 19. This
Chapter 11
Application
361
Fig. 19Modified-impedance characteristics provides improved selectivity between heavy load swings and line faults.
characteristic provides adequate discrimination between load and faults in a majority of cases. However, increased use is being made of the Type HZM relay having the modified-impedance characteristics shown in Fig. 2(k) for the protection of long or heavily-loaded lines. It is necessary that the relays permit the maximum expected steady or swing loads without tripping. On long or heavily-loaded lines, and especially if high-speed reclosing is used, it becomes more and more difficult for the relays to distinguish between heavy swing loads and fault conditions. The modified-impedance relays provide the improved discrimination necessary in these cases. Their operation can best be shown by plotting on a single diagram the impedances corresponding to three things. 1. Faults on the protected line. 2. Heavy load swing conditions. 3. The relay characteristics. These are shown for a simple case in Figure 19. Considering the simplified case of a two-machine system with all impedances having the same R/X ratio, let the system impedance be laid out on a resistance-reactance chart as the line OABO representing the impedance from one machine to the other. The transmission line being studied is a tie-line constituting the section AB, shown heavy, of this total impedance. If R and X axes are drawn through A, as shown, the impedance with respect to these co-ordinate axes are those seen by the relays at A. Thus, the locus of impedances seen by the relay at A for a solid fault on the protected line consists of the line A-B, having impedances from zero up to the full line impedance. It can be shown that if the generator voltages at the two ends of the system are equal in magnitude and are at first in phase, then are moved out-of-phase, resulting in load transfer over the line A-B, the impedance locus viewed from A is along the line EL. That is, the impedance seen by a relay at A, for the load condition, is the impedance
vector drawn from the origin, A, to a point on the line EL. This line bisects O-O perpendicularly. The no-load points, corresponding to zero angle between machines are at infinity either way along EL, whereas the 180 degree out-of-phase condition is at the intersection, C. All intermediate loads are somewhere along the line L-L, the point L corresponding to power flow A to B and point L to power flow from B to A. The large circle with center at the origin, A, represents a pure impedance characteristic, as in Fig. 2(i). The smaller circle having the same reach beyond B is a modified impedance characteristic, Fig. 2(k). It can equally well trip for all faults on the protected line but is less likely to trip for very heavy loads or load swings. Successively heavier loads are represented by progressing through points L, D, and E along the load locus. The modified characteristic taken with directional element, trips for faults or loads in the cross-hatched area F, while the pure impedance element trips also in the areas G and G. Thus the modified characteristic permits heavier loads without tripping. When carrier relaying is used so that tripping requires closure of the relay at each end of the line, the small circle, for the relay at A can be advantageously shifted further to the right, so that the combined action of this relay and its mate at the other end of the line limits carrier tripping to the restricted zone between the two arcs at H and H. The back-up protection must be given sufficient timing to ride through swings or eliminated entirely except for a high-set, long-time element. The modified impedance element provides for independent adjustment of the radius of the circle and the location of the center as shown in Fig. 2(k) and hence makes possible the superior discriminating characteristics needed for long or heavily-loaded lines as outlined before.
13. Carrier Pilot Relaying

A pilot channel such as that obtainable by carrier current over the power circuit, or by a microwave beam, provides the possibility of simultaneous high speed tripping of both circuit breakers in one to three cycles for faults throughout the entire section. The significance of faultclearing speed on system stability is treated fully in Chap. 13. However there are, altogether, a number of reasons why carrier-current relaying has been employed in preference to other systems. These are: and increases the loads that can be safely carried over parallel interconnecting lines. imultaneous tripping is essential to 2. Quick Reclosing53S fast reclosing, the combination being particularly effective in increasing stability with single tie lines. 3. Shock to System-System shock, evidenced by voltage dips and dropping of synchronous load is lessened by fast clearing. 4. SystemDesign FlexibilityDesirable system arrangements that can not be relayed with sufficientspeeds otherwise,are possible with carrier relaying.
5. Growth of Faults-The more serious three-phase and double-ground faults generally originate as line-to-line or singleground and with sufficient speed of clearing the spreading to other phases is greatly reduced. 6. Ground Relaying Improved-On systems where high-speed 1. Stability-Simultaneous clearing improves system stability
362
Application
Chapter 11
ground relaying is not feasible otherwise, carrier pilot relaying provides an ideal solution. 7. Out-of-Synchronism*-The carrier channel provides means
for preventing operation of protective relays by power swings or out-of-synchronism conditions, yet clearing
Corresponding second and third zone ground over-current elements, Fig. 2(b), are provided. The carrier control has
the net effect faults within present of eliminating the protected is stopped all second-zone section. This because time delay for is accomplished of
flow
faults during such conditions.

8. Simultaneous Faults-The
by closing of the contact RR, Fig. 17, whenever a fault is

and carrier the direction
makes possible superiorrelay performanceunder simultaneously occurring faults.
added
basis for discrimination
is inward at each end of the line. The mechanism is indicated generally in Fig. 20. Occurrence of a fault anywhere within reach of the third zone re-
9. Joint Use-From an economic point of view joint use of the carrier channel for point-to-point communication, or for control or remote metering, may indicate the use of carrier pilot where the relaying requirements alone do not justify it. Carrier provided relaying fault power operates does on the not principle at any of tripping other ter-
quickly all terminals through which power flows into a line

flow out
minal. If fault power flows out at any terminal, that terminal continues to transmit a straight telegraphic carrier signal over the line, which is picked up by all other terminals on that particular line and prevents tripping. No time delay is necessary for internal faults since tripping for external faults can be prevented by the carrier signal. Since carrier is not transmitted for internal faults, the short circuiting of the carrier channel by the fault is of no consequence. Directional Comparison System-The type HZ or HZM directional comparison system utilizes the stepped impedance elements, Fig. 2(i) or (k), as its basic actuating elements, (see section on High-Speed Impedance Relay).
*Can also be accomplished without carrier,*9 with some differences.
lays, closes 23 starting carrier and setting up a circuit so that the receiver relay will close if carrier is removed by some other action. If the fault is internal, D and 22 close, stopping carrier transmission from either end of the line. The receiver relays, RR, immediately complete the trip circuit. If the contact circuit RR is opened manually the carrier can be cut out and a stepped-distance relay scheme remains. Thus the carrier is thought of as simply eliminating the time delay in the end zones, indicated by the shaded areas in Figure 16. The stepped-distance elements and an inverse-time ground current element provide the backup protection in this system. Phase Comparison System-The phase comparison system differs functionally in that the current directions or phases at the two ends of the line are compared rather than the power directions. Networks are used to derive a single-phase function of the line currents as in the pilotwire relay. This function may be referred to simply as the current at each end of the line since it is a measure of the several phase currents.
Fig. 20-Carrier
current relay system including
relays, carrier current transmitter-receivers,
coupling capacitors, and
chokes.
Dotted lines indicate symbolically the carrier controls MO-Master Oscillator REC-Receiver PA-Power Amplifier LTU-Line Tuning Unit
Chapter 11
Relay and Circuit-Breaker Application
363
If the currents at the two ends of the line are in phase and of fault magnitude, carrier is transmitted on alternate half cycles of current from either end of the line, resulting in substantially continuous carrier on the line from one end or the other. For an internal fault the current at one end of the line reverses or remains below the fault detector setting so that carrier is sent only half of the time. The relay is arranged so that this produces tripping. Figure 21 shows (heavy) the relative positions of the locally and remotely transmitted carrier pulses for internal
Fig. 22Typical
overall characteristics of the Type HKB carrier relaying system.
A few of the factors considered in determining whether to use one or the other of these systems are the following:
Favoring the Phase Comparison System
1. Can be added for high-speed

2. 3.
4. 5.
with any existing relay scheme for back-up. No potential transformers are required. Low-tension potentials are sometimes adequate for back-up relays. Inherently trip proof on out-of-step conditions. (Out-ofstep blocking relays are included in the other scheme if needed.) Not subject to trip by induced ground current from a parallel line. Back-up relays entirely separate. Can take either highspeed or back-up out of service without affecting the other.
protection
Favoring the Directional-Comparison System Using SteppedDistance Relays

2.
1. More generally, applicable to multi-terminal lines. Provides better discrimination between loads (tapped from lines) and faults. When transformers are tapped along the lines, it is not desired to trip the line for faults on the low-
tension feeders.
Fig. 21Relay tube circuit and typical wave diagrams of type HKB carrier relaying system.
3. Can trip with fault currents less than twice load currents. 4. More flexible for system changes.
faults (left) and external faults (right). Local pulses of operating voltage are applied to the relay tube every half cycle. For the internal fault the pulses of restraining voltage caused by the carrier occur in the opposite half cycle from the operating pulses. Hence pulses of plate current occur and result in tripping. For the external fault, restraining pulses occur during both half cycles, and since these pulses are, by design, greater than the operating pulses, no trip current results. With the entering and leaving line currents not quite in phase, some relay current flows. However, as shown in Fig. 22, a substantial phase difference can be tolerated without causing tripping. In this system, the carrier portion is purely pilot protection. Back-up protection must be added as an entirely separate entity. Stepped-distance relays, or simply directional-overcurrent relays, are used for back-up protection.
However, on many lines either system is entirely applicable and might equally well be used. Fig. 20 shows the complete equipment required for a carrier current relay system. Relays shown are of the directional-comparison type. The carrier components are the same with the phase-comparison type relays.
(a) The relays; practically the same as for high-speed distancetype protection except with the addition of the receiver relay, directional auxiliary relays, and out-of-step elements which are housed together. The Type HZ relay is shown in Fig. 23. (b) The carrier current transmitter-receivers operated from the station battery and with an output of 5 to 40 watts at 50 to 150 kilocycles when keyed. The outdoor set contains line-tuning equipment for matching through the coupling capacitor to the high tension line. When the set is located indoors it is connected by coaxial cable to the line tuning equipment in a separate housing located near the coupling capacitor. The connections to ground, and (c) The coupling capacitor. to the potential device if used, are through radio-frequency
364
Application
Chapter 11
including the relays, carrier set, coupling capacitor, choke and installation, costs considerably more than the terminal equipment required with the cost of the pilot-wire pilot-wire protection. However circuit increases almost directly
with the length of line; consequently there is an economic

dividing line between carrier-current and pilot-wire applications. The average dividing line is at about 10 miles but it varies widely with such factors as: the availability of existing pilot-wire circuits, the number of other pilot wire services with which the cost of the pilot cable can be shared, the inductive exposure conditions which determine the test voltage of the pilot cable, and the cost and complexity of the necessary carrier-current channel. The ideal pilot-wire relay systems should:
Fig. 23The Type HZ high-speed, three-zone impedance relay. The relay is arranged for use either as a conventional step-type distance relay for phase-fault protection, or in the directional comparison scheme of pilot-wire or carrier-current relaying.
1. Require only two pilot wires. 2. Provide complete phase and ground protection with a single relay at each terminal. 3. Permit wide variations in current transformer performance. 4. Be suitable for use over leased telephone circuits. 5. Not operate incorrectly when the system is out-ofsynchronism. 6. Provide adequate insulation between the pilot wires and the terminal equipment. 7. Have provisions for dealing readily with longitudinal induced voltages in the pilot circuits or with differences in station ground potential. 8. Have provision for supervising the pilot wires. 9. Operate at high speed.
chokes. Thus, at carrier frequency, the coupling capacitor is simply a series capacitor between the carrier set and the hightension line. (d) The tuned carrier-current choke or wave trap, of sufficient capacity to carry the line current but imposing a high impedance to carrier current of the frequency used. Its purpose is to prevent loss of the carrier energy into other sections, so that ample signal strength is available in the protected section.
A number of d-c or a-c pilot wire schemes based on directional comparison or on current differential have been used to a limited extent. For example an arrangement similar to the carrier-current protection described has been used with pilot wire. However by far the greatest number of pilot-wire relay applications employ the Type HCB relay thereby meeting all of the above requirements. The arrangement is shown in Fig. 24. At each end of the line a voltage proportional to positive-sequence current
Microwave Relaying-Either the directional-comparison or phase-comparison systems of relaying can be used over microwave channels as well as power-line carrier channels without significant alteration. However, because the microwave channel is not subjected to line faults it does not necessarily have to be used in a blocking manner, but is suitable also for transfer tripping. 14. Pilot-Wire Relaying
Pilot wire relaying is to the short transmission line what carrier current protection is to the long one. It provides uniform simultaneous tripping of the circuit breakers at both ends of a section, with all that such operation implies in the way of increased stability, lessened shock and damage to the system, and simplified coordination with other relay protection. In short high-voltage lines, discrimination is often impossible with distance type relays; pilot relaying by wire or carrier becomes the only method of discrimination not based on time delays. The cost of carrier-current protection is practically unaffected by the length of the line. The terminal equipment,
*Positive-sequence and zero-sequence segregating network. The secondary currents are fed in. An internal voltage is produced proportional to II +Klo. The network as viewed from the relay element terminals has this internal voltage and an internal impedance. A saturating transformer, not shown, is used between the network and the relay element.
Fig. 24Alternating current pilot wire scheme using the HCB relay. Simplified schematic. Only two wires are required and continuous supervision of them can be obtained as shown dotted.
Chapter 11
Application
365
Fig. 25Typical
operating
characteristics of the HCB relay for a phase A to ground fault with the currents two terminals 30 degrees out of phase, and a 2000-ohm pilot wire.
through
the
istics. Each relay contains a filter segregating the positiveand zero-sequence currents and combining the right amount of each into a single relaying quantity. Thus, by selecting the proper relay tap, tripping is obtained at any desired ground-fault current and phase-fault current. This reduction to a single differentiating quantity makes comparison over only two pilot wires possible. The ability to trip correctly even with badly mismatched current transformers is illustrated in Fig. 25. The vector positions of the sequence currents for various types GFfaults are illustrated in Fig. 32.
15. Ground Relaying

