Industrial Power Systems Handbook - Donald Beeman

Industrial Power Systems Handbook D O N A L D BEEMAN, Editor Manager, Industriaf P w e r Engineering Industrial Engineering Seclwn General Electric Company, Schenectady, New Yorlc

FIRST EDITION

McGRAW-HILL BOOK COMPANY, INC.

1955

New York

Toronto

London

Ch.UPh?r 1

by Donald Beeman, Alan Graeme Darling, and

R. H. Kaufmann

Short-circuit-current Calculating Procedures FUNDAMENTALS OF A-C SHORT-CIRCUIT CURRENTS The determination of short-circuit currents in power distribution systems is just as basic and important as the determination of load currents for the purpose of applying circuit breakers, fuses, and motor starters. The magnitude of the shoncircuit current is often easier to determine than the magnitude of the load current. Calculating procedures have been so greatly simplified compared with the very complicated procedures previously used that now only simple arithmetic is required to determine the short-circuit currents in even the most complicated power systems. SHORT-CIRCUIT CURRENTS AND THEIR EFFECTS

If adequate protection is to he provided for a plant electric system, the size of the electric power system must also be considered to determine how much short-circuit current i t will deliver. This is done so that circuit breakers or fuses may he selected with adequate interrupting capacity (IC). This interrupting capacity should be high enough to open safely the maximum short-circuit current which the power system can cause to flow through a circuit breaker if a short circuit occurs in the feeder or equipment which it protects. The magnitude of the load current is determined by the amount Of work that is being done and hears little relation to the size of the system supplying the load. However, the magnitude of the short-circuit current is somewhat independent of the load and is directly related to the size or I

SHORT-CIRCUIT-CURRENT CALCULATING PROCEDURES

2

capacity of t,he power source. The larger the apparatus which supplies electric power t o the system, the greater the short-circuit current will be. Take a simple case: A 440-volt three-phase lo-lip motor draws about 13 amp of current a t full load and will draw only this amount whether supplied by a 25-kva or a 2500-kva transformer bank. So, if only thc load currcnts arc considered when selecting motor branch circuit breakers, a 15- or 20-amp circnit, breaker wnuld he specified. However, the size of t,he power system back of the circuit breaker has a real bearing on the amount of the short,-circuit,current. which can flow as a result of a short circuit on the load side of the circuit breaker. Hence, a much larger circuit breaker would be required to handle the short-circuit current from a 2500-kva bank than from a 25-kva bank of transformers. A simple mathematical example is shown in Fig. 1.1. These numbers MUST BE CAPABLE OF

El

INTERRUPTING

MOTOR IOOV 100 A

APPARENT IMPEDANCE 20 OHMS

CIRCUIT CURRENT =

E ZT

MUST

LOAD

CURRENT 5 AMP

~ ~ 1 0O.HM 1S

SHORT

1000 AMPERES

:

I00 = 1000- AMPERES 0.1

BE CAPABLE OF INTERRUPTING 10,000 AMPERES

w

I000 A 2 1 = 0.01 OHMS

FIG. 1.1

MOTOR LOAD CURRENT 5 AMP

Illustrotion showing that copocity of power source has more effect on rhortcircuit-current magnitude than load.


3

have been chosen for easy calculation rather than a representation of actual system conditions. The impedance, limiting the flow of load current, consists mainly of the 20 ohms apparent impedance of the motor. If a short circuit occurs at F , the only impedance t o limit the flow of short-circuit current is the transformer impedance (0.1 ohm compared with 20 ohms for the motor); therefore, the short-circuit current is 1000 amp, or 200 times as great as the load current. Unless circuit breaker A can open 1000 amp, the short-circuit current will continue to flow, doing great damage. Suppose the plant grows and a larger transformer, one rated a t 1000 amp, is substituted for the 100-amp unit. A short circuit a t F , (bottom in Fig. 1.1) will now be limited by only 0.01 ohm, the impedance of the larger transformer. Although the load current is still 5 amp, the shortcircuit current will now he 10,000 amp, and circuit breaker A must be able t o open that amount. Consequently it is necessary to coiisider the size of the system supplying the plant as well as the load current, to be sure that circuit breakers or fuses are selected which have adequate interrupting rating for stopping the flow of the short-circuit current. Short-circuit and load currents are analogous t o the flow of xvater in a hydroelectric plant, shoivn in Fig. 1.2. The amount of water that flows under normal conditions is determined by the load on the turbines. Within limits, it makes little difference whether the reservoir behiiid the dam is large or small. This flow of water is comparable to the flow of load current in the distribution system in a factory. On the other hand, if the dam breaks, the amount of water that will flow will depend upon the capacity of the reservoir and will bear little relation to the load on the turbines. Whether the reservoir is large or small will make a great difference in this case. This flow of water is comparable t o the flow of current through a short circuit in the distribution system. The load currents do useful work, like the water that flows down the penstock through the turbine water wheel. The short-circuit currents produce unwanted effects, like the torrent that rushes madly downstream when the dam breaks. SOURCES OF SHORT-CIRCUIT CURRENTS

When determining the magnitude of short-circuit currents, it is extremely important that all sources of short-circuit current he considered and that the reactance characteristics of these sources be known. There are three basic sources of short-circuit current: 1. Generators 2. Synchronous motors and synchronous condensers 3. Induction motors

4


All these can feed shorecircuit current into a short circuit (Fig. 1.3). Generators are driven by turbines, diesel engines, water wheels, or other types of prime movers. When a short circuit occurs on the circuit fed by a generatar, the generator continues t o produce voltage because the field excitation is maintained and the prime mover drives the generator at substantially normal speed. The generated voltage produces a shortcircuit current of a large magnitude which flows from the generator (or generators) to the short circuit. This flow of short-circuit current is limited only by the impedance of the generator and of the circuit between the generator and the short circuit. For a short circuit a t the terminals of the generator, the current from the generator is limited only by its own impedance.

FIG. 1.2

Normal load and short-circuit currents are analogous to the conditions shown in

the hydroelectric plant.

SHORT-CIRCUIT-CURRENT ULCULATlNG PROCEDURES

5

METAL CLAD SWITCHGEAR

SHORT CIRCUIT

CURRENT FROM INDUCTION MOTOR

FIG. 1.3

Generators, synchronous motors, and induction motors all produce short-circuit

current.

HOW SYNCHRONOUS MOTORS PRODUCE SHORT-CIRCUIT CURRENT

Synchronous motors are constructed substantially like generators; i.e., they have a field excited by direct current and a stator winding in which alternating current flows. Normally, synchronous motors draw a-c power from the line and convert electric energy to mechanical energy. However, the design of a synchronous motor is so much like that of a generator that electric energy can be produced just as in a generator, by driving the synchronous motor with a prime mover. Actually, during a system short circuit the synchronous motor acts like a generator and delivers shortcircuit current to the system instead of drawing load current from it (Fig. 1.4). As soon as a short circuit is established, the voltage on the system is reduced to a very low value. Consequently, the motor stops delivering energy to the mechanical load and starts slowing down. However, the inertia of the load and motor rotor tends to prevent the motor from slowing down. In other words, the rotating energy of the load and rotor drives the synchronous motor just as the prime mover drives a generator.

6


The synchronous motor then becomes a generator and delivers shortcircuit current for many cycles after the short circuit occurs on the system. Figure 1.5 shows an oscillogram of the current delivered by a synchronous motor during a system short circuit. The amount of current depends upon the horsepower, voltage rating, and reactance of the synchronous motor and the reactance of the system to the point of short circuit.

LOAD CURRENT

FIG. 1.4 UlILITY SYSTEM

SYNCHRONOUS MOTOR

,-

\

Normally motors draw load current from the source or utility system but produce rhortcircuit current when a short cirw i t occurs in the d a d .

-€t SHORT CIRCUIT CURRENT FROM MOTOR

. .-.. . SYSTEM

SYNCMOYOUS

'

Yoroll

'.

SHORT CIRCUIT

-

I

-.

__

FIG 5 IBmlowl c. e f. 0s. . 1.._ ,. ..,. l.r o. .o . . . cillogrclm of short-circuit current produced by a synchronous motor

SHORT CIRCUIT CURRENT DELIVERED BY A SYNCHRONOUS MOTOR.

SHORT.CIRCUIT-CURRENT CALCULATING PROCEDURES

7

HOW INDUCTION MOTORS PRODUCE SHORT-CIRCUIT CURRENT

The inertia of the load and rotor of an induction motor has exactly the same effect on an induction motor as on a synchronous motor; i.e., it drives the motor after the system short circuit occurs. There is one major difference. The induction motor has no d-c field winding, but there is a flux in the induction motor during normal operation. This flux acts like flux produced by the d-c field winding in the synchronous motor. The field of the induction motor is produced by induction from the stator rather than from the d-c winding. The rotor flux remains normal as long as voltage is applied to the stator from an external source. However, if the external source of voltage is removed suddenly, as it is when a short circuit occurs on the system, the flux in the rotor cannot change instantly. Since the rotor flux cannot decay instantly and the inertia drives the induction motor, a voltage is generated in the stator winding causing a short-circuit current to flow to the short circuit until the rotor flux decays to zero. To illustrate the short-circuit current from an induction motor in a practical case, oscillograms were taken on a woundrotor induction motor rated 150 hp, 440 volts, 60 cycles, three phase, ten poles, 720 rpm. The external rotor resistance was short-circuited in each case, in order that the effect might he similar to that which would he obtained with a low-resistance squirrel-cage induction motor. Figure 1.6 shows the primary current when the machine is initially running light and a solid three-phase short circuit is applied a t a point in the circuit close to its input (stator) terminals a t time TI. The current shown is measured on the motor side of the short circuit; so the shortcircuit current contribution from the source of power does not appear, but only that contributed by the motor. Similar tests made with the machine initially running a t full load show that the short-circuit current produced

T.