Overcurrent and directional relaying used for ground protection usually follows the same schemes as employed for phase protection. That is, the overcurrent ground relays for a radial system may be progressively timed, as in Fig. 13. For loop systems the directional element is added, polarized either by zero-sequence voltage or by a transformer bank neutral current which is proportional to zerosequence voltage. This results in progressive timing as in the loop circuit of Fig. 12. Stepped-distance relaying is likewise used but with limitations which will be discussed. There are a number of differences between ground and phase relaying. (a) The zero-sequence impedance of the average transmission line is 2 to 5 times its positive-sequence impedance, while the zero-sequence impedance of the source, comprising frequently only the transformer-bank impedance, may be lower than the positive-sequence impedance of the source. Therefore as the fault is moved along the line, the
Fig. 26The Type HCB pilot-wire relay. The relay has a single operating element, which functions for all types of phase and ground faults.
and a constant times the zero-sequence current is derived. The relay has variable-percentage differential character-
366
Application
Chapter 11
that
ground-fault current falls off more rapidly than the phasefault current and current magnitude can be used more for discrimination, at least where ground wires are used. (b) Usually there are many more sources of ground current than of phase current. This improves the selectivity obtainable with overcurrent ground relays. (c) A system may have several unconnected portions of zero-sequence network, in which case a ground fault in one section does not draw ground current from other parts. This makes the coordination of ground relays simpler than that of phase relays. (d) Fault resistance is likely to be much higher for ground faults than for phase faults on the higher voltage lines. At currents of 1000 amperes or more the arc voltage is 300 to 500 volts per foot so that a lOOO-ampere line-toline fault through a 5-foot arc involves a fault resistance of
used if the line terminates at the far end in a transformer will not pass residual current.
Reactance Relaying-has an inherent advantage over impedance relaying for ground fault protection in that the relay measurement is generally much less affected by fault
resistance. the relay erally If the currents to have less some with supplied from the two ends of
the line are not in phase, the fault resistance does appear to
reactance. a reactance Nevertheless, element than the error with an
in distance measurement caused by fault resistance is genmuch
impedance element. For ground relaying it is desired that the relay measure the zero sequence reactance from the relay to the fault.a2 It may be noted that the ratio of line-to-neutral voltage at the relay to zero-sequence current is:
E, -= IO &lZo+Ml+~2Z2 IO
approximately 2500 volts divided by 1000 amperes or 2.5 ohms. Compared with this, pole grounds, which may be the fault impedance for a ground fault, are usually in the 5 to 50-ohm range. A wire on the ground can have almost any fault resistance. Being unaffected by load current, a ground-current relay can be set lower than a line-current relay. Thus it can be set low enough to operate even though the fault current is choked down considerably by
fault parallel resistance. lines is important, although positive-sequence mu-
where 10, 11, I2 are the sequence current at the relay and ZO, 21, 22 the sequence impedances from relay to fault. Thus the zero sequence impedance can be measured asZ. = E, I121 IO I222 _ Lzl IO
(e) The zero-sequence mutual reactance between two tual reactance is usually unimportant. The zero-sequence
mutual reactance leads to circulating residual currents in one line for a fault in the other, even though the lines are part of two separate systems. It also interferes somewhat with distance-type ground relaying, although
The positive and negative sequence voltage drops from relay to fault are deducted from the line to neutral voltage, E,, by compensators and the resulting voltage divided by IO is a measure of the zero sequence impedance. A reactance element using this voltage and current willmeasure the zero sequence -reactance as desired. The type HXS ground reactance relay operates on this
principle, three elements being used to provide stepped-
methods are available for compensating this effect in some cases.32l36 Another factor of importance is that the fault, rather than the supply end of the line, is the source of zerosequence voltage. That is, the zero-sequence voltage tapers down from the fault towards the relay as outlined in Sec. 23 and illustrated in Fig. 30. These factors lead to difficulties in applying impedance or other distance measuring relays for protection against ground faults, While they have been used in some cases where conditions are favorable and where discrimination
would be even limited. more difficult by other means, their use is
no separate fault detector is required. Only one HXS relay is used for all three phases, the voltage of the faulty phase being connected to it by a type HPS faulty-phase selector relay illustrated in Fig. 2(t). To provide a single high-speed step, as for example where
distance protection. for fault conditions,
As there is zero-sequence current only
existing relays provide adequate backup protection,

HXL relay is used in conjunction of the line length, with the HPS
the
phase fea-
selector. This provides one ground reactance step, usually

set about 75 percent and a load-loss
Overcurren t Ground
Relaying-The
vast
majority
of ground relaying is essentially overcurrent, with direction where needed. The more common elements follow:
Type CO-Induction-overcurrent trip attachment. The instantaneous relay with instantaneoustrip is set below the max-
ture which opens the second breaker instantly after the feature utilizes three first breaker opens. The load-loss overcurrent elements to recognize by the closing of at least one back contact and one or more front contacts that a
fault is present load conditions and the far breaker is open. Under all three front contacts are closed. normal
Negative-Sequence
Directional
Relaying-The
imum ground current in the line for a fault at the next bus.
Nonsimultaneous closure of the circuit breaker poles during load switching may result in momentary ground current sufficient to operate the instantaneous ground relays, where the relays are set sensitively. This has been avoided by connecting a residualvoltage-relay contact in series with the trip circuit. The latter does not operate during load switching operations. As above but with induction-directional element, Fig. 2(e), controlling the induction overcurrent element. The directional element may be polarized either by residual voltage, Type CR, or bank-neutral current if available, Type CRC.
negative-phase-sequence directional element can frequently be used to advantage with an overcurrent ground relay to obtain selective clearing of ground faults. This results from three facts1. Only two potential transformers are required. 2. On solidly-grounded systems the negative-phase-sequence voltage at the relay may be of larger magnitude than the zero-sequence voltage at that point, hence, a more positive relay operation can be obtained. 3. The negative-sequence directional element is not affected by zero-sequence mutual induction from parallel transmission circuits.
Instantaneous-overcurrentelements, such as the SC, can be
368
Application
Chapter 11
In some very important stations complete duplicate sets of relays are used, operated from separate transformers to provide the back-up function. Also in certain large generating stations, with double circuit breakers separating major bus sections with power concentrations of the order of 20000-kw load, the back-up partial-differential circuit not only uses separate current transformers but trips the other circuit breaker of the double-breaker combination. The partial-differential protection can often be arranged so that it backs up generator, reactor, and bus protection as well as feeder protection. Since carrier, pilot-wire or distance protection is sometimes required to obtain 100 percent selective primary protection, it is not always possible to obtain selective action of the back-up relays. There is some trend in the case of pilot-wire protection of short-line sections to cut the backup relays in automatically when the supervision relays indicate that the pilot wires are inoperative. Some use is being made of carrier in a somewhat similar philosophy, to block certain back-up elements when carrier is being received.
process damage, resulting from sustained operation at low voltage. However, where continuous operation is more important than motor protection this feature is eliminated or used for alarm only. Thermal-alarm devices applied directly on the motors are becoming increasingly popular in attended stations where the operator can determine from the ammeters or other indications, the source of trouble.
18. Power House Auxiliaries

The power house is the heart of any electric system and its functioning rests on the motors which drive its fans, pumps, gates, and other auxiliaries. Hence reliability is paramount, and in addition to provision of a most suitable power supply and spare units for certain auxiliaries, much attention has been given to the relay protection. A recent survey of United States practiceM resulted in the following recommendations, attendance being assumed. The 2300-volt motors should use long-time-delay phase-overcurrent relays for overload and internal motor faults, set at approximately 150 percent of rated current. They also should be equipped with instantaneousovercurrent relays for short-circuit protection set above maximum-inrushcurrent. If the auxiliaries are transferred, the instantaneous relay must be set above a higher inrush current. On essential motors* the time-delay overload relays may be used for alarm purposes only, and the instantaneousrelays used to trip. In this case the time-delay relays can be set more sensitively than 150 percent of rated current. Low voltage motors (208, 440, 550 volts) should use a thermal device for overload protection, and an instantaneous-trip device for short-circuit protection.
17. Motor Protection

For the larger motors which are equipped with protective relays, the protection supplied varies with the importance of the service and whether automatic or not and whether attended or non-attended. Many motors 500 horsepower and larger, and some smaller ones are protected by long-time (geared, 40-second) induction-type overcurrent relays set in the neighborhood of 150 percent of maximum-load current, but with timing to permit the starting in rush of several-hundredpercent current for several seconds. A typical case might be 600 percent for ten seconds for an across-the-line-started induction motor, although these figures vary widely, depending on the application, and must be determined in each individual case. These relays will operate for a stalled motor, for internal motor faults, or for a heavy overload. The overcurrent relays are usually provided with instantaneous-trip attachments set above the motor-starting current to trip quickly for severe faults in starter, leads, or motor. Thermal relays are frequently provided in addition to the overcurrent relays, to provide more sensitive overload protection. Having a time constant of several minutes as contrasted with seconds for the overcurrent relay, they follow the motor temperature more closely. They will not protect the motor under stalled conditions, however, where the motor ventilation is missing; the overcurrent relay provides this function. In automatic applications, and where starting in reverse would be serious, a phase rotation relay is used which closes contacts to permit a starting only if the phase sequence is correct and voltage is present on all three phases. This does not guard against phase unbalances which might occur during operation, however. For this purpose, a phase-balance current relay, Fig. 2(p), is provided in many automatic schemes to trip if the 3-phase currents become excessively unbalanced. Time-delay under-voltage protection is frequently applied to guard against over-current or
In addition, the report notes the desirability of eliminating undervoltage protection except for alarm purposes, so that the loss of auxiliaries due to system disturbances will be minimized.
19. Industrial
Interconnections
When a line is tapped to an industrial plant having generation, it is common practice to segregate essential loads for operation from the plant generator and dump others in event of a line outage. If the same line is tapped for other plants, the problem arises of separating the plant under consideration from the line under conditions hazardous to its operation. One scheme in successful use on many industrial interconnections consists of separation based on any of three indications provided power flow has reversed and is toward the power company. The three indications are: under frequency, undervoltage, or generator overload. Any of these occurrences, provided power flow is away from the plant, is taken as sufficient cause for separating and at the same time dumping nonessential loads so that the remaining plant load may be brought within the capacity of the plant generation. The relays normally employed are : Induction-type overcurrent for generator overload. Induction-type under-frequency relay. Induction-type undervoltage relay.
High-speed-type three-phase directional relay.
The generator overload relay is directional controlled so *Essential motors are in this case defined as those motors whose failureresultsin the shut-downof generatingcapacity.
Chapter 11
Application
369
that it will not start timing unless direction in the interconnection has reversed. Directional relays are also used, without the voltage, frequency, or current fault detectors, for this purpose.
20. Three Terminal Lines

Lines having three or more terminals3* are generally more difficult to protect than two-terminal lines. Alternating current pilot wire protection is applicable in many cases although the limiting values of pilot-wire capacitance and resistance are less per terminal than for two-terminal lines. Carrier schemes, particularly of the blocking type are applicable to multi-terminal lines. However, sequential operation of circuit breakers occurs if fault current of appreciable magnitude flows out at one terminal for an internal fault near another terminal. The first-zone impedance element nearest the fault, acting independent of carrier, opens the first circuit breaker, after which carrier is stopped by the directional elements permitting clearing of the other circuit breakers.
the importance of the service. However, the major calibration tests are generally scheduled for periods more of the order of six months to two years. One year is a quite common period. There is a decided feeling that too frequent testing may cause more harm from mistakes and inadvertent damage than the good that is accomplished. The tests vary from the over-all or primary test in which current is passed through the primary of the current transformer, and the circuit breaker tripped by the resulting relay action, to much less complete checks. A quite usual procedure would be to remove the relays from service and test and calibrate on a load box, and to check the instrument transformers for continuity and grounds. The instrument transformer-relay circuit is grounded at only one pointQ3so that the intentional ground can be lifted for this test. If feasible the circuit breaker may be tripped by closing the relay contact.
23. Relaying tained
Quantities
and How They Are Ob-
21. Out-of-Step
Protection
Practically all utilities,Q0 except those consisting of steam stations connected rigidly together electrically, have experienced system instability. Most utilities have experienced some undesired operation of fault-protective relays as a result of system instability. Quite a number of utilities attempt either to block line relays from tripping because of out-of-step conditions, or to set the relays so that tripping will occur at a preselected point. Out-of-step blocking in conjunction with carrier relaying is the method most commonly employed. Synchronous frequency changers interconnecting two systems may suffer mechanical damage to shafts and couplings if permitted to operate with the systems out-of-step. The resulting power pulsations may be close to the natural frequency of the two-mass system composed of the two rotors with connecting shaft. Out-of-step relays are available which detect a slip cycle by the power reversal at high current and can be set to trip after two or three slip cycles, or before serious torque oscillations build up. Quick clearing of faults by modern 8-, 5-, and 3-cycle circuit breakers and high-speed relays is well accepted as a measure of prime importance in improving system stability and reducing damage and permanent outages. Case after case could be cited where these improvements have been realized as circuit breakers and relays have been modernized up to present-day standards. High speed reclosing has been made possible by simultaneous operation of circuit breakers at the two ends of a transmission line by carrier-current or pilot-wire relaying. This measure is generally accepted as economically of greatest benefit in improving stability and service reliability. Three-poleQ1 reclosing has been most widely used. However, there are a number of applications of single pole reclosingQ2which further enhance the stability by leaving the sound phases in service while the faulted ones are opened and reclosed.
The prime requisite of all protective relaying is a fundamental basis of discrimination, which has been variously referred to as a discriminating function or quantity, an operating principle, or a relaying quantity. This discriminating quantity must be one to which a protective relay can be made to respond, and one which separates the desired tripping values from the desired non-tripping values. The common discriminating quantities, such as current, voltage, time, impedance, direction, and power are well known, and the methods of obtaining them from current transformers, potential transformers, and potential devices are generally understood and described elsewhere.1*32 No general treatment of this subject can be given here. However, some of the more important characteristics of these quantities will be briefly outlined. Some special consideration will be given the newer sequence quantities arising from the method of symmetrical components, given in Chap. 2.
Voltages and Currents
During Fault Conditions-
Ten different faults of four kinds can occur at one point on the system: three-phase line- to-line double line-toground single line-toground ABC AB ABG AG BC BCG BG CA CAG CC
22. Testing and Maintenance