FIG. 1.6

Tracer of oxillograms of short-circuit currents produced running a t light load. ,

by an induction motor

8


by the motor when short-circuited is substantially the same, regardless of initial loading on the motor. Note that the maximum current occurs in the lowest trace on the oscillogram and is about ten times rated full-load current. The current vanishes almost completely in four cycles, since there is no sustained field current in the rotor to provide flux, as in the case of a synchronous machine. The flux does last long enough to prodnce enough short-circuit current to affect the momentary duty on circuit breakers and the interrupting duty on devices which open within one or two cycles after a short circuit. Hence, the short-circuit current produced by induction motors must he considered in certain calculations. The magnitude of short-circuit current produced by the induction motor depends upon the horsepower, voltage rating, reactance of the motor, and the reactance of the system to the point of short c. "cuit. The machine impedance, effective a t the time of short circuit, cmesponds closely with the impedance a t standstill. Consequently, the i iitial symmetrical value of Short-circuit current is approximately equnl to the full-voltage starting current of the motor. TRANSFORMERS

Transformers are often spoken of as a source of short-circuit current. Strictly speaking, this is not correct, for the transformer merely delivers the short-circuit current generated by generators or motors ahead of the transformer. Transformers merely change the system voltage and mag; nitude of current but generate neither. The short-circuit current delivered by a transformer is determined by its secondary voltage rating and reactance, the reactance of the generators and system to the terminals of the transformer, and the reactance of the circuit from the transformer to the short circuit. ROTATING-MACHINE REACTANCE

The reactance of a rotating machine is not one simple value as it is for a transformer or a piece of cable, but is complex and variable with time. For example, if a short circuit is applied to the terminals of a generator, the short-circuit current behaves as shown i n Fig. 1.7. The current starts out a t a high value and decays to a steady state after some time has elapsed from the inception of the short cirroit. Since the field excitation voltage and speed have remained snbstantially constant within the short interval of time considered, a change of apparent react,ance of the machine may he assumed, to explain the change in the magnitude of short-circuit current with time. The expression of such variable reactance at any instant after the


9

occurrence of any short circuit requires a complicated formula involving time as one of the variables. For the sake of simplification in short-circuit calculating procedures for circuit-breaker and relay applications, three values of reactance are assigned to generators and motors, viz., subtransient reactance, transient reactance, and synrhronous reactance. The three reactances can be briefly described as follows: 1. Subtransient reactance X y is the apparent reactance of the stator winding at the instant short circuit occurs, and it determines the current Row during the first few cycles of a short circuit. 2. Transient reactance X i is the apparent initial reactance of the stator winding, if the effect of all amortisseur windings is ignored and only the field winding considered. This reactance determines the current following the period when subtransient reactance is the controlling value. Transient reactance is effective up to 45 see or longer, depending upon the design of the machine. 3. Synchronous reactance X d is the apparent reactance that determines the current flow when a steady-state condition is reached. It is not effective until several seconds after the short circuit occurs; consequently, it has no value in short-circuit calculations for the application of circuit breakers, fuses, and contactors but is useful for relay-setting studies. Figure 1.8 shows the variation of current with time and associates the various reactances mentioned above with the time and current scale. Previous loading has an effect on the total magnitude of short-circuit

CURRENT DETERMINED BY SYNCHRONOUS

OCCURS AT THIS TIME.

OF TOTAL OSCILLOGRAM

ONLY TWO ENDS SHOWN HERE. THIS REPRESENTS THE BREAK BETWEEN THE TWO PARTS.

FIG. 1.7 Trace of orcillograrn of hart-circuit current produced by a generator.


10

/-

MAX, SUBTRANSIENT

CURRENT- USE SUBTRANSIENT REACTANCE X"d

TIME(8)

FIG 1.8

Variation of generotor short-circuit current wilh time.

current delivered by a generator. The value of X i or X y generally given by the machine designer is the lowest value obtainable. Hence, its use will show maximum short-circuit current. Certain characteristics of short-circuit currents must he understood before a system analysis can he made. SYMMETRICAL AND ASYMMETRICAL SHORT-CIRCUIT CURRENTS

These terms are used to describe the symmetry of the a-c waves about the zero axis. If the envelopes of the peaks of the current waves are symmetrical about the zero axis, the current is called symmetrical current (Figs. 1.9 and 1.10). If the envelopes of the peaks of the current waves are not symmetrical about the zero axis, the current is called asymmetrical ENVEWPES OF PEAKS OF SINE WAVE ARE SYMMETRIGAL ABOUT THE ZERO AXIS. ZERO

AXIS

FIG. 1.9 Symmelrical a-c wove.