Routine tests are made by many companies at quite frequent intervals such as one to three months, depending on
When one of these four kinds of faults occurs along the line, the voltage and current relations at the relay are somewhat as shown in Fig. 28. For a three-phase fault the currents are balanced and lag the line-to-neutral voltages by the impedance angle of the line. In an average highvoltage line this angle is about 60. The addition of fault resistance tends to lower it. For line-to-line fault, say BC, the current in line B lags the collapsed BC voltage by a line impedance angle of about 60. For a two-line-to-ground fault, for example BCG, a similar situation pertains, except that the line-to-neutral voltages B and C also collapse to an extent depending on how solidly the system is grounded. Consider two easily visualized cases. If the system is
370
Application
Chapter 11
Fig. 28Typical
unbalanced conditions of current and voltage occurring during various types of fault on a three-phase system with an angle of sixty degrees. Load current flow neglected for fault conditions.
grounded through a very high impedance (2, very high), the currents will be nearly the same as for a line-to-line fault except for a small added ground current. However, the fault will establish the mid-point between B and C phase voltages as ground potential. Another example is the case Z1=Z2=Z0 for all parts of the system. For this case the phases are independent and the three-phase systern acts exactly like three independent single-phase systerns. The two-line-to-ground fault BCG, is the same as the three-phase fault except that phase A voltage is not collapsed, and only load current flows in phase A. The majority of systems fall between these two limits. For a single line-to-ground fault on phase A, the corresponding line-to-neutral voltage collapses and the phase A current lags the line-to-neutral voltage of phase A by the impedance angle of the line-ground-return circuit including the fault impedance. If Z0 = Z1 = Z2 throughout the system this fault will not influence the other two phases. If Z0 is higher than 21, corresponding to high impedance grounding, the condition of ungrounded operation or full displacement of the voltage triangle is approached.
Single- Phase
Directional
Element
Response
Considering the different kinds of faults, and also their occurrence with symmetry to A, B, or C phases, the angle between voltage and current for a fault on the line varies over rather wide limits. In using a single-phase directional element, as in the CR, CZ, or HZ relay, a particular voltage must be associated with a particular current. 30 Connections. One of the common connections associates the phase A current with the phase CA voltage and is known as the 30 connection, because at unity power factor under balanced three-phase conditions the current leads the voltage by 30. A watt-type* directional element
closes for current from approximately 90 ahead to 90 behind the voltage applied to its potential coils. For a threephase fault on a 60 impedance-angle line the current lags behind its unity power factor position by 60. With a 30 connection it lags the reference or polarizing voltage used on the directional element by 30. Fault resistance (plus any modification of the relay characteristics by lagging) brings the fault current nearly to the maximum torque position. For the other kinds of faults on different phases the current is shifted one way or the other, but the wide closing band of the relay allows for this variation. The 30 connection uses star currents and delta voltages. The same system is followed in naming other connections, although the relay used, including its phase-shifter if any, does not always have a closing zone for current from -90 to +90 with respect to voltage. The SO Connection uses delta currents and voltages; the IA- IB current being used with the phase CA voltage. A relay with a closing zone approximately +90 to -90 is used. Delta-connected main or auxiliary current transformers are needed to obtain the delta currents. The 90 Connection uses star currents and delta voltages; the phase A current being used with the phase CB voltage. In this case, however, a 45 voltage advancing phaseshifter is employed with the relay element giving it for star currents a closing zone approximately from 135 ahead to 45 behind the delta voltage. For a three-phase fault on a 60 impedance-angle line the phase A current leads the
*Other directional elements may have their closing zone shifted as much as 45 in the leading or lagging direction. The element used with the 30 connection may be a watt type or may have its closing zone shifted 10 to 20 in the lagging direction.
Chapter 11
Application
371
phase CB voltage by 30, and a small fault resistance would swing it toward 45 leading. The closing zone extends 90 either side of this position and affords optimum opportunity for the relay to give correct directional indication with other kinds of faults. Usually any of these three connections gives correct directional indication although in individual cases advantages can be found for one or the other, depending on such factors as the impedance angle of the line, the possible fault impedance, and the likelihood of an undesired amount of directional element operation caused by leading load currents near the directional boundary. For distance carrier relaying using single-phase directional elements the SO connection using delta currents is preferred since the same delta current is used on the impedance element. 11
TABLE &-SEQUENCE POWER RESPONSE OF THREE-PHASE DIREC, TIONAL ELEMENTS
Fig. 29Diagrams showing the relative directions of positive-, negative- and zero-sequence power during fault conditions. The Chart (b) indicates what part of the system fixes the power factor for each sequence.
*These are the commonly used 30, 60, and 90 connections, **Other phases connected symmetrically in sequence A, B, C. A new two-element 90 connection giving the desirable P-N torque. Two-wattmeter connection with zero-sequence current removed by a filter. Two-wattmeter connection. (4) Parasitic torques. Note: Torque is the real part of the expression in the last column. P =E&; N =Eu% Z = Ed. If the system were pure resistance throughout all Es and Is would be in phase, or phase opposition. For faults, N and Z would be negative at the relay, and P positive, but all three power terms would be pure scalars. The arrows show the vector position (not magnitude) of the torque expressions for this idealized pure resistance system case, taking into account that the values of N and 2 are negative as shown in Fig. 29. If instead of being pure resistance the system were pure 60 impedance angle throughout the effect would be to rotate all currents negatively 60. leaving all voltages unchanged. As the power terms are of the form El this will rotate similarly all the P, N, and Z quantities and hence the torque vectors. The real components of these vectors are the torques. Hence conclusions can be drawn as to whether the torques associated with P, N, and Z are additive and how much voltage phase shift is needed for optimum condition on a system of given impedance angle. (1) (2) (3)
impedance element operation is caused by a line-to-ground fault its associated directional element is influenced by fault current. This overcomes any possible load current effect.
Three-Phase
Directional
Element
Response-The
same connections are used with three-phase directional elements*. In this case another factor influences selection of the connection. Table 4 shows the functions of sequence power to which various connections respond. As shown in Figs. 29 and 30, positive-sequence power flows toward the fault; negative- and zero-sequence power flows away from the fault, since the fault is the source of negativeand zero-sequence voltage. Therefore the positive-sequence
*This discussion relates to three independent single-phase elements on the same shaft. No three-phase rotating field is involved.
372
Application
Chapter 11
Fig. 30Simplified equivalent system showing the sequence voltage distribution during fault conditions.
minus the negative-sequence power at the relay (P-N) is more positive for a fault out on the line than is (P+N), and hence provides a better fault directional indication, having higher torques for fault currents in relation to those due to load currents. In table 4 the connections are shown in the first three columns, and the resulting three-phase torque on the element in the last column. The relative phase positions of the torques produced by positive- and negative-sequence power are shown in the 4th, 5th, and 6th columns, taking into account that when the impedance angles are the same throughout a system, the values of N and 2 are opposite to P under fault conditions. The positive-sequence torque vector is drawn in the position of the conjugate of the current vector to produce maximum torque, the horizontal axis being the unity power factor position. Thus for the star-voltage, star-current, 0 connection maximum torque is for unity power factor. If a watt responsive element is used the voltage must be retarded 60 by a phase-shifter to obtain maximum torque for current lagging 60. The star-voltage, star-current, SO connection* can be obtained with the phase A current associated with the negative of the phase B-to-neutral voltage, the other phases being symmetrically connected. This is perhaps the best star-current, star-voltage connection as no phase shift is required to get maximum positive-sequence torque with *This is not,the commonly referred to 60 connection.
current lagging 60. Also the negative-sequence torque is maximum for current lagging 120 and hence gives 50 percent of the maximum possible assistance to positive-sequence torque for 60 system-impedance angle. As the system-impedance angle is likely to be above 60, this is quite favorable. The zero-sequence torque is also in the right direction though not maximum. The delta-voltage, star-current, 30 connection has adverse negative-sequence torque, while the 90 connection is ideal in this respect, the positive- and negative-sequence torques have their maximums in the same direction for 90 lagging current. The usual 45 voltage-phase-shifter brings the maximum torque to a desirable point and maximum assistance is secured from the negative-sequence torque. The delta-voltage, delta-current, 60 connection, like the corresponding star-star connection, has good negative-sequence torque in the proper direction for a 60 impedanceangle system. It has no zero-sequence torque. Maximum positive-sequence torque occurs for a 60 impedance angle without using a phase-shifter. In addition to symmetrical connections, one unsymmetrical connection is worthy of note in that it is capable of securing the desirable P -N accurately. It uses the B minus C current with the CA voltage and the A minus C current with the CB voltage. It has maximum torque for a 90 impedance-angle system and hence can be used to advantage with a 45 voltage advancing phase-shifter similarly to the better known 90 connection of three elements. Impedance Measurement-Referring to Fig. 30, the difference between E1 and Ez at the relay is the positive-sequence drop from the relay to the fault plus the negativesequence drop back to the relay. Recognizing that & = 21 for the line, it can be readily shown that the line impedance to the fault is: z -- Ed%
lk-I2 (1)
This applies for three-phase, line-to-line, or double lineto-ground faults. For line-to-ground faults a higher impedance than 21 is measured by the ratio (E1--Z&)/(11 -2r2). From Eq. (1) the delta connection is derived
Ib = Io+a211+aI~ I,= lo+aIl+a212 Ib-I,=(a2-a)(Il--2) Eb-E,= (a-a)(&-EJ Et,-% ErE,_Z -z----Ib-Ic
(2) (3) (4) (5)
11-12
(6)
The last expression shows that for fault at a given location the delta voltage divided by the delta current is the line impedance Z1 for three-phase, line-to-line and double line-to-ground faults. As shown previously a higher value is measured for line-to-ground faults. Lewis and Tippett32 give the fundamental basis for distance relaying in the most comprehensive paper on this subject. Among other things it is brought out that use of delta current and delta voltage on the impedance element, for example the A minus B current with the BA voltage, as This refers to torque due to negative-sequence currents and voltages. Actually it is torque in the positive direction.
Chapter 11
Application
373
outlined in the preceding paragraph, avoids a 15 percent difference in distance measured for line-to-line and threephase faults, which is present if only one line current (star current) is used. Use of Sequence Quantities*-In using sequence quantities the point of view should be developed, first, that the fault is the source of negative- and zero-sequence voltage and power, and that negative- and zero-sequence power (volt-amperes at system impedance angle) flow away from the fault at the relay location; second, that the sequence voltage is measured with respect to the bus-of-no-voltage or point n in the particular sequence network considered. These relations are brought out in Figs. 29 and 30.
Sequence
Voltage
Distribution
During
Faults-
In Fig. 30 the voltage gradients are shown very generally. For a three-phase fault the voltage tapers off from the generator to the fault. For a line-to-line fault the positivesequence voltage tapers off until, at the point of fault, it equals the negative-sequence voltage, which in turn tapers to zero back at the generator neutral, or point back of which there is no impedance to negative-sequence current. In some cases this may be an infinite bus. For a double line-to-ground fault the positive-sequence voltage again tapers off to the point of fault where it equals the negative- and zero-sequence voltages. These taper to zero in going back through the network until a point of no voltage of the respective sequence is reached. At a line-to-ground fault the positive-sequence voltage is the negative of the sum of the negative- and zero-sequence voltages and these taper to zero back through the network. It is well to note that if the zero-sequence impedance is high (high-impedance grounded system), the zero-sequence voltage is nearly equal to the normal positive-sequence or line-to-neutral voltage for a line-to-ground fault, and approximately half as much for a double-line-to-ground fault where the generated voltage divides between the positiveand negative-sequence networks, thereby applying about half voltage to the zero-sequence network. As a result on lightly grounded systems all zero-sequence and residual voltages and currents are approximately half as much for double-line-to-ground as for line-to-ground faults. Sequence-Segregating Filters-Sequence currents and voltages may be segregated from the corresponding line currents and voltages by segregating networks or filters. The methods of obtaining zero-sequence currents or voltages are already well known as these quantities are simply one-third of the corresponding residual quantities. Typical sequence-segregating networks are given in Fig. 31. The performance of each network is expressed by giving its equivalent circuit112and also by giving the equations of operation. Polyphase networks for segregating positive- and negative-sequence voltage are shown in parts (a) and (b), and are useful for operating a polyphase device in response to either of these quantities. The positive-sequence filter is also useful for obtaining a balanced three-phase supply from an unbalanced (or single-phase) supply. The remaining filters shown all have single-phase output. Parts (c) and (d) are auto-transformer type voltagesegregating networks and parts (e) and (f) are the all-im*Refer to Chapt. 2.
pedance type and require a special potential transformer connection. The star series transformer connection for obtaining zero-sequence voltage is shown in part (g). Parts (h) and (i) are three-winding-reactor type current filters whereas parts (j) and (k) use an auxiliary current transformer to produce the reactive drop due to B minus C current in a single-winding reactor. Note: The Type CRS negative-sequence directional relay uses filters (d) and (i) for negative-sequence voltage and current respectively. Parts (l), (m), (n), and (0) are all impedance-type current filters, (1) and (m) being suitable only when there is no zero-sequence current; (n) and (0) are not affected by zerosequence current, but require double the number of current transformers. Part (p) is a zero-sequence current filter, which is merely the neutral connection of star-connected current transformers. A combination positive- and zero-sequence currentsegregating filter is illustrated in part (q). This filter is used in the Type HCB pilot-wire relay.62 The relative weighting of zero and positive sequence is determined by the relative magnitude of & and &. For example, if it is desired that the same internal voltage Ei be produced by one-tenth as much zero-sequence current as positive-sequence current,, the zero-sequence weighting factor k must be set equal to 10. Then the required value of Ro may be determined as follows:
R. = $kRl= 6.67R1
The relative phase positions and magnitudes of various sequence quantities of the reference or A phase vary with the type and phase of the fault. The response of a combination filter varies accordingly. Fig. 32 illustrates the relative positions and magnitudes of the vectors comprising the quantity I,+k10 on a system for which the groundfault current is one-tenth of the phase-fault current, and using a zero-sequence weighting factor of k = 15. The 11 vectors have been magnified somewhat in the line-toground fault diagram to make them visible; their actual
Fig. 32Vectors comprising the relaying quantity Il+Klo shown for K== +15. 11 vectors magnified in line-to-ground fault cases.
376
Application
Chapter 11
length being one-fifteenth of the krOvector. The combination, II+klO, is the discriminating quantity used in the Type HCB pilot-wire relay which in effect totalizes the two ends of the circuit. It is a single quantity having, for a majority of systems, a much greater value for fault conditions than for load conditions and thus is an ideal discriminating quantity.
24. Reclosing
Many of the faults occurring on power systems are transient in nature and if the circuit is opened momentarily, permitting the arc to become extinguished, the circuit can be reclosed successfully. The necessary power-off time for deionization of the arc is given in Chapt. 13. For example, Logan and MilesSs have found that on the Georgia power system the number of successful reclosures is as follows. Number of trip-outs. . . . . . . Successful reclosures 1st immediate. , . . . . . . . . 2nd 15-45 seconds. . . . . . 3rd 120 seconds. . . . . . . . . Circuit lockouts. . . . . . . . . . 10090
8400
100%
83.25%
1084 143
553
10.05% 1.42a/,
5.28%
This knowledge is used in a variety of ways. Many radial distribution feeders are provided with reclosing relays. A very common arrangement, using the Type RC reclosing relay shown in Fig. 33, provides for one immediate and sev-
era1 time-delay reclosures. In the event of a tripout after the third reclosure the line is locked out until the relay is reset manually. However if the line holds, even on the third try, the reclosing relay resets automatically, and is prepared to repeat the same performance at a later time. For a feeder sectionalized by a number of fuses, the replacement of a fuse involves a service trip. However, an opening and reclosing operation interrupts the current before any fuses blow, and if the fault is transient, the service trip is avoided. Single-shot reclosing which is also widely used may be accomplished by the Type SGR-12 reclosing relay. As shown by Logans data, it takes care of the large proportion of cases. Also a self-reclosing single-pole circuit breaker is used principally on single-phase feeders to perform the single-shot reclosing function without the use of relays. Multiple-shot fuses are also used but require delayed action of subsequent fuses and necessitate refilling manually after each operation. On tie lines or single lines serving important industrial loads reclosing is used for quite another purpose, namely, to keep the systems from going out-of-step or to prevent loss of essential loads. This phase of the problem is covered in Chapt. 13.
III. CONTROL SCHEMES

To secure the system benefits expected from the circuit breakers and protective relays, it is essential that the control scheme used result in prompt and reliable tripping of the circuit breaker when the relays indicate a fault within their protective zone. Two factors are involved: a reliable source of control power, and a scheme that permits the breaker to trip free in spite of all manual or automatic closing agencies. A control battery provides the most reliable source of tripping energy, and is connected to the shunt-trip coil by the protective relay contacts as shown in Fig. 34. As
Fig. 33The Type RC recloslng relay. A single instantaneous plus several time-delay reclosures can be initiated with this relay.
Fig. 34A protective relay and circuit breaker. When the breaker opens, the auxiliary switch interrupts the trip circuit to prevent burning of the relay contacts.
Chapter 11
Relay and Circuit-Breaker by an
Application
377
the circuit breaker opens, the trip circuit is broken auxiliary switch linked to the circuit breaker.
25. Electrically
Trip-Free
Scheme
breaker remains open after the circuit breaker has once closed in and tripped out. This situation continues until the control switch is restored to neutral or the closing contact of the automatic device opens.
Fig. 35 shows a typical circuit-breaker control scheme, one of the several commonly used, known as the X-Y relay scheme. A station battery provides power for closing and
26. Mechanically
Trip-Free
Arrangements
Fig. 35The X-Y
scheme of circuit breaker control using battery.
C-Closing contact of control switch. T-Tripping contact of control switch. X-Closing contactor or relay. CC-Closing coil of main circuit breaker. Y-Releasing contactor or relay. P-Typical protective relay contact. B-Auxiliary switch (closed when main circuit breaker is closed). W-D-C shunt trip coil of main circuit breaker.
If a circuit breaker is to be closed manually against a possible fault, it should be mechanically trip-free from the The mechanically trip-free feature proclosing linkage. vides somewhat faster tripping for three reasons. First, the circuit breaker contacts can be tripped free anywhere in the closing stroke without waiting for the closing current to be cut off before acceleration towards the open position can start. Second, with the contacts tripped free from the closing solenoid the mass to be accelerated is less. Third, because of eddy currents the flux in the closing solenoid does not decay immediately when the circuit is opened; thus there is appreciable magnetic retardation in the opening of a mechanically non-trip-free breaker which has just been closed. High-speed reclosing requires that the circuit breaker be mechanically non-trip-free so that the contact motion can be arrested before the full open position and the breaker declosed again. To meet this need without encountering layed opening if the reclosure takes place on a permanent
tripping. The control switch merely controls the application of battery power to the closing solenoid or shunt-trip coil as desired. The protective relay, when it operates, applies battery power to the shunt trip coil. The X-Y relay scheme prevents pumping and makes the circuit breaker electrically trip-free. The closing contact of the control switch picks up the X relay that energizes the closing solenoid of the circuit breaker. As the circuit breaker reaches its closed position, an auxiliary switch B energizes the Y relay that seals in through its own front contacts. The Y relay contacts shunt the X relay, which opens and interrupts current to the circuit breaker closing solenoid. If the circuit breaker trips automatically when it is closed in on a fault, it will open and will not reclose even though the operator holds the control switch in the closing position. The X relay remains shunted by the Y relay until the control switch is returned to the neutral position. The trip-free relay scheme provides a similar action through the use of a specially designed contactor for controlling the heavy current to the breaker-closing solenoid. The moving contact assembly of this contactor is tripped free from the operating armature by a release coil energized by an auxiliary switch when the circuit breaker reaches its closed position. Thus even though the closing contact of the control switch or of an automatic closing device remains closed, thereby holding the operating armature closed, the circuit to the closing solenoid of the circuit
Fig. 36A-C
tripping
echemee.
378
Application
Chapter
11
fault, a mechanism has been designed that is mechanically trip-free on the second opening but not on the first, except when the circuit breaker is initially closed.
27. A-C Tripping

In less important locations, where the cost and maintenance of a control battery is not justifiable, various forms of a-c tripping are employed for the smaller, lower voltage circuit breakers. Some of the more desirable of these are shown in Fig. 36. The series-trip and transformer-trip schemes are used where the accurate magnitude and timing characteristics of a protective relay are not required. The transformer trip is used where the primary voltage or current is too great for the series trip. The capacitor trip and a-c shunt trip require a source of a-c control power. The capacitor trip is much to be preferred in most cases because its ability to trip is not impaired by the momentary drop in voltage at the time of a fault. The a-c supply is taken from the source side of the circuit breaker so that the capacitor is charged before the circuit breaker is closed. A-c shunt tripping can be used only where the reduction of voltage at time of fault on the protected circuit will not prevent tripping by some tripping agency. The circuit-opening relay scheme and the tripping-transformer scheme are similar to the transformer-trip scheme in that the line current transformer supplies the trip-coil energy. However, a protective relay is added and must be supplied by the same or by a different current transformer. The trip coil imposes a heavy burden on the current transformer, and there is a definite lower limit to the primary current at which tripping can be secured. The relay must, of course, be set above this value.
against overloads and short-circuits. Separate protective equipment is required for utilization devices such as motors. For more important and higher capacity circuits, metalenclosed, drawout assemblies of air circuit breakers, as in Figs. 38 and 39, are used. Typical applications are found in main feeders of the lighting and utility circuits described above, and for the low-voltage power circuits of such buildings as well as industrial plants and generating stations. These breakers may be operated either manually or electrically (under some conditions only electrical operation is recommended) and may be obtained with directacting series overload trips or relays which will give selec-
IV. APPLICATION
OF CIRCUIT
BREAKERS
The application of circuit breakers to power and lighting circuits involves the choice of the type of breaker and its mounting or housing as well as determination of the specific ratings required for the particular service.
28. Typical Circuit Breaker Construction Practice
and
LOW Voltage Circuit Breakers-Circuit breakers intended for service on a-c circuits up to 1500 volts and d-c circuits up to 3000 volts are classified as low-voltage breakers.lo8 For such service air breakers have many advantages and are generally used in preference to oil breakers. They are inherently fast in operation, free from fire hazard, require little maintenance on repetitive service, and, because of the low voltage, are simpler, more compact, and easier to handle than oil breakers. For low-capacity branch lighting and utility circuits such as are found in commercial and public buildings, small, molded case thermal-breakers are grouped in panelboards such as that shown in Fig. 37. Such breakers are usually operated manually and are available in ratings up to a maximum of 600 amperes load current and 25 000 amperes interrupting current. They provide automatic inverse-time overload tripping to protect circuit wiring
Fig. 381600 ampere, 600 volt, Type DB-50 air circuit breaker -50,000 amperes interrupting capacity.
Chapter
11
Application
Fig. 41Typical
metal-clad drawout switchgear indoor service.
for 13.8 kv
tive isolation of a faulty circuit. The metal-enclosed gear is factory assembled and tested and provides maximum reliability, safety and ease of maintenance with minimum interruption to service. These breakers may be used to provide control as well as running overload and shortcircuit protection for individual motor circuits. Even where unusual atmospheric conditions are encountered (see Sec. 30), low voltage air breakers may be used if they are mounted in suitable sealed enclosures.
For indoor service at from 1.5 to 15.0 kv and up to 500 mva interrupting duty, magnetic-type air breakers in metal-clad assemblies have become predominant, although metal-clad oil breakers are also used under adverse atmospheric conditions. A typical breaker and assembly are shown in Figs. 40 and 41. Although interrupting time and space required are the same for medium voltage air and oil breakers, the freedom from oil-fire hazard and lower maintenance on repetitive service are distinct advantages of the air breakers. Many such breakers are used, both for power and lighting feeders and to control individual large industrial or powerhouse-auxiliary motors. For indoor service at interrupting ratings above 500 mva, and for any rating at voltages between 15 and 34.5
Power Circuit
Breakers Medium
Voltage
Circuit breakers intended for service on a-c circuits above 1500 volts are classified as power circuit breakers.
Fig. 40Magnetic type air circuit breaker4160 volts, mva interrupting capacity, type 50-DH-150.
150-
Fig. 42Compressed air circuit breaker in station 15-H kv insulation class.
cubicle-
380
Application
Chapter
11
kv, compressed-air breakers mounted in station cubicles have become standard. A typical unit is shown in Fig. 42. Circuit-breaker practice in outdoor substations is more varied than in indoor service because there is a greater range in the requirements. In rural and outlying substa-
tions both normal load and interrupting kva are relatively low, and such factors as fire hazard, space requirements, appearance and rapid maintenance may not be critical. For such service frame-mounted oil breakers with open buses and disconnecting switches are frequently used because of lower cost. A typical installation is shown in Fig
43.
For suburban and urban outdoor service up to 15 kv and up to 500-mva interrupting duty the many advantages of metal-clad oil-less switchgear (freedom from oil-fire hazard, compactness, appearance, ease of maintenance and flexibility) have resulted in the use of such gear for the
_ .~-~~
Fig. 43Installation view of outdoor 23 kv, 250 000 kva, oil circuit breakers.
Fig.
4534.5-kv,
lOOO-mva, frame-mounted, breakers.
oil
circuit
Fig. 44 Typical outdoor coordinated
unit substation,
majority of these substations. The metal-clad gear is usually throat connected to the transformer(s) although roof bushings are sometimes used, especially with singlephase transformers. A typical installation is shown in Fig. 44. The cost comparison between open gear and metal clad varies with the voltage and kva rating, the type of open structure used, the cost of real estate, the labor facilities of the utility, and the method used in estimating overhead and fixed charges. Oil circuit breakers are used where severe atmospheric conditions are encountered, either frame mounted or in metal-clad structures. For outdoor service at interrupting ratings above 500 mva and for all interrupting ratings at voltages between 15- and 34.5-kv, oil breakers are essentially standard. A typical installation is shown in Fig. 45. High- Voltage Breakers-Almost all circuit breakers rated above 34.5 kv are mounted outdoors and are oil
Chapter 11
Application
381
Standard values of rated voltage are based on operation at altitudes of 3300 feet or less. Standard equipment may be operated at higher altitudes if the maximum operating voltage is not more than the maximum design voltage times a correction factor, as follows: Altitude in Feet Voltage Correction Factor
3300 4000 5000

10000
0.98 0.95
0.80
1.00
Operation at altitudes other than those listed is covered in
AIEE BlB but operation above 10 000 feet should be given

special consideration because of possible influence of altitude on interrupting capacity. Rated Impulse Withstand Voltage-Impulse ratings of standard power circuit breakers are listed in ASA C 37.6. A correction (the same as that given above for rated voltage) should be made for the effect of altitude above 3300 feet on impulse strength. No such ratings are given for low-voltage air breakers because such breakers are seldom exposed to impulse voltages. When necessary, impulse strength could be taken as the crest of the 60-cycle test voltage.l1 The surge protection of the system should be coordinated with the impulse strength of the breaker, both across the open contacts and to ground. Attention should also be given to increase in surge voltage because of reflections which occur at breakers when their contacts are open, especially where cables are involved.lo4 Frequency-Standard power circuit breakers are rated at 60 cycles. Service at other frequencies must be Although standard 60-cycle given special consideration. power circuit breakers are given corresponding continuous-current ratings for 25-cycle service,lo5 other ratings (e.g. interrupting capacity) must be checked and accessories must be made suitable. Low-voltage breakers are listed for 60-cycle service and for direct current. Service at other frequencies requires special consideration. Rated Continuous CurrentThis rating is based on operation of the circuit breaker or switchgear assembly where the ambient temperature (measured outside the enclosure where such is supplied) does not exceed 40C and the altitude does not exceed 3300 feet. Operation in higher ambient temperature must be given special consideration. Molded-case, thermal-trip, low-voltage circuit breakers are calibrated on the basis of an ambient temperature of 25C. Operation in ambient temperatures other than 25C will affect the tripping characteristic and must be Standard equipment may be taken into consideration. operated at higher altitudes by reducing the continuous current rating in accordance with the following table. Altitude in Feet
Fig. 46230-kv, 3500-mva, floor-mounted,
oil circuit breaker.
breakers of the grounded metal-tank type. Pneumatic operating mechanisms predominate and these can be arranged to give 20-cycle reclosing. A 230-kv installation is shown in Fig. 46. Some experimental installations have been made of airblast and oil-poor breakers at high voltage and a table of ratings for such breakers has been included in the standards as a guide for development. At the present time it is difficult to produce and install an oil-less outdoor breaker having current transformers and potential devices to compete economically with conventional oil breakers. There is also the hazard of fragile porcelain structures. So far there is no definite indication of the place such breakers will fill in normal practice.
29. Standard
Ratings of Circuit Breakers
The standard ratings of the several classes of circuit breakers are defined in ASA, AIEE, NEMA, and Underwriters Laboratories standards. 106-111It is not the intention here to review these ratings in detail, only to discuss the principal factors involved in the selection of a circuit breaker for a particular application. Rated VoltageIn general a circuit breaker is given a rated voltage which designates the maximum nominal system voltage for which the breaker is intended and also a maximum design voltage which designates the maximum operating voltage for which it is intended. For certain low-voltage breakers this distinction is not made and rated voltage should be taken as maximum. Standard voltage ratings of power circuit breakers are in terms of threephase line-to-line voltage.
3300 4000 5000

10000
Current Correction Factor 1.00
0.996 0.99
0.96
Rated Interrupting Current-Rated ing Mva-Operating Duty-Interrupting