11

THE ENVELOPES OF PEAKS ARE SVHHETRICAL ABOUT

ZERO AXIS

FIG, 1.10

Symmetrical

d t e r n a t i n g current f r o m a short-circuited generotor.

ENVELOPES OF PEAKS ARE NOT SYMMETRICAL ABOUT ZERO AXIS

AX1 S TOTALLY 0 F F SET PARTIALLY O F F S E l

FIG. 1.11 Asymmetrical (I-c waver. The conditions shown here ore theoreticol a n d ore for the purpose of illustration only. D-C component will r a p i d l y d e c a y to zero i n a c t u a l circuits.

FIG. 1.12

Trace of o r c i l l o g r a m of a t y p i c a l short-circuit current

12


current (Fig. 1.11). The envelope is a line drawn through the peaks of the waves, as shown in Figs. 1.9 to 1.12. For the sake of explanation, many of the illustrations, such as Figs. 1.11, 1.15 to 1.19, show sine waves of current uniformly offset for several cycles. It should be noted that in practical circuits the amount of asymmetry decreases rapidly after the occurrence of the short circuit in the system. This decrease of asymmetry is shown qualitatively in illustrations such as Figs. 1.12, 1.20, 1.23, and 1.24. Oscillograms show that short-circuit currents are nearly always asymmetrical during the first few cycles after the short circuit occurs. They also show that the asymmetry is maximum at the instant the short circuit occurs and that the current gradually becomes symmetrical a few cycles after the occurrence of the short circuit. The trace of an oscillogram of a typical short-circuit current is shown in Fig. 1.12. WHY SHORT-CIRCUIT CURRENTS ARE ASYMMETRICAL

In the usual industrial power systems the applied or generated voltages are of sine-wave form. When a short circuit occurs, substantially s i n e wave short-circuit currents result. For simplicity, the following discussion assumes sine-wave voltages and currents. In ordinary power circuits the resistance of the circuit is negligible compared with the reactance of the circuit. The short-circuit-current power factor is determined by the ratio of resistance and reactance of the circuit only (not of the load). Therefore the short-circuit current in most power circuits lags the internal generator voltage by approximately 90" (see Fig. 1.13). The internal generator voltage is the voltage generated in the stator coils by the field flux. If in a circuit mainly containing reactance a short circuit occurs at the peak of the voltage wave, the short-circuit current would start at zero and trace a sine wave which would be symmetrical ahout the zero axis (Fig. 1.14). This is known as a symmetrical short-circuit current. If in the same circuit (i.e., one containing a large ratio of reactance to resistance) a short circuit occurs at the zero point of the voltage wave, the current will start a t zero but cannot follow a sine wave symmetrically about the zero axis because such a current would be in phase with the voltage. The wave shape must be the same as that of voltage hut 90' behind. That can occur only if the current is displaced from the zero axis, as shown in Fig. 1.15. In this illustration the current is a sine wave and is displaced 90' from the voltage wave and also is displaced from the zero axis. The two cases shown in Figs. 1.14 and 1.15 are extremes. One shows a symmetrical current and the other a completely asymmetricd current.

WORT-CIRCUIT-CURRENT CALCULATING PROCEDURES

13

GENERATOR TRANSFORMER INTERNAL VOLTAGE OF GENERATOR APPLIED HERE

ONE LINE IMPEDANCE

ioxazx

7x 0.m

REACTANCE, X = 19% RESISTANCE. R = 1.4%

I

RESISTANCE I S LESS THAN OF THE REACTANCE BE NEGLECTED WITHOUT AN APPRECIABLE ERROR

HENCE MAY

INTERNAL VOLTAGE OF GENERATOR

-

NEARLY 90'

SHORT CIRCUIT CURRENT

DIAGRAM SHOWING SINE WAVES CORRESPONDING TO VECTOR DIAGRAM FOR ABOVE CIRCUIT

FIG. 1.13

Diagrams Illustrating the phase relations of voltage and short-circuit current.

14

SHORT-CIRCUll-CURRENT CALCULATING PROCEDURES

GENERATED VOLTAGE SHORT CIRCUIT CURRENT

ZERO AXIS

SHORT CIRCUIT OCCURRED AT THIS POINT

FIG. 1.14

Symmetric01 short-circuit current and generoted voltage for zero-power-factor

cirwit.

-SHORT CIRCUIT CURRENT

FIG. 1.15 circuit.

Asymmetrical short-circuit current and generated voltage in zero-power-factor Condition i s theoretical and is shown for illustration purposes only.