rated interrupting
Interrupt-
TimeThe
current of a power circuit breaker is
382
Application
Chapter 11
based on the rms total current in any pole of the breaker at the time the breaker contacts part. (Note that this time may be considerably shorter than the interrupting time.) The correct value of rated interrupting current for an operating voltage other than rated value can be calculated by the following formula : Amperes at operating voltage= amperes at rated voltrated voltage Operating voltage should, of age X operating voltage. course, not exceed the maximum design voltage. Also, no matter how low the voltage, the rated interrup ting current is not increased above the rated maximum interrupting current. Standard rating tables P give the value of rated interrupting current at rated voltage as well as the rated maximum interrupting current and the corresponding operating voltage. Over this range of voltages the product of operating voltage and current is constant and this product times a phase-factor is called rated interrupting mva. For 3-phase circuits the factor is 1.73, for 2-phase circuits 2.0, and for l-phase circuits 1.0. However, standard breakers are rated only on a three-phase basis and rules are provided (given later in this chapter) for determining the equivalent three-phase interrupting ratings. The above values of rated interrupting current are based on specified conditions of circuit recovery voltage, breaker performance, and also on standard operating duty. For power circuit breakers rated 50 mva and higher (oil-tight or oil-less) this consists of two unit operations (CO) oeparated by a 15-second interval. Each unit operation consists of breaker closing followed by its opening without inten-
Fig. 47Breaker
rating factors for the usual reclosing duty cycles.
C. The Breaker Rating Factors for the Usual Reclosing Duty Cycles are obtained by reference to Figure 47. This gives the percentage rating factor for the duty cycle in question at the rated interrupting current at the circuit voltage of the system to which the breaker will be applied. D. The breaker interrupting rating at the specified reclosing duty cycle is obtained by multiplying the rated interrupting amperes by the rating factor. Example: Determine the interrupting rating of a 34.5 kv, lOOO-mva breaker used at 23-kv on Reclosing Duty Cycle I. 1. Breaker Interrupting Rating on Standard Duty Cycle: 17 000 Amperes at 34.5-kv 25 000 Amperes at 23 kv 2. Figure 1, Curve A, gives a rating factor of 91.5 percent at 25 000 amperes for reclosing duty cycle I. 3. Breaker Interrupting Rating on Duty Cycle I at 23 kv =Rating Factor, 91.5 percent X25 000 Amperes=22 900 Amperes. E. Breaker Rating Factors for Other Than the Usual Reclosing Duty Cycles 1. The standard for the number of operations is two (2). Additional operations increase the duty on the contacts and therefore a rating factor is applied to enable circuit breakers to meet standards of interrupting performance. (See SG6-40, Par. C) (a) For Three Operations-Use mean factor between 2 and 4 operations, (Standard and Duty Cycle I). Example: 0+15 Sec.+CO+lfi Sec.+CO y0 Factor = 97 for 20 000 Amperes
tional time delay. This standard operating duty is designated by the expression, CO+ 15 sec.+CO. For power circuit breakers rated 25 mva and lower (non-oil-tight) the standard operating duty consists of two unit oierai tions separated by a two-minute interval (CO+2 min. + CO). For any other operating duty the standard interrupting ratings should be reduced in accordance with rules given in NEMA standards *log The following revision of the current rules is now being recommended by AIEE to ASA.
NEMA STANDARDS-RECLOSING DUTY CYCLE FACTORS FOR OIL-TIGHT AND OIL-LESS POWER CIRCUIT BREAKERS-REVISION OF 11/17/49
SG-6-90 BREAKER RATING SERVICE (Rev.)
FACTORS
FOR
RECLOSING
A. The interrupting ratings of power circuit breakers may be reduced for operating duty cycles other than the standard, CO +15 Sec.+CO (See SG6-40, Par. B) to enable them to meet the standard of Interrupting Performance. (See SG6-40, Par. C) Note: Such factors do not apply to highly repetitive duty at or near the continuous rating of the breaker. B. For purposes of this section, the following duty cycles shall be considered as representing the Usual Duty Cycles for Reclosing Reclosing Duty Cycle IO+15 Sec. **+CO+l5 Sec.** +CO+15 Sec.**+CO Reclosing Duty Cycle II 0+0 Sec.*+CO Reclosing Duty Cycle III 0+0 Sec.*+C0+15 Sec.** +C@+ 15 Sec.**+CO *Zero seconds shall be interpreted delay. **15 seconds or longer. to mean no intentional time
(b) For Five Operations-Use Factor obtained by reducing that for 4 operations (Duty Cycle I) by difference between 2 and 4 operations. Example: 0+15 Sec.+C0+15 Sec.+C0+15 Sec. +CO+15 Sec.+CO y0 Factor = 94 - 6 = 88 for 20 000 Amperes
2. The standard for the interval between operations is 15 seconds. Reducing this interval increases the interrupting duty and therefore a rating factor is applied to enable breaker to meet standards of interrupting performance. (See SG6-40 Par. C)
Chapter
11
Application
383
(a) For Two Operations-Reduce 100 percent factor for 15 seconds by an amount equal to proportionate part of the reduction for zero interval as determined by the ratio of the times. Example: 0+5 Sec.+CO a/,Factor=lOO-#(100-94)=96for20 Amperes 990
(b) For Four Operations-Multiply the factor for Duty Cycle I by the appropriate factor determined under paragraph 2 (a) for each interval less than standard. 0+5 Sec.+CO+Fi Sec.+C0+5 Sec.+CO y0 Factor=94X96X96X96=83 for 20 000 Amperes (c) For Three or Five Operations-Use combination of Rule 1 and Rule 2 (a) applied for each interval. Example: 0+5 Sec.+C0+5 Sec.+C y. Factor = 97 X96X 96 = 89 for 20 900 Amperes Example:
the interruption of currents from 25 percent to 100 percent of the rated value. The interrupting rating of low-voltageair circuit breakers is based on the rms total short-circuit current which would occur at the end of s cycle at the breaker location if the line terminals of the breaker were short-circuited. The impedance of the breaker which interrupts the circuit should not be included in calculating the interrupting duty. Also, for three-phase a-c circuits the breaker rated interrupting current should be chosen on the basis of the average of the currents in the three phases. For single-phase. circuits the average current which would occur for three
successive short-circuits should be used. For average tems the average 3-phase or l-phase rms total current syswill
3. The usual instantaneous reclosing cycles are Duty Cycles II and III. Variations are in the number of operations. (a) For Three Operations-Use mean between factors for Duty Cycles II and III. Example: Sec.+C0+15 Sec.+CO 94+t35 y0 Factor = =89.5 for
2
0+0
20 960
Am-
peres (b) For Five Operations-Use factor by reducing that for 4 operations (Duty Cycle III) by difference between 2 and 4 operations. Example: 0+0 Sec.+C0+15 Sec.+C0+15 Sec.+CO +15 Sec.+CO % Factor = 85 - (94 -85) = 76 for 20 000 amperes duty cycle factors for breakers rated 25
be equal to 1.25 times the initial subtransient symmetrical current. Low-voltage air breakers for d-c service are also applied on the basis of the short-circuit current without the breaker in place; however, the maximum current is measured. The standard rated interrupting current of low-voltage air circuit-breakers is based on a standard operating duty designated 0 +2 min+CO. The breaker opens the circuit and, after a 2-minute interval, is reclosed on the fault, which it opens without purposely delayed action. For other interrupting duty the siandard interrupting rating should be multiplied by factors given in Table 6.
TABLE 6**-OPERATING DUTY FOR REGLOSING SERVICE FOR LARGE AIR CIRCUIT BREAKERS
The
reclosing
mva and below (non-oil-tight Table 5.
circuit
breakers)
are given in
TABLE 5* -RRECLOSING DUTY CYCLE FACTORS FOR NON OILTIGHT OIL POWER CIRCUIT BREAKERS
This table does not apply to molded-case breakers cause they are not used on reclosing service. Rated Momentary CurrentThe maximum
total current (including breaker, measured during the d-c component) the maximum cycle,
berms
through a should not
exceed the rated momentary breakers this rating applies

taken individually and
current. For power circuit to each pole of the breaker

worst condition of asym-
for the
metry.
The interrupting time of maximum interval from the at normal control voltage This time is published for
a power circuit breaker is the time the trip coil is energized until the arc is extinguished. standard power breakers for
Rated Four-Second CurrentA four-second current rating is given for power circuit breakers based on the rms total current measured or calculated at the end of one second. For standard breakers it is numerically equal to the rated maximum interrupting current, and l/1.6 times For normal circuits this means the momentary current. that the permissible duration of the maximum permissible
384
Application
Chapter 11
fault current is four seconds. No current rating is given for times longer than four seconds but less than continuous. No similar short-time rating is given for low-voltage breakers because such breakers are normally equipped with direct acting series overload trips. Rated Making CurrentThis rating is given to power operated power circuit breakers only. No provision is made for manual closing of oil-less breakers or for oil breakers above 250 mva. It is essentially a design requirement to preclude welding of contacts or other undue damage when a breaker closes on a fault. It is required that the breaker be immediately opened without purposely delayed action. n The values of momentary and making-current rating for present standard breakers have been so selected that these ratings will not normally limit application of the breakers when they are applied in accordance with the recommended simplified procedure, which will be described later in this chapter. An exception may occur where motors produce a large portion of the fault current. Breakers may also be applied on the basis of decrement curves or detailed calculations and, under unusual conditions or in existing installations, momentary current may be the limiting rating of the breaker. Rated Latching CurrentThis rating is distinguished from rated making current in that the breaker must latch when it closes on a fault of the specified rms total current magnitude. Thus delayed tripping is permissible within this rating if the magnitude, duration, and operating duty are within the short-time and interrupting ratings. For present standard power circuit breakers the latching current rating is numerically equal to both the four-second rating and the maximum interrupting current rating. However, the latching current is measured during the maximum cycle whereas the interrupting rating is measured at the time the contacts part and the foursecond rating is measured at the end of one second. Reclosing TimeFor outdoor reclosing oil-circuit breakers standard and fast reclosing times are shown in Table 7. These values apply only to breakers which have a continuous current rating of 1200 amperes or less when operated in conjunction with an automatic reclosing device.
30. Selection of Circuit Breakers for Specific System Conditions GeneralThe great majority of circuit breakers are
applied as three-pole, gang-operated breakers in threephase power systems which are ungrounded or grounded at the neutrals of generators or transformers. Consequently the standard ratings of most circuit breakers are given on this basis. For such an application it is sufficient (in so far as rating is concerned) to select a breaker such that none of its standard ratings will be exceeded under any condition of system operation. For example, the voltage rating should include allowances, where applicable, for such factors as line voltage regulation, shunt or series capacitance, overexcited or overspeed operation of synchronous machines, line-dropcompensation of tap changers or feeder voltage regulators and the operation of transformers on tap positions other than the nominal values. Both voltage and continuous current ratings should take into account future load growth and the contingencies associated with circuit or apparatus outages. The various ratings associated with interruption of faults should include allowances for increase in generation, addition of parallel circuits or transformers, and any other system changes which would increase interrupting duty. The calculation of fault currents and their interpretation in terms of interrupting ratings will be considered separately in Sec. IV. System frequency will usually be substantially constant at GO-cycles. Operation at other frequencies or at varying frequency requires special consideration. The interrupting time of the circuit breakers themselves is subject to choice in few cases. Considerations of transient stability may dictate one or another type of relaying system in order to obtain sufficiently fast clearing. However, considerations of system operation and stability may or may not call for fast reclosing, and a choice should be indicated.
Determination of Equivalent Three-Phase Voltage and Interrupting RatingsThe standard ratings of most power circuit breakers are given in terms of three-pole breakers for three-phase systems. These voltage ratings are based on the line-to-line voltage of the circuit, and the interrupting ratings are given in amperes and approximate three-phase kva. In order to select the proper l-, 2-, 3-, or 4-pole circuit breaker for special services on three-phase circuits, and for use on two-phase the equivalent three-phase and single-phase circuits, breaker rating can be determined from Tables 8, 9, and 10. First, the three-phase voltage rating of the breaker type must be equal to or greater than the voltage determined from column 5 of the tables. Second, make a tentative breaker selection on the basis of equivalent three-phase kva in accordance with column 6 of the tables. If the computed equivalent three-phase kva is more than 95 percent of the approximate kva rating of the breaker type a further check must be made. In such cases the product of rated voltage times rated interrupting current times 1.73 (standard three-phase ratings) must equal or exceed the equivalent three-phase kva calculated in column 6.
TABLE
7*RECLOSING TIME FOR OUTDOOR RECLOSING OIL CIRCUIT BREAKERS (60-CYCLE BASIS)
Chapter 11

THREE-PHASE VOLTAGE
Application
AND INTERRUPTING
385
RATINGS FOR THREE-PHASE SYSTEMS
TABLE8DETERMINATION OF EQUIVALENT
386
Application
Chapter 11
TABLE %-DETERMINATION OF EQUIVALENT THREE-PHASE VOLTAGE AND INTERIWP~~NG RATINGS FOR TWO-PHASE SYSTEMS
Chapter
TABLE
11
10 DETERMINATION

OF EQUIVALENT THREE-PHASE VOLTAGE
Application
AND INTERRUPTING RATINGS FOR SINGLE-PHASE SYSTEMS
387
Third, the short-time current rating and the interrupting capacity current limitation must not be exceeded. The fault current may be calculated by one of the methods described in the next section, and should be checked for all types of faults. Switching of Capacitive CurrentWhen circuit breakers are used to switch the charging current of lines or cables or to switch capacitor banks, abnormally high voltages can be produced by restriking in the breaker. Experience 101 has indicated that transient voltages which result from such restriking will seldom exceed 2.5 times normal line-to-neutral crest voltage on circuits having effectively grounded neutrals. There is relatively little hazard to either the breakers or to other apparatus on such circuits. There are insufficient data on ungrounded or impedance-grounded systems to draw conclusions. Lightning arresters may be damaged if the voltages developed are sufficient to cause them to discharge and if, in addition, the line capacitance is large. Because of the random nature of the phenomena involved it is not possible
at this time to give specific limits for capacitive switching. As an approximate guide special consideration should be given when one desires to switch 69-kv cables which exceed 9 miles in length or 115 and 138kv cables longer than 7 miles. Another problem to be considered is that a large momentary current may flow when one capacitor bank is switched in parallel with another capacitor bank. This current is a function of the capacitance involved and the inductance of the leads connecting the two banks. This current may be calcu1ated102 and should not exceed the momentary rating of the circuit breaker.
Conditions
Affecting Construction
or Protective
FeaturesThere
are unusual conditions which, where they exist, should be given special consideration in the selection and design of the apparatus. Among such unusual conditions are:
(1) Exposure
to damaging fumes or vapors, excessive or abrasive dust, explosive mixtures of dust or gases, steam, salt spray, excessive moisture, or dripping water, etc.;
388
(2) (3) (4) (5) (6)
Relay and Circuit-Breaker Exposure to abnormal vibration, shocks or tilting; Exposure to excessively high or low temperatures; Exposure to unusual transportation or storage conditions; Unusual space limitations; Unusual operating duty, frequency of operation, difficulty of maintenance, etc.
Application
Chapter 11
31. Requirements for Low-Voltage Air Breakers in Cascade Arrangement*
Circuit
When a plurality of low voltage air circuit breakers are connected in series in a distribution system, and the breakers beyond those nearest to the source are applied in the following correlated manner, they are said to be in a cascade arrangement. In this cascade arrangement, breakers toward the source are provided with instantaneous tripping for current values which may obtain for faults beyond other breakers nearer the load. Hence, breakers in the series, other than the breaker closest to a fault may trip and interrupt loads on other than the fault circuit. Such arrangements are used only where the consequent possible sacrifice in service conWhere continuity of service is tinuity is acceptable. desired, selective tripping arrangements of fully rated breakers are required. Where continuity of service is not important, properly selected breakers may be applied in cascade. The following requirements shall be observed:
(b). All circuit breakers subjected to fault currents in excess of their interrupting rating shall be electrically operated. (c) . Where cascading is proposed, recommendations shall be obtained from the manufacturer in order to insure proper coordination between circuit breakers. (d). The operation of breakers in excess of their interrupting rating is limited to one operation, after which inspection, replacement, or maintenance may be required. In calculating the short circuit current through each step in (al), (a2), and (a3) above it is permissible to include the impedance of all circuit elements (including breaker trip coils) between the line terminals of the breaker in question and the source, but not the impedance of the breaker for which the interrupting current rating is being determined. For example, the impedance of breakers in the first two steps may be included in the calculation to determine the fault current to which the breakers in the third step will be exposed. However, the impedance of the third-step breakers should not be included.
32. Selective Tripping of Low Voltage Air Circuit Breakers*