IS

If,in a circuit containing only reactance, the short circuit occurs a t any point except a t the peak of the voltage wave, there will be some offset of the current (Fig. 1.16). The amount of offset depends upon the point on the voltage wave at which the short circuit occurs. It may vary from zero (shown in Fig. 1.14) to a maximum (shown in Fig. 1.15). I n circuits containing both reactance and resistance, the s~,?&&,R&!~~ amount of offset of the shortCURRENT circuit current may vary between the same limits as for circuits containing only reactance. However, the point on the voltage wave a t which the short circuit must occur to produce maximum asymmetry dependsupon the ratioof reactance to resistance of the circuit. Maximum asymmetry is obtained when the short circuit occurs a t a time angle equal to 90" 0 (measured forward in degrees from the zero point of the voltage wave) where tangent 0 equals thereASYMMETRICAL actance-to-resistance ratio of FIG. 1.16 Short-circuit current and generated the circuit' The short-circuit voltage in zero-Dower-factor circuit. Short circurrent will be symmetrical cuit occurred between the when the fault occurs 90"from point and peak of the generated voltctge wove. that point onthe voltage wave. This condition i s theoretical and for illustration an example, assumeacir- purporer only. The short-circuit current will gradually become symmetrical in practical cuit that has equal resistance CiTCUit., and reactance, i.e., the reactance-to-resistance ratio is 1. The tangent of 45" is I ; hence, maximum offset is obtained when the short circuit occurs a t 135' from the zero point of the voltage wave (Fig. 1.17).

+

D-C COMPONENT OF ASYMMETRICAL SHORT-CIRCUIT CURRENTS

Asymmetrical alternating currents when treatedas a single current wave are difficult to interpret for circuit-breaker application and relay-setting purposes. Complicated formulas are also required to calculate their magnitude unless resolved into components. The asymmetrical alternating currents are, for circuit-breaker applications and relay-setting

16

SHORT-CIRCUIT-CURRENT CALCUUTING PROCEDURES

MAXIMUM OFFSET

FIG. 1.17 Short-circuit current and generated voltage in circuit with equal reactance and resistance. This condition i s theoretical and is shown for illustration purposes only. The short-circuit current will gradually become symmetrical in practical circuits.

purposes, arbitrarily divided into simple components, which makes it easy to calculate the short-circuit magnitude a t certain significant times after the short circuit occurs. The asymmetrical alternating current behaves exactly as if there were two component currents flowing simultaneously. One is a symmetrical a-c component and the other a d-c component. The sum of those two components a t any instant is equal t o the magnitude of the total asymmetrical a-c wave a t the same instant. The d-c component referred to here is generated within the a-c system with no external source of direct current being considered. I n some cases, particularly in the neighborhood of the d-c railways, direct current from the railways flows through neighboring a-c systems. This type of d-c current is not considered in this discussion or in the calculating procedures which follow. As an example of the resolution of asymmetrical alternating currents into components, refer to Fig. 1.15 which shows an asymmetrical shortcircuit current which is resolved into a symmetrical a-c and a d-c component in Fig. 1.18. If the instantaneous values of the two components (dashed lines) are added a t any instant, the resultant will be that of the asymmetrical current wave.


17

F I N S T A N T AT WHICH SHORT CIRCUIT OCCURS

ASYMMETRICAL

AC COMPONENT

FIG. 1.18 current.

Theoretical Ihort-circuit-cvrrent wove illustrating components of asymmetrical In practical circuits, d-c component would decay to zero in o few cycler.

INSTANT

OF SHORT CIRCUIT

TOTAL CURRENT

DC COMPONENT AC COMPONENT

ZERO A X I S

a = b = D C COMPONENT FIG. 1.19 Components of asymmetrical short-circuit current in which short circuit occurred at some point between the zero point and p e a k of the generated voltage wave. This is a lhsoretical condition similar to that shown in Fig. 1.18.

I8


As mentioned previously, the examples shown in Figs. 1.13 and 1.18 are for purposes of illustration only. In practical circuits the d-c component decays very rapidly, as shown in Fig. 1.20. INITIAL M A G N I T U D E OF D-C C O M P O N E N T

The magnitude of the d-c component depends upon the iustant, the short circuit occurs and may vary from zero, as in Fig. 1.14, to a maximum initial value equal to the peak of the a-c symmetrical compoiieiit, as i n Figs. 1.15 and 1.18. When the short circuit occurs at any other point, such as shown in Fig. 1.19, the initial magnitude of the d-c componciit is equal to the value of the a-c symmct,riral component a t thc instant of short circuit. The above limit,s hold true for the initial magiiitudc of d-c eomporient in a system regardless of the reactance and resistance. Ilowever, the d-c componeut does not continue to flo~va t a constant value, as shown i n Figs. 1.18 and 1.19, unless there is zero resistauce i i i the circuit. DECREMENT

There is uo d-c voltage in the system t o sustaiu the flax of direct current; therefore the energy represeuted by the dirert. component of current will be dissipated as ZZR loss from the direct current flowiug through the resistance of the circuit. If the circuit had zero resistance, the direct current would flow at a constant value (Figs. 1.18 and 1.19) TOTAL ASYMMETRICAL CURRENT C

COMPONENT AC

COMPONENT

FIG. 1.20 Trace of orcillogrom showing decay of d-c component and how orymmetricd short-circuit currenl gradually becomes symmetrical when d-c component diroppearr.