Properly selected air circuit breakers may be applied to low voltage circuits to obtain selective tripping. The following requirements shall be observed: (a) Each air circuit breaker must have an interrupting rating equal to or greater than the available short circuit current at the point of application. (b) Each air circuit breaker, except those having instantaneous trips (such as the one farthest removed from the source of power), must have a short-time rating equal to or greater than the available current at the point of application. (c) The time-current characteristics of each air circuit breaker at all values of available overcurrent shall be such as to insure that the circuit breaker nearest the fault shall function to remove the overcurrent conditions, and breakers nearer the source shall remain closed and continue to carry the remaining load current. (d) To insure that each breaker shall function to meet the requirements of paragraph (c) above, the time current characteristics of adjacent breakers must The pickup settings and time delay not overlap. bands of both the long-time and short-time delay elements must be properly selected. (e) Manually operated circuit breakers shall be limited to applications in which delayed tripping requirements do not exceed 15 000 amperes or 15 times the coil rating, whichever is greater. (f) The time-current characteristics of a breaker in a selective system shall be such that up to four breakers may be operated selectively in series, when required. One of these breakers shall be a load breaker equipped with an instantaneous trip element. NOTE: Attention is directed to the fact that operation of selective tripping requires coordination with the rest of the system; as for instance, the low voltage side of a trans-
(a) Cascading shall be limited to either two or three steps of interrupting rating. (1). The interrupting rating of the breaker or breakers nearest the source of power shall be equal to at least 100 percent of the short-circuit current as calculated in accordance with section 29. The breaker or breakers in this step shall be equipped with instantaneous features set to trip at a value of current that will give back-up protection whenever the breaker in the next lower step carries current greater than 80 percent of its interrupting rating. (2) The breaker or breakers in the second step shall be selected so that the calculated short circuit current through the first step plus motor contribution in the second step, will not exceed 200 percent The breaker or of their interrupting rating. breakers shall be equipped with instantaneous trip set at a value of current that will give back-up protection whenever the breaker in the next lower step carries current greater than 80 percent of its interrupting rating. For the second step of a two step cascade the breaker or breakers shall have an instantaneous trip setting above the starting inrush current of the load. (3) The breaker or breakers in the third step shall be selected so that the calculated short circuit current through the first step, plus motor contribution of the second and third steps, will not exceed 300 percent of their interrupting rating. The breaker or breakers shall have instantaneous trips set above the starting inrush current of the load. *Taken in pa rt from the current proposed revision of NEMA
Standards 46-109not applicable to molded-case breakers
Chapter 11
Application
389
Fig. 48Typical
fault on a three-phase
a-c system.
former bank requires that in the application of relays or fuses on the high side, proper coordinating steps should be taken.
V. FAULT CALCULATIONS
In order to determine the momentary and interrupting duty on circuit breakers and to make preliminary relay settings it is necessary to predict the fault currents that may occur at each circuit breaker location. This information is sometimes available from tests or from previous calculations on adjacent circuits, but must frequently be calculated for a new system or extension. The rigorous determination of short-circuit currents as a function of time involves too laborious a calculation to be practical. Thus, some approximation is required, and a degree of judgment must be used in the application of any method proposed. In the following paragraphs several such methods will be discussed, including the simplified procedure suggested by the AIEE Protective Devices Committee. In using any of these calculating procedures it is necessary to determine the system impedance as viewed from the point of fault, and the current distribution for different kinds of faults. Such calculations for relatively simple systems or parts of systems can be made directly. The network solutions described in Chap. 10 and the method of symmetrical components given in Chap. 2 are helpful in such calculations. The calculation of faults by these methods on many modern interconnected systems may become entirely too involved. Such systems can be represented in miniature on an a-c or d-c network calculator. Fault currents can be determined from calculator readings in a relatively short time. A description of an a-c network calculator is given in Sec. 35 of Chap. 13. A d-c network calculator can be used for studies where either resistance or reactance alone is sufficient to represent the system. Network calculators are also used in studies of load-current distribution, voltage regulation, transient overvoltage, and transient and steady-state stability.
33. Components
of Fault Current
EX of Fig. 48. In this diagram OX is the line of zero current and 0 represents the time at which the fault has occurred. The current to the left of OY is the load current prior to the fault. The short-circuit current wave is unsymmetrical with respect to the OX axis immediately after the short circuit, but during increasing increments of time it approaches a position of symmetry. This asymmetry is dependent upon the point of the voltage wave at which the short circuit occurs. It is possible, by short circuiting at different points on the normal voltage wave, to secure short-circuit current waves ranging anywhere from those symmetrical about the OX axis to those totally asymmetrical. CD is a curve passing through the maxima of the wave of the total current, and EF is a curve passing AB is a curve cutting the vertical through the minima. everywhere midway between CD and EF. The wave of fotaZ current with crests along curves CD and EF and with ordinates measured from the axis OX can be resolved into two components, namely: 1. A direct-current component. 2. An alternating-current component. The direct-current component is determined at any instant by the ordinate GH of the curve Al?, at the time X. The alternating-current component is a wave with a crest value at any time equal to the difference between the ordinates of the curves CD and AB. This difference at the time X has the value HJ. The rms values of this alternating-current component are shown on curve ST. At any instant, this component is considered to have the same rms value as an alternating wave of constant amplitude with crest value one-half the distance between curves CD and EF at that instant. The rms value of the total current wave under short circuit at any instant is the square root of the sum of the squares of the direct-current component and the rms value of the alternating-current component at that instant. The rms values of this total current are shown on the curve RT. The rms value of the total current at the time of parting of the circuit-breaker contacts determines the interrupting rating of a power circuit breaker.
Before discussing specific methods of fault calculation for circuit-breaker and relay application the current components of a typical fault on an a-c system will be reviewed briefly. A more complete analysis is given in Chap. 6. The current in one phase for a three-phase fault on an a-c system is shown as a function of time by the curve
34. Simplified Procedure for Calculating ShortCircuit Currents for the Application of Circuit Breakers and Relays
A simplified procedure for the calculation of short-circuit currents has been presented in reports 11~116 sponsored
390
Application
Chapter
11
TABLE llREACTANCE QUANTITIES AND MULTIPLYING FACTORSFOR APPLICATIONOF CIRCUIT BREAKERS
by the Protective Devices Committee of the AIEE. This method has been found satisfactory and is intended for general use by the industry as a simplified method of approximating the magnitude of fault currents. However, other more rigorous methods should be used when required. The new method is based upon the determination of an initial value of rms symmetrical current (a-c component) to which multiplying factors are applied for application purposes. In the determination of this current, the following symbols are used : E = line-to-neutral voltage. X1=positive sequence reactance viewed from the point of fault, including transient or subtransient directaxis rated voltage reactance of machines as specified in Tables 11 and 12 in ohms per phase. X0 = zero-sequence reactance. Z&I zero-sequence = resistance.
(a) Circuit Breaker Application(1) Determine the highest value of rms symmetrical current for any type of fault equal to E/Xl or 3E/(2X1+X,J, whichever is greater, except that when Ro is greater than 2.23X1 no consideration need be given to the latter expression. This value should be taken for the maximum connected synchronous capacity. (2) Multiply this current by the proper factors from Table 11. (3) The resulting interrupt-
ing and momentary currents should be used to select the circuit breaker. The factors given in Table 11 represent the ratio between the rms total current at the instant of contact parting and the initial value of rms symmetrical current. In determining these factors it was assumed that circuit breakers should be installed which would permit the use of high-speed relays at some later date, and the time of contact parting was selected on this basis. Contact parting times of 4,3,2, and 1 cycles were assumed for 8-, 5-, 3-, and 2-cycle breakers. Note that the total fault current calculated above may in some cases divide between two or more circuits. It is necessary to determine the maximum fault current that must be interrupted by each breaker under any circuit condition (see example). For most apparatus and circuits the resistance may be neglected as a justifiable approximation. For underground cables and very light aerial lines the resistance may be as great as the reactance. For these elements the impedance should be used instead of the reactance. Unless it constitutes a major part of the total circuit impedance this impedance may be added arithmetically to the reactance of the rest of the circuit without appreciable error. (b) Overcurrent Protective RelaysIn approximating the settings of overcurrent relays, the fault currents for two conditions should be determined:
Chapter 11
Application
TABLE 12
391
1. The maximum initial symmetrical current for maximum connected synchronous capacity as determined by E/X1 or 3E/(2X1+Xo), whichever is greater, except that, when & is greater than 2.23X1, no consideration need be given to the expression 3E/(2X1 +X0). 2. The minimum symmetrical current for minimum connected synchronous capacity as determined by 0.866E/X1, or 3E/(2X1+X0) for reactance grounded systems. In particular situations, allowance should be made for remote fault locations and fault resistance. Ground, distance, balanced, and other types of relays require special consideration. For each of these conditions use machine impedances and multiplying factors in accordance with Table 12. (c) ExampleIn order to illustrate the use of the above method of calculation, circuit breaker ratings for several locations in the system shown in Fig. 49 will be determined. The approximate impedance data references in Sec. 37 will be used. From Table 4 of Chap. 6, Part XIII, the waterwheel gen-
erators would have a subtransient reactance of 24 percent on their own base. The 138-kv transformers would have an impedance of about 11 percent according to Table 1 of Chap. 5. The combined positive-sequence impedance of all four generators and transformers, viewed from the 138kv bus is thus 35 percent on 200 mva or 17.5 percent on 100 mva. The zero-sequence impedance would be that of the transformers alone or 5.5 percent on 100 mva. From Tables 2 and 6 of Chap. 3 each transmission circuit has a positive-sequence reactance of 0.77 ohms per mile. For the two 40 mile circuits in parallel the reactance is 8.1 percent on 100 mva. From Table 14 of this chapter the zero-sequence reactance may be estimated at 24.3 percent on 100 mva. The 100-mva step-down transformers will also have an impedance of 11 percent on their kva rating or a net for the two of 5.5 percent on 100 mva. The turbine-generators (see Table 4 of Chap. 6) will be taken as 9 percent each on 50 mva or a total of 4.5 percent on 100 mva for the four units. The above impedances may be combined into the equivalent circuit shown in Fig. 50. For a fault at A in Fig. 49 the 3-phase fault will govern breaker interrupting duty because of the limiting effect of
Fig. 49Hypothetical
system for example.
Fig.50Equivalent
circuit for system of Fig. 49. Impedances in percent on 100 mva.
392 the generator grounding
Application
chapter
11
resistor and the delta-connected 1.0 per unit =420 mva. transformers. This fault is -=4.2 0.239 Since this is iess than 500 mva the general case of Table 11 applies and an &cycle breaker of 420 mva interrupting ratFor normal interrupting duty ing would be adequate. (CO + 15 seconds + CO) a standard 500-mva breaker could be chosen without further analysis since the calculated interrupting mva does not exceed 95 percent of the breaker rating. If we desire to provide instantaneous single-shot reclosing on this feeder-breaker the interrupting current rating must be calculated. The interrupting rating of a 500-mva breaker at 13.8 kv is 21 000 amperes. According to Fig. 47 reclosing duty cycle II requires reduction of this rating to 21 000X0.94 = 19 700 amperes. The calculated fault level of 420 mva is equal to 17 600 amperes, so a standard 500mva breaker would still be adequate. Regardless of the interrupting duty the momentary rating required would be 1.6 X 17 600 = 28 200 amperes. For a fault at B, E/X1 =&$ = 25.5 per unit or 2550
for a fault on 138-kv bus at Fig. 51Current distribution steam stationcurrents in per unit on 100-mva base.
mva. However, the portion of the fault contributed by generator B does not go through breaker B. This is & = 5.5 per unit or 550 mva. Thus E/Xl for breaker B is . 2000 mva. Since this value is greater than 500 mva and all standard 13.8-kv breakers have 8 cycle interrupting time, breaker B should have an interrupting rating of 1.1 X 2000 = 2200 mva. This is less than 95 percent of 2500 mva and a standard 2500-mva breaker may be chosen without further study. A fault at C will give the highest fault current on any of the main 13.8-kv breakers. The three-phase fault will govern as before. Although it is an abnormal condition, the greatest fault current will flow when breaker D is open. For this condition X1 is 4.0 percent and E/X1=25.0 per unit or 2500 mva. Since this fault is produced predominantly by the 13.8-kv turbine-generators the 1.1 multiplier is required for 8-cycle breakers and the duty exceeds that of the largest standard 13.8-kv breaker. In view of the close margin between the breaker rating and the calculated duty a more accurate check would be in order as suggested in Sec. 30. If such a check still indicated duty in excess of 2500 mva it would be necessary to increase the X1 by modification of generator design or the addition of current-limiting reactors. In order to determine the interrupting duty on the 138 kv breakers at the steam station it is necessary to consider both three-phase and single-line-to ground faults as well as several fault locations and switching conditions. With all breakers closed E/X1 = && . fault and = 13.9 per unit or 1390 mva
It is apparent that ground faults will govern interrupting duty. A fault at D will produce the same fault currents but breaker D will carry I1 = I2 = 3.36 and IO = 3.04. IA = 9.76 per unit. For such a fault the required interrupting duty for a 5 cycle breaker is 9.76X 1.1 = 10.7 per unit or 1070 mva. In order to be safe it is also necessary to consider a fault at D with breaker C open. See Fig. 52. The total fault 3.0 = 12.5 per unit or now becomes 0.0965+0.0965+0.0465 1250 mva but part of lo does not pass through breaker D. = 10.72. The reIn the breaker Ia= 4.16+4.16+2.40
for a three-phase
3E 3.0 2X1+Xo=2(0.0719)+0.0465 = 15.7 per unit or 1570 mva for a line-to-ground fault. The current distribution for such a fault is shown in Fig. 51. The transformer and line circuits have been shown separately in order to study different fault locations. The smallest 138-kv breaker is rated 1500 mva and 5 cycles.
Fig. 52Current distribution for single-line-to-ground fault at 138-kv terminals of transformer with 13.8-kv breaker open.
Chapter 11
Application
393
Fig. 53Short-circuit
decrement
cu rves for similar
parallel
machines.
quire Id interrupting rating for this condition would be 1.1X 10.72 = 11.80 per unit or 1180 mva which is greater than the value calculated in the previous study. Thus a 1500. -mva breaker would be adequate for breaker D on the
basis of the system as shown. If additional generation or lines were contemplated the 1500 mva rating might be exceeded and such changes would have to be considered Reclosing would not normally be used on these breakers
394
Application
Chapter 11
For breaker E a fault at E with all breakers closed required an interrupting rating of 1.1 (4.52+4.52+4.84) If breaker F were open such = 15.2 or 1520 mva. rating of interrupting a fault would require an
0.077+0.077+0.05 Breaker E would probably
1.1
3.0
(k) All generators are assumed to have an open-circuit transient time constant (Td) of 5 seconds and an armature short-circuit time constant of 0.15 second. (l) A subtransient time constant of 0.05 second was used for all curves.
= 16.1
per
unit
or
1610
mva.
have instantaneous single shot reclosing. A standard 1500-mva breaker would be good for only 1425 mva for such reclosing duty (see Fig. 47). Although it might be possible to reduce the interrupting duty of the system to 1425 mva by increasing the transformer impedance, a reasonable amount, some allowance should be made for future addition of generators or transmission lines. Thus a 3500-mva breaker would probably be chosen for the assumed system.
35. Short-Circuit Machines
Calculations
for Similar Parallel
It is intended that the simplified procedure given in Sec. 34 be used for normal circuit-breaker applications and for In some special cases it may be preliminary relay settings. desired to make a more accurate analysis of current decrement such as to take into account abnormal machine time constants or to obtain relay currents a relatively long time after the fault has occurred. When all of the machines which contribute to a fault have similar reactances and time constants, are equally loaded, and are symmetrically located with respect to the fault the group of machines can be represented as a single equivalent generator. The fault current can then be calculated with relative accuracy by the methods described in Parts II, III and VI of Chap. 6 in which the effects of individual machine characteristics, loads, external impedance and change of excitation can be included. A somewhat easier analysis may be made by the use of short-circuit decrement curves which have been published,117*118 and are reproduced in Fig. 53, if the assumptions on which they are based hold for the system under consideration. These are:
The short-circuit current from a synchronous machine consists of an a-c and a d-c component. The a-c component in general can be resolved into a transient component having a relatively large time constant and a subtransient component having a relatively small time constant. The values of these constants are such that during the first one-tenth second the transient component changes very little, but the subtransient component disappears almost entirely. Because of this relation it is possible to plot the two a-c components on one set of curves as shown by the "a" curves of Fig. 53. The numbers of these curves refer to the combined external reactance (exclusive of loads) and machine subtransient reactance. A subtransient time constant of 0.05 and a transient open circuit time constant of
5.0 were used in the preparation of these curves, but the
effect of other time constants can be included by reading vertically from the intersection of the horizontal line corresponding to the particular time constant and the inclined line corresponding to the particular time. The dotted lines show the transient component only of a-c current. These curves are intended primarily for turbine-generator systems as indicated by assumption (g). The assumed relation between transient and subtransient reactance is Xd = (1.4Xd+ .02) per unit. The curves may be used with fair accuracy for salient pole generators with dampers. For salient pole machines without dampers the curves may be used with the following adjustments: (a)
Calculate the total system reactance to the point of fault using the subtransient reactance of the machines, and then subtract 5 percent. (b) Enter the curves with the above modified value of reactance. (For example if the system reactance is 25 percent use the curve marked 20.) (c) The proper a-c component of current will be approximately midway between the dotted and solid portions of the curves of 53(a) in the short time periods where a distinction is made. of current is given by the (d) The proper d-c component dotted curves of 53(b).
(a) Transient (b) (c) (d) (e) (f) (g) (h) (i)
characteristics of alternating-current generators of normal design determined from oscillograph tests. That the effect of capacitance and resistance is neglected, except in so far as decrements are concerned, which effects are included by average decrement factors. That the contact resistance at short circuit is zero. That the alternating-current generators are carrying full load at 80 percent power factor previous to short circuit. That the short circuit is established at the point of voltage wave corresponding to the maximum possible instantaneous current. That the effect of automatic generator voltage regulators is neglected. All reactance up to and including 15 percent is considered within the generator. For values of reactance greater than that the difference is considered external. All machine emfs are assumed to be and remain in phase. The load is assumed to be located at the machine terminals and the fault to occur on an unloaded feeder. The actual system subjected to fault may be represented by a single equivalent generator of the same total rating as the synchronous apparatus of the system and an equivalent external reactance.
With the above general qualifications the curves may be used to calculate three-phase, line-to-line or single-line-toground faults. The following symbols are used:
Xr = percent positive-sequence impedance viewed point of fault, based on the total synchronous X2= negative-sequence impedance viewed from the fault. XO= zero-sequence impedance viewed from the point T, = time constant of direct-current component. Td = short-circuit subtransient time constant. Th, = open-circuit transient time constant. from the kva. point of of fault.
For a three-phase fault use the curves of Fig. 53 (b) for which X=X1 and read the components of on the ordinate scales designated 3-phase. The a-c components may be combined into the rms total for maximum asymmetry by the formula,
ha total =A&qj.
(a) and current and d-c current
Chapter 11
Application of equivalent circuits
395 is given in the
Fig. 53(c) may be used to perform this calculation by laying off the components along the two axes and reading the rms total current on the circular scales. A line-to-line fault is read in a similar manner except the curves are used for which X =X1+X2 and the magnitude of current is read on the ordinate scale headed L-L. For single-line-to-ground faults enter the curves with X=X1+Xz+XO and use the ordinate scale headed L-G. The curves of Fig. 53(b) are plotted against a basic time scale corresponding to T,=0.15 sec. If the d-c time constant is known to be different, read vertically from the intersection of the horizontal line corresponding to the desired time and the inclined line corresponding to the desired time constant.
36. The Internal Voltage Method