19

until the circuit was interrupted. However, all practical circuits have some resistance; so the d-c romponent decays as shown in Fig. 1.20. The combination of the decaying of d-c and symmetriral a-(*components gives an asymmetrical wave that changes to a symmetriral wave whcti the d-c component has disappeared. The rate of decay of the currents is called the decrement.

X/R

RATIO

The X / R ratio is the ratio of the reactance to the resistance of the circuit. The decrement or rate of decay of the d-c component is proportional to the ratio of reactance to resistance of the complete circuit from generator to short circuit. The theory is the same as opening the circuit of a battery and an inductive coil. If the ratio of reactance to resistance is infinite (i.e., zero resistance), the d-c component never decays, as shown in Figs. 1.18 and 1.19. On the other hand, if the ratio is zero (all resistance, no reartance), it decays instantly. FOFany ratio of reactarice to resistance in between these limits, the d-c component takes a definite time to decrease to substantially zero, as shown in Fig. 1.20. ! I n generators the ratio of subtransient reactance to resistance may be as ?much as 7 0 : l ; so it takes several cycles for the d-c component to disappear. In circuits remote from generators, the ratio of reactance to resistance is lower, and the d-c component decays more rapidly. The higher the resistance in proportion to the reactance, the more IaRloss from the d-c c.omponent, and the energy of the direct current is dissipated sooner. D-C TIME CONSTANT

Often it is said that generators, motors, or circuits have a certain d-c time constant. This refers again to the rate of decay of the d-c compoO C COMPONENT

a

C-

= 37Y. OF b (APPROX

TIME

CONSTANT I N SECONDS FIG. 1.21

OF D C COMPONENT

Graphic illustration of time constant.

)


20

nent. The d-c time constant is the time, in seconds, required by the d-c component to reduce to about 37 per cent of its original value a t the instant of short circuit. I t is the ratio of the inductance in henrys to the resistance in ohms of the machine or circuit. This is merely a guide to how fast the d-c component decays. Stated in other terms, it is the time in seconds for the d-c component to reach zero if it continued t o decay a t the same rate it does initially (Fig. 1.21). RMS VALUE INCLUDING D-C COMPONENT

The rms values of a-c waves are significant since circuit breakers, fuses, and motor starters are rated in terms of rrns current or equivalent kva. The maximum rrns value of short-circuit current occurs at a time of about one cycle after short circuit, as shown in Fig. 1.20. If there were no decay in the d-c component, as in Fig. 1.18, the rrns value of the first cycle of current would be j.732 times the rrns value of the a-c component. I n practical circuits there is always some d-c decay during the first cycle. An approximate rrns value of one cycle of an offset wave whether it is partially or totally offset is expressed by the equation

where C

=

a b

= =

rrns value of offset or asymmetrical current wave over one cycle rrns value of a-c component value of d-c component at one-half cycle

MULTIPLYING FACTOR

Calculation of the precise rrns value of an asymmetrical current a t any time after the inception of a short circuit may be very involved. Accurate decrement factors to account for the d-c component a t any time are required, as well as accurate factors for the rate of change of the apparent reactance of the generators. This precise method may he used if desired, but simplified methods have been evolved whereby the d-c component is accounted for by simple multiplying factors. The multiplying factor converts the rrns value of the symmetrical a-c wave into rms amperes of the asymmetrical wave including a d-c component. The magnitude of the d-c component depends upon the point on the voltage wave a t which the short circuit occurs. For protective-device application, only the maximum d-c component is considered, since the circuit breaker must be applied to handle the maximum short-circuit current that can occur in a system.


21

In the general case for circuits rated above 600 volts, the multiplying factor to account for d-c component is 1.6 times the rms value of the a-c symmetrical component at the first half cycle. For circuits rated 5000 volts or less where there is no local generation, that is, where the supply t,o the bus is through transformers or long lines, the multiplying factor to ralculate the total current at the first half cycle may be reduced to 1.5. For circuits 600 volts and less, t,he multiplying factor to calculate the total current at the first half cycle is 1.25 when the circuit breaker is applied on the average current in three phases. Where single-phase conditions must be considered in circuits GOO volts and less, then to account for the d-c component in one phase of a three-phase circuit a multiplying factor to calculate the total current at the first half cycle of 1.5 is used. For some calculations, rms current evaluations a t longer time intervals than the first half cycle, such as three to eight cycles corresponding to the interrupting time of circuit breakers, are required. Multiplying factors for this purpose may be taken from the curve in Fig. 1.22. Table 1.2 gives the multiplying factors commonly used for applying

e

FIG. 1.22 Charts showing multiplying factors to account for decoy of d-c component for various X / R ratio of circuits.