For critical relay and circuit-breaker applications where synchronous machines are dissimilar and unsymmetrically located with respect to the fault, a more accurate shortcircuit analysis can be made by means of the Internal Voltage Method. This method lends itself to the use of a network calculator and can be used to include the effect of a change in the excitation of the machines. Because of the limited application of this method of calculation the reader is referred to a series of articles by C. F. Wagner, entitled Decrement of Short-Circuit Currents, which appeared in the March, April, and May 1933, issues of the Electric
Journal.
Feeder Voltage RegulatorsThe impedance of single-phase induction regulators referred to the through kva of the circuit varies with regulator position from approximately 0.7 percent at maximum buck or boost position to approximately 2.5 percent at points midway between the neutral and maximum positions. At the neutral position the impedance is approximately 1.5 percent. The impedance of polyphase induction regulators does not vary greatly with regulator position and lies between 1.0 percent and 1.5 percent on the circuit kva base. For line voltages not exceeding the 15 kv insulation class level, single-core step regulators are used when the line current does not exceed 400 amperes. Two-core step regulators are used for higher current circuits to reduce the current handled by the tap changer to 400 amperes. Two-core four-winding construction is used where the line voltage exceeds the normal 15 kv insulation class level. The impedance of plus or minus 10 percent regulators in single-phase and balanced three-phase circuits is given in Table 13.
TABLE 13 -IMPEDANCE OF feeder REGULATORSCIRCUIT KVA BASE-PLUS OR minus 10 PERCENT PERCENT ON REGULATION
chapter and a table appendix.
37. Approximate lations
Impedance Data for Fault Calcu-
In fault calculations, impedance data applicable specifically to the apparatus and circuits under consideration should be used whenever possible. Such data can usually be obtained from the manufacturers for existing apparatus and can be calculated with the aid of tables referred to below for overhead lines and cables. The necessity for accurate data is particularly important for circuit elements which have a major influence on the fault magnitude. For estimating fault currents on proposed new circuits, and for approximate data on the less important elements of existing circuits, the following references and tables are offered as typical of present-day practice.
densersTable 4 in Chap. 6 lists both average values and
Synchronous
Generators,
Motors,
and
Con-
the probable range of the several impedances and time constants of 60-cycle three-phase synchronous machines. In most simplified fault calculations subtransient reactance is used to represent the positive-sequence impedance of synchronous machines, and its relation to the other impedances is assumed on the basis of typical designs, Exceptions to this assumption are noted in Sec. 34. The effect of external impedance on the time constants is discussed in Sec. 10 of Chap. 6. Induction MotorsThe effect of induction motors on the short-circuit current is discussed in Chap. 6.
Power and Distribution
TransformersTypical
impedance values for distribution and power transformers are given in Table 1 of Chap. 5. The relation between the positive- and zero-sequence impedances for each of the principal types of transformers is also discussed in this
Aerial LinesThe characteristics of aerial lines are given in Chap. 3. When the conductor size and spacing of an aerial line cannot be determined and a rough value of impedance is known to be satisfactory, the reactance of lines above 15 kv class can be taken as 0.8 ohms per mile without serious error. The resistance of such lines will usually be negligible from the standpoint of circuit breaker and relay application. For lines rated 15 kv and below conductor size and spacing vary greatly and typical figures should not be used. If the actual line data cannot be obtained (and an approximate figure is known to be satisfactory) the conductor size and spacing may sometimes be estimated on the basis of thermal and regulation limits of the circuit. The zero-sequence reactance of aerial lines can be estimated from the positive-sequence reactance by the use of Table 14. This approximation is sufficiently accurate for most circuit breaker applications, but when greater accuracy is required refer to Chap. 3 and other references given in that chapter. CablesThe impedance of single- and three-conductor cables is given in Chap. 4. The effect of iron conduit in increasing the reactance and resistance of cables has been investigated by L. Breiger of the Consolidated Edison Co. with both laboratory and field tests.124 These tests show that if the cables are held
396
Application
Chapter
11
TABLE 14APPROXIMATE RATIO OF X0 TO X1 FOR Transmission LINES ANDCABLES
in close triangular arrangement, the reactance is increased by only about 10 percent because of the iron conduit. In many cases, however, cables lie at random in the conduit and the reactance may be increased by as much as 50 percent. For circuit breaker applications, in the absence of specific information, the reactance of iron conduit circuits may be taken at from 40 to 45 microhms per foot for one conductor per phase and 20 to 25 microhms per foot for two conductors per phase. For non-magnetic duct corresponding figures are 35 to 40 microhms per foot for one conductor per phase and 18 to 22 microhms per foot for two conductors per phase. The increase in resistance caused by the iron conduit is not sufficient to justify consideration in circuit breaker applications.
Bus ConductorsThe reactance of most busbar arrangements for low or medium voltage circuits is of the order of 50 microhms per foot. Values for practical circuits range from 30 to about 70 or 80 microhms. In lowvoltage circuits bus reactance may be an appreciable part For example 50 feet of bus at 50 of the circuit impedance. microhms per foot will cause a drop of 50 volts at 20 000 amperes. Fig. 54126gives the 60-cycle reactance per conductor per foot of two rectangular bars in a single phase circuit. The reactance per phase of a transposed three-phase bus may also be obtained from Fig. 54 by replacing s by an equivalent spacing equal to the cube root of the product of the three distances between phase conductors. If the bus is not transposed, the reactance Sequlv. = iYSlS2S3. corresponding to the minimum spacing should be used for circuit breaker applications in order to obtain the maximum current in any pole. For other applications it may be desirable to use the equivalent spacing in order to determine the average reactance per phase. The reactance of bus runs composed of several closely spaced bars per phase may be determined approximately by considering each phase group as a solid conductor having the same overall dimensions. This approximation will give values of reactance accurate within about 5 percent if the distance s is more than twice the equivalent a. For the arrangement in Fig. 55 the error is 15 percent for s = 8 inch-
Fig. 55Reactance voltage drop in each three-phase bus caused by one ampere of balanced three-phase current.
es and 6 percent for 30 inches. A more accurate method of calculation is given in Reference 126. The reactance of irregularly shaped conductors can be determined from Figs. 56 and 57 and similar data published by bus bar manufacturers, such as References 131 to 133. A rough approximation may be obtained by the method described in the preceding paragraph.
Low-Voltage Air Circuit Breakers (600 Volts and

Below)The reactance of low-voltage air circuit breakers with series trip coils may be an appreciable part of the total circuit impedance when the full-load rating of the breaker is small compared with the remainder of the system. Care should be taken to make sure that trip coils are included on all three poles of a breaker before using the values in the accompanying table. If only two coils are used, one-third
Fig. 54Reactance
of rectangular
bar conductors.
Chapter 11
Application
397
Fig. 56Reactance
of square tubular
hue
bars.
(a) Reactance of thin square tubes. (b) Increase caused by round corners. (c) Increase caused by thickness of tubes.
of the single-coil impedance should be used in calculating the maximum pole current in a three-phase fault, and the impedance should be omitted entirely in calculating a single line-to-ground fault. Values of series trip coil impedances are given in Table 15 for Westinghouse DA 50, DB 25, and DB 15 air circuit
breakers, which have interrupting ratings of 50 000, 25 000, and 15 000 amperes respectively at 600 volts or below. The resistance per pole of AB-10 thermal breakers is given in Table 16. The reactance of the main current-carrying loop of an air circuit breaker can not readily be separated from the influence of the bus or cable to which it is connected. It may be calculated along with the bus or neglected. Current Transformers-The reactance of the smaller wound current transformers in a low-voltage circuit may be appreciable when fed from a relatively heavy supply
TABLE 15D-C RESISTANCE AND 60-CYCLE REACTANCE OF 600 VOLT AIR CIRCUIT BREAKER SERIES TRIP COILS*
Fig. 57Reactance
of channel
bus.
398
TABLE 16RESISTANCE PER POLE OF AB-10 CIRCUIT BREAKERS

600-VOLTTHERMAL
Application
Chapter 11
Bibliography of Relay Literature, 1940-1943, AIEE Relay Committee. A.I.E.E. Transactions, Vol. 63, 1944, October section, pages 705-09. Bibliography of Relay Literature, 1927-1939, AIEE Relay Committee. A.I.E.E. Transactions, Vol. 60, 1941, pages 1435-47. Books on Relaying 1. Silent Sentinels (a book), Westinghouse Electric Corporation 1949. 2. Relay Systems, Theory and Application (a book), by I. T. Monseth and P. H. Robinson, McGraw-Hill Book Co., Inc., 1935. Bus and Apparatus Protection
TABLE 17Impedance
of CURRENTTRANSFORMERS
circuit. Approximate values for specific Westinghouse current transformers are given in Table 17. The limiting effect of secondary burden has been neglected for the sake of simplicity.
REFERENCES
RELAYING For a more complete Bibliography refer to Silent Sentinels or to the Bibliographies of the Relay Literature as follows: AIEE Relay Bibliographies. Bibliography of Relay Literature, 1944-1946, AIEE Relay Committee. A.I.E.E. Transactions, Vol. 67, 1948, pages 24-7.
11. The Fault Ground Bus, Its Use and Design in Brunot Island Switch House of the Duquesne Light Co., by R. M. Stanley and F. C. Hornibrook, A.I.E.E. Transactions, Vol. 49, Jan. 1930, pages 201-212. 12. Protecting Frequency Changers from Out-of-Step Conditions, by L. N. Crichton, The Electric Journal, Vol. 34, Feb. 1937, pages 71-76. 13. Relay Protection for Station Buses, by W. A. Lewis and R. M. Smith, The Electric Journal, Vol. 34, Nov. 1937, pages 457-458. 14. Harmonic Current-Restrained Relays for Differential Protection, by L. F. Kennedy and C. D. Hayward, A.I.E.E. Transactions, Vol. 57, May 1938, pages 262-271. 15. Relay Protection for a Large Regulating Transformer, by W. C. Marter, The Electric Journal, Vol. 36, March 1939, pages 87-88. 16. Bus Protection, A.I.E.E.-E.E.I. Committee Report, A.I.E.E. Transactions, Vol. 58, May 1939, pages 206-211. 17. Considerations in Applying Ratio Differential Relays for Bus Protection, by R. M. Smith, W. K. Sonnemann and G. B. Dodds, A.I.E.E. Transactions, Vol. 58, June 1939, pages 243252. 18. Current Transformers and Relays for High-Speed Differential Protection, with Particular Reference to Off-Set Transient Currents, by E. C. Wentz and W. K. Sonnemann, A.I.E.E. Transactions, Vol. 59, Aug. 1940, pages 481-488. 19. A High Speed Differential Relay for Generator Protection, by W. K. Sonnemann, A.I.E.E. Transactions, Vol. 59, Nov. 1940, pages 608-612. 20. Prolonged Inrush Currents with Parallel Transformers Affect Differential Relaying, by C. D. Hayward. A.I.E.E. Technical Paper 41-65. 21. Linear Couplers for Bus Protection, by E. L. Harder, E. C. Wentz, W. K. Sonnemann, E. H. Klemmer. A paper presented at the A.I.E.E. Winter Convention, Jan. 1942. 22. Predicting Performance of Bus Differential Systems, by E. C. Wentz and W. K. Sonnemann, Westinghouse Engineer, Feb. 1942, page 28. 23. The Effect of Direct Current in Transformer Windings, by E. L. Harder, Electric Journal, Vol. 27, Oct. 1930, page 601. 24. Bus Protection Independent of Current Transformer Characteristics, by G. Steeb, A.I.E.E. Technical Paper No. 41-99. 25. System Protection Analysis Precedes Oswego Design, by G, Steeb, Electrical World, Vol. 116, Nov. 1, 1941, page 57. 26. Vibration Protection for Rotating Machinery, R. L. Webb, C. S. Murray. A.I.E.E. Transactions, Vol. 63, July 1944, pages 534-7. 27. Linear Couplers. Field Test and Experience at York and Middletown, Pa., E. L. Harder, E. H. Klemmer and R. E. Neidig. A.I.E.E. Transactions, Vol. 65, Mar. 1946, pages 10713. 28. Motoring Protection for A-C Generators, L. L. Fountain. Westinghouse Engineer, Vol. 6, November 1946, pages 190-91. 29. Relay Protection of Power Transformers, A.I.E.E. Relay Committee. A.I.E.E. Transactions, Vol. 66, 1947, pages 911-15.
Chapter 11
30.
Application
399
Recommended Practices for the Protection of Electrical Apparatus, A.I.E.E. Relay Subcommittee. A.Z.E.E. Transactions, Vol. 52, 1933, pages 607-13. 30.1 Relay Protection for Large Regulating Transformers, W. E. Marter. Electric Journal, Vol. 36, March 1939, pages 87-8. 30.2 Current Transformer Excitation under Transient Conditions, D. E. Marshall, P. 0. Langguth A.I.E.E. Transactions, Vol. 48, 1929, pages 1464-74. Transmission 31. Line Protection (except carrier and pilot wire)
63.
64. 65.
32.
33.
34.
35.
36.
37.
38.
39.
40. 41.
High-Speed Relays Increase System Stability, by S. L. Goldsborough, The Electric Journal, Vol. 27, July 1930, pages 400401. Fundamental Basis for Distance Relaying on Three-Phase Systems, by W. A. Lewis and L. S. Tippett. A.I.E.E. Transactions, Vol. 66, 1947, pages 694-709. A New High-Speed Distance Relay, by S. L. Goldsborough and W. A. Lewis, A.I.E.E. Paper 32M1, Presented Winter Convention, Jan. 26, 1932. Abstract, Electrical Engineering, Vol. 51, March 1932, pages 157-160. High Speed Distance Relay (Type HCZ Relay), by L. N. Crichton, The Electric Journal, Vol. 32, Dec. 1935, pages 537542. A New Distance Ground Relay, by S. L. Goldsborough and R. M. Smith, A.I.E.E. Transactions, Vol. 55, June 1936, pages 607-703; Disc. Vol. 55, Nov. 1936, pages 1255-1256. High Speed Balanced-Current Relays for Parallel Lines, by S. C. Leyland, The Electric Journal, Vol. 33, Aug. 1936, pages 347-350. A System Out-of-Step and Its Relay Requirements, L. N. Crichton, A.I.E.E. Transactions, Vol. 56, Oct. 1937, pages 1261-1267; Disc. Vol. 57, May 1938, pages 284-286. High-Speed Relaying Experience and Practice, A.I.E.E. Committee Report, A.I.E.E. Transactions, Vol. 58, Nov. 1939, pages 588-592. Out-of-Step Blocking and Selective Tripping with Impedance Relays, by H. R. Vaughan and E. C. Sawyer, A.I.E.E. Transactions, Vol. 58, Dec. 1939, pages 637-646. An Improved Polyphase Directional Relay, by Bert V. Hoard, A.I.E.E. Paper 41-62. A Distance Relay with Adjustable Phase Angle Discrimination. S. L. Goldsborough. A.I.E.E. Transactions, Vol. 63, 1944, pages 835-838. Relaying
66.
67.
E. L. Harder and M. A. Bostwick, The Electric Journal, Vol. 35, Nov. 1938, pages 443-448. Ratio Differential Protection of Transmission Lines, by R. M. Smith and M. A. Bostwick, A.I.E.E. Technical Paper, No. 39-154, Aug. 1939. An Improved A-C Pilot Wire Relay, by J. H. Neher and A. J. McConnell, A.I.E.E. Technical Paper, No. 40-127, July 1940. Pilot Wire Circuits for Protective Relaying-Experience and Practice, A.I.E.E. Relay Committee. A.I.E.E. Transactions, Vol. 62, May 1943, pages 210-14. Protection of Pilot Wire Circuits, E. L. Harder, M. A. Bostwick. A.I.E.E. Transactions, Vol. 61, September 1942, pages 645-51. Protection of Pilot Wires from Induced Potentials, R. B. Killen, G. G. Law. A.I.E.E. Transactions, Vol. 65, 1946, May section, pages 267-70. Transformers
Instrument 71.
Current Transformer Excitation Under Transient Conditions, by D. E. Marshall and P. 0. Langguth, A.I.E.E. Transactions, Vol. 48, Oct. 1929, pages 1464-1474. 72. What the Tests Show (Refers to Indianapolis Tests on Breakers, Relays and Potential Devices), by P. 0. Langguth and R. M. Smith, The Electric Journal, Vol. 30, March 1933, pages 98-101. 73. Capacitor Potential Devices, by P. 0. Langguth, The Electric Journal, Vol. 31, March 1934, pages 107-109, 112. 74. Overcurrent Performance of Bushing Type Current Transformers, by C. A. Woods, Jr. and S. A. Bottonari, A.I.E.E. Transactions, Vol. 59, Sept. 1940, pages 554-560. 75. Transient and Steady State Performance of Potential Devices, by E. L. Harder, P. 0. Langguth and C. A. Woods, Jr., A.I.E.E. Transactions, Vol. 59, Feb. 1940, Pages 91-102. 76. Rating of Potential Devices and Suggested Material for a Standard, by J. E. Clem and P. 0. Langguth, A.I.E.E. Technical Paper, No. 40-94, June 1940. Relaying
General
Carrier-Current 51.
One Cycle Carrier Relaying Accomplished, by P. Sporn and C. A. Muller, Electrical World, Vol. 105, Oct. 12, 1935, pages 26-28. 52. A Faster Carrier Pilot Relay System, by 0. C. Traver and E. H. Bancker, A.I.E.E. Transactions, Vol. 55, June 1936, pages 689-696; Disc. Vol. 55, Nov. 1936, pages 1252-1254. 53. Carrier Relaying and Rapid Reclosing at 110 Kv, by R. E. Pierce, R. E. Powers, E. C. Stewart and G. E. Heberlein, A.I.E.E. Transactions, Vol. 55, Oct. 1936, pages 1120-1129. 54. A New High-Speed Distance-Type Carrier Pilot Relay System, by E. L. Harder, B. E. Lenehan and S. L. Goldsborough, A.I.E.E. Transactions, Vol. 57, Jan. 1938, pages 5-10; Disc. Vol. 57, May 1938, pages 291-294. 55. A New Carrier Relay System, T. R. Halman, S. L. Goldsborough, H. W. Lensner, A. F. Drompp. A.I.E.E. Transactions, Vol. 63, 1944, August issue, pages 568-72. Pilot-Wire 61. Relaying
D-C Pilot-Wire Loop Protects 66 Kv. Cable Circuits, by J. H. Neher, Electrical World, Vol. 101, Mar. 25, 1933, pages 384387. 62. A Single-Element Differential Pilot Wire Relay System, by
81. Building a New Power System, by F. S. Douglass and A. C. Monteith, The Electric Journal, Vol. 30, Feb. 1933, pages 55-59. 82. Planned Protective System, by A. C. Monteith and W. A. Lewis, Electrical World, Vol. 106, Nov. 21, 1936, pages 40-42. 83. Coordination of Fuse Links, Method of Comparing Fuse Links and Relays for Sectionalization and Protection, by E. M. Adkins, Electrical World, Vol. 107, March 13,1937, pages 877,952. 84. Pennsylvania Railroad New York-Washington-Harrisburg Electrification-Relay Protection of Power Supply System, by E. L. Harder, A.I.E.E. Transactions, Vol. 58, June 1939, pages 266-276. 85. Sensitive Ground Protection for Radial Distribution Feeders, by L. F. Hunt, and J. H. Vivian, A.I.E.E. Transactions, Vol. 59, Feb. 1940, pages 84-90. 86. Factors Contributing to Improving Electric Service by Means of High-Speed Switching and Utilization of Stored Energy, by J. T. Logan and John H. Miles, A.I.E.E. Transactions, 1941, page 1012. 87. Principles and Practices of Relaying in the United States, by E. L. Harder and W. E. Marter. A.I.E.E. Transactions, 1948, pages 1005-22. 88. Protection of Power House Auxiliaries, A.I.E.E. Relay Committee. A.I.E.E. Transactions, Vol. 65, 1946, pages 746-751, Disc. pages 1115-1116. 89. Protection of Three-Terminal Lines. M. A. Bostwick, E. L. Harder. Westinghouse Engineer, Aug. 1943, pages 76-79. 90. Interim Report on Operation and Application of Out-of-Step Protection. A.I.E.E. Relay Committee. A.I.E.E. Transactions, Vol. 62, Sept. 1943, pages 567-73.
400
91.

Nine Years Experience with Ultrahigh-Speed Reclosing of High-Voltage Transmission Lines, Philip Sporn, C. A. Muller. A.I.E.E. Transactions, Vol. 64, 1945, May section, pages 225-8. High-Speed Single-Pole Reclosing, J. J. Trainor, J. E. Hobson, H. N. Muller, Jr. A.I.E.E. Transactions, Vol. 61, February 1942, pages 81-7. Grounding of Instrument-Transformer Secondary Circuits, A.I.E.E. Relay Subcommittee, A.I.E.E. Transactions, Vol. 66, 1947, pages 419-20. High-Speed Relaying Experience and Practice. A.I.E.E. Relay Committee, A.I.E.E. Transactions, Vol. 58, 1939, pages 58891.
Application
Chapter 11
114.
92.
115. 116. 117. 118. 119. 120.
93.
94.
CIRCUIT 100.
BREAKERS 121. 122. 123.
101. 102.
103. 104. 105.
106. 107. 108. 109. 110. 111. 112. 113.
A Brief Review of Switchgear and Circuit Breaker Practice in the United States, by M. H. Hobbs, A.I.E.E. Transactions, Vol. 67, Part II, page 893. The Interruption of Charging Current at High Voltage, by W. M. Leeds and R. C. Van Sickle, A.I.E.E. Transactions, 1947. The 13,500 KVAR Static Capacitor Installation at Newport News, by E. L. Harder and V. R. Parrick, A.I.E.E. Transactions, 1944. Power Circuit Breaker Ratings, by R. C. Van Sickle, A.I.E.E. Transactions, Vol. 60, page 882. Surge Protection of Cable Connected Equipment, by R. L. Witzke and T. J. Bliss, A.I.E.E. Paper 50-83. American Standards for Alternating-Current Power Circuit Breakers, C37.4 and C37.5-1945, C37.6-1949, C37.7 and C37.8 1945, c37.9-1945. Report on Guiding Principles for the Specification of Service Conditions in Electrical Standards, A.I.E.E. No. lB, 1944. Air Circuit Breakers, A.I.E.E. No. 20, 1930. Low-Voltage Air Circuit Breakers, Proposed A.I.E.E. No. 20A, 1946. Power Circuit Breaker Standards, NEMA 46-116. Large Air Circuit Breaker Standards, NEMA 46-109. Standard for Branch-Circuit and Service Circuit Breakers, Underwriters Laboratories No. 489. Standards for Switchgear Assemblies, A.I.E.E. No. 27, 1942. NEMA No. 44-92. Power Switchgear Assemblies Standards.
124. 125.
126. 127. 128.
129. 130. 131. 132. 133.
FAULT CALCULATIONS System Short-Circuit Currents by W. M. Hanna, H. A. Travers, C. F. Wagner, C. A. Woodrow, and W. F. Skeats, A.I.E.E. Transactions, Vol. 60, page 877. New Fault-Current-Calculating Procedure Recommended to Industry, Electrical Engineering, Vol. 60, page 596. Simplified Calculation of Fault Currents, an A.I.E.E. Committee Report, Electrical Engineering, November 1948. Standard Decrement Curves, By W. C. Hahn and C. F. Wagner, A.I.E.E. Transactions, Vol. 51, 1932, page 353. Decrement of Short-Circuit Currents, by C. F. Wagner, Electric Journal, March, April and May 1933. Symmetrical Components, by C. F. Wagner and R. D Evans (a book) McGraw-Hill Book Company, 1933. Sequence Network Connections for Unbalanced Load and Fault Conditions, by E. L. Harder, Electric JournaZ, Vol. 34, Dec. 1937. Thevenins Theorem, by E. L. Harder, Electric Journal, Vol. 35, Oct. 1938. Calculation of Fault Current in Industrial Plants, by Raymond C. R. Schulze, EZectricaZ Engineering, June 1941. Enclosed Bus-Bar Electrical Distribution Systems for Industrial Plants, by E. T. Carlson, A.I.E.E. Transactions, Vol. 60, page 297. Impedance of Three-Phase Secondary Mains in Nonmetallic and Iron Conduits, by L. Brieger, EEI Bulletin, Feb. 1938. Formulas and Tables for the Calculation of Self and Mutual Inductance, by E. B. Rosa and F. W. Grover, Scientific Paper No. 169, Bureau of Standards. Reactance Values for Rectangular Conductors, by H. B. Dwight, Electric JournaZ, June 1919. Inductance and Reactance of Rectangular Bar Conductors, by 0. R. Schurig, General Electric Review, May 1933. Calculations of Inductance and Current Distribution in LowVoltage Connections to Electric Furnaces, by C. C. Levy, A. I.E.E. Transactions, Vol. 51, page 903. Current Carrying Capacity of Bare Conductors, by H. W. Papst, Electrical World, Sept. 21, 1929. Reactance of Square Tubular Busbars, by H. B. Dwight and T. K. Wang, A.I.E.E. Transactions, Vol. 57, page 762. Alcoa Aluminum Bus Conductors, Aluminum Company of America. Chase Electrical Handbook, Chase Brass and Copper Company. Anaconda Copper Bus Conductors, Publication C-25.

Relay Circuit Breaker Application

Uploaded by

Copyright:

Available Formats

Relay Circuit Breaker Application

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Relay Circuit Breaker Application

Uploaded by

Copyright:

Available Formats

CHAPTER

RELAY AND CIRCUIT-BREAKER

system showing protective zones.

Relay and Circuit-Breaker

2. Basic Relay Element87

Relay and Circuit-Breaker

Relay and Circuit-Breaker

Fig. 2Relay elementsContinued

Relay and Circuit-Breaker

Fig. 2Relay elementsContinued

Relay and Circuit-Breaker

Relay and Circuit-Breaker

Relay and Circuit-Breaker

Ungrounded, and Potential-TransformerGrounded Generators-are those grounded only through

Relay and Circuit-Breaker

Relay and Circuit-Breaker

TABLE 2RELAY APPLICATION CHART-Continued

Relay and Circuit-Breaker

Relay and Circuit-Breaker

Fig. 6Typical large regulating

Relay and Circuit-Breaker

Trip for Transformer

Relay and Circuit-Breaker

of a six-circuit bus using two multi-restraint relays per phase.

Relay and Circuit-Breaker

Relay and Circuit-Breaker

power system illustrating

typical protective problems

and their solution.

Relay and Circuit-Breaker

of overcurrent protection power system.

were as follows: Conductors 56 11 11 11 11 percent percent percent percent percent

Lightning Sleet, Wind, Jumping Apparatus Failure Close-in on Fault Miscellaneous

Relay and Circuit-Breaker

Relay and Circuit-Breaker

12. High-Speed Impedance

Relays, HZ and HZM

Relay and Circuit-Breaker

13. Carrier Pilot Relaying

Relay and Circuit-Breaker

faults during such conditions.

by closing of the contact RR, Fig. 17, whenever a fault is

makes possible superiorrelay performanceunder simultaneously occurring faults.

basis for discrimination

quickly all terminals through which power flows into a line

current relay system including

relays, carrier current transmitter-receivers,

coupling capacitors, and

Relay and Circuit-Breaker Application

overall characteristics of the Type HKB carrier relaying system.

1. Can be added for high-speed

Favoring the Directional-Comparison System Using SteppedDistance Relays

Relay and Circuit-Breaker

with the length of line; consequently there is an economic

Relay and Circuit-Breaker

15. Ground Relaying

Relay and Circuit-Breaker

in distance measurement caused by fault resistance is genmuch

distance protection. for fault conditions,

As there is zero-sequence current only