22

SHORT-CIRCUIT-CURREM CALCULATING PROCEDURES

short-circuit protective devices. These factors range from 1 t o 1.6, depending upon whether the short-circuit calculation is being made t o determine the interrupting or momentary duty on the short-circuit protective device. SHORT-CIRCUIT RATIO OF GENERATORS

This term is referred t o frequently in short-circuit discussions. With present AIEE procedures of short-rircuit ralrulations, it has become a n accessory with no practical significance from this standpoint. For the sake of completeness, a definition is given here. Short-circuit ratio field current t o produce rated voltage a t no load -~ field current t o produce rated current at sustained short circuit

No further mention will he made of short-circuit ratio. TOTAL SHORT-CIRCUIT CURRENT

The total symmetrical short-rirruit current is made up of currents from several sourves, Fig. 1.23. At the top of the figure is shown the shortcircuit current from the utility. This act,ually comes from ut,ility generators, but generally the industrial system is small and remote electrically from the utility generators so that the Symmetrical short-rircuit current is substant,ially constant,. If there are generators in the indust,rial plant, then they cont,ribute a symmet,rical short-circuit rurreiit which for all practical purposes is constant over the first few cycles. There is, however, a slight decrement, as indicated in Fig. 1.23. The other sources are synchronous motors which act something like plant generators, except that t,hey have a higher rate of decay of the symmetriral component, and induction motors whirh have a very rapid rate of dccay of the symmetrical component of current. When all these currents are added, the total symmetrical short-circuit rurrent is typical of that shown a t the bottom of Fig. 1.23. The magnitude of the first few cycles of the t,otal symmetrical shortcircuit, current is further increased by the presence of a d-c compouent, Fig. 1.24. The d-c component, offsets the a-c ware and, therefore, makes it asymmetrical. The d-c component decays t o zero within a few cycles in most indust,rial power systems. It is this total rms asymmetrical short-circuit current, as shown in Fig. 1.24, that must he determilied for short-circuit protective-derice appliration. The problem of doing this has been simplified by standardized procedures to a poiut xhere t o determine the rms asymmetriral current one need only divide t,he line-to-neutral roltage by the proper reactance


RG. 1.23 Tracer of orcillogramr of rymmetrical short-circuit currents from utility, panerator, synchronous motors, and induclion motors. The shape of the total combined currents is illurtmted by the bottom hace.

23

FIG. 1.24 Arymmelrical short-circuit current from dl sources illustrated in Fig. 1.23 plus d-c component.

24

SHORT.CIRCUIT-CURRENT U L C U U l l N G PROCEDURES

or impedance and then multiply by the proper multiplying factor from Table 1.2. BASIS OF RATING A-C SHORT-CIRCUIT PROTECTIVE DEVICES

The background of the circuit-breaker rating structure as well as the basic characteristics of short-circuit currents must be understood to enable the engineer to select the proper rotating-machine reactances and multiplying factors for the d-c component to determine the sbort-circuitcurrent magnitude for checking the duty on a particular circuit breaker, such as momentary duty or interrupting duty. The rating structure of circuit breakers, fuses, and motor starters is designed to tell the application engineer how circuit breakers, fuses, or motor starters will perform under conditions where the short-circuit current varies with time. In discussing these rating bases, and for the sake of clarity, they will be arbitrarily divided into two sections, i.e., the rating basis of high-voltage short-circuit protective devices above 600 volts and the rating basis of low-voltage Short-circuit protective devices 600 volts and below. HIGH-VOLTAGE SHORT-CIRCUIT PROTECTIVE DEVICES (ABOVE 600 VOLTS)

Power-circuit-breaker Rating Basis. The standard indoor oilless power circuit breakers as used in metal-clad switchgear will be used here t o explain power circuit-breaker ratings. The same fundamental principles apply to all other high-voltage power circuit breakers. The circuit-breaker rating structure is complicated because of the time of operation of the circuit breakers after a short circuit occurs. The few cycles needed for the power circuit breaker to open the circuit and stop the flow of short-circuit current consist of the time required for (1) the protective relays to close their contacts, (2) the circuit-breaker trip coil to move its plunger to release the breaker operating mechanism, (3) the circuit-breaker contacts to part, and (4)the circuit breaker to interrupt the short-circuit current in its arc chamber. During this time, the short-circuit current produces high mechanical stresses in the circuit breaker and in other parts of the circuit. These stresses are produced almost instantaneously in phase with the current and vary as the square of the current. Therefore, they are greatest when maximum current is flowing. The foregoing discussion showed that t,he short-circuit current is maximum during the first cycle or loop, because of the presence of the d-c component and because the motors contribute the most short-circuit current a t that time. Thus, the short-circuit stresses on the circuit breakers and other parts of the circuit are maximum during the first loop of short-circuit current. During the time from the inception of the short circuit until the circuitbreaker contacts part, the current decreases in magnitude because of the


25

decay of the d-c component and the change in motor reactance, as explained previously. Consequently, the current that the circuit breaker must interrupt, four or five cycles after the inception of t.he short circuit, is generally of less magnitude than the maximum value of the first loop. The fact that the current changes in magnitude with time has led to the establishment of two bases of short-circuit-current ratings on power circuit breakers: (1) the momentary rating or its ability to withstand mechanical stresses due to high short-circuit current and (2) the interrupting rating or its ability t,o interrupt the flow of short-circuit current within its interrupting element. What Comprises the Circuit-breaker-rating Structure. Circuitbreaker-rating structures are revised and changed from time to time. It is suggested that where specific problems require the latest information on circuit-breaker ratings the applicahlc American Standards Association (ASA), National Electrical Manufacturers Association (XEMA), or American Instituteof Elect,rical Engineers (AIEE) standards he referred to. To illustrate the various factors that comprise the circuit-breakerrating structure, an oilless power circuit breaker for metal-clad switchgear rated 4.16 kv 250 mva* has been chosen. The complete rating is shown on line 5, Table 1.1. The following will explain the meaning of the several columns of Table 1.1, starting at the left. The rircuit-breaker-type designation, column 1, varies among manufacturers. For the sake of completeness the General Electric Company nomenclature is used in this column. The remainder of the items are uniform throughout the industry. 1. Type of Circuit Breaker (AM-4.16-250) AM = magne-blast circuit breaker 4.16 = for 4.16-kv class of circuits (not applicable to 4800- and 4800volt circuits) 250 = interrupting rating in mva a t 4.16 kv

2-4. Voltage Rating 2. Rated kv (4.16): the nominal voltage class or classes in which the circuit breaker is rated. 3. Maximum design kv (4.76): the maximum voltage a t which the circuit breaker is designed to operate. The 4.16-kv circuit breakers, for example, are suitable for a 1330-volt system plus 10 per cent for voltage regulation or 4.76 kv. (Note: 4330 is 4% X 2500.) Some utility syst.ems operate a t 1330 volts near the substation. 4. Minimum operating kv a t rated mva (3.85) : the minimum voltage a t which the circuit breaker will interrupt its rated mva or in this case it is 3.85 kv. At any voltages below this value, the circuit breaker

* blegavalt-amperes t

i.

(see Appendix).

16


I !

I (

I

t

a /


27

is not designed to interrupt the rated mva but will interrupt some value less than rated mva. This is very significant in the rating of power circuit breakers for, as poiuted out later, the circuit hreaker will interrupt a maximum of only so many amperes regardless of voltage. At any voltage less than the minimum operating voltage the product of the maximum kiloampere interrupting rating times the kv times the square root of 3 is less than the mva interrupting rating of the circuit breaker. 5-6. Insulation Level (Withstand Test) 5 . Low-frequency rrns kv (19): the 60-cycle high-potential test. 6. Impulse crest kv (60) : a measure of its ability to withstand lightning and other surges. This is applied with an impulse generator as a design test.

7-9. Current Ratings in Amperes 7. Continuous 60 cycles (1200 or 2000): the amount of load current which the circuit breaker will carry continuously without exceeding the allowable temperature rise. 8-9. Short-time Rating 8. Momentary amperes (60,000) : the maximum rms asymmetrical current that a circuit breaker will withstand including short-circuit cnrrents from all sources and motors (induction and synchronous) and the d-c component. This rating is independent of operating voltage for a given circuit breaker. This is just as significant a limitation as mva interrupting rating. It defines the ability of the circuit breaker to withstand the mechanical stresses produced by the very large offset first cycle of the shortcircuit current. This rating is nnusually significant because the mechanical stresses in the circuit hreaker vary as the square of the current. It is the only rating that is affected by the square law, and therefore is one of the most critical in the application of the circuit breakers. The rating schedules of power circuit breakers are so proportioned that the momentary rating is about 1.6 times the maximum interrupting rating amperes. 9. Four-second (37,500): the maximum current that the circuit breaker will withstand in the closed position for a period of 4 sec to allow for relaying operating time. This value is the same as the maximum interrupting rating amperes.

10-13. Interrupting Ratings 10. Three-phase rated mva (250): the three-phase mva which the circuit breaker will interrupt over a range of voltages from the maximum design kv down t o the minimum operating kv. In this case the

28


interrupting rating is 250 rnva between 4.76 and 3.85 kv. The mva to be interrupted is obtained by multiplying the kv a t which the circuit breaker operates times the symmetrical current in kiloamperes to be interrupted times the square root of 3. The product of these must not exceed the rnva interrupting rating a t any operating voltage. 11. Amperes a t rated voltage (35,000): the maximum total rms amperes which the circuit breaker will interrupt a t rated voltage, i.e., in the case of the example used above 35,000 at 4.16 kv (4.16 X 35.000 x fi = 250 mva). These figures are rounded. This figure is given for information only and does not have a limiting significance of particular interest to the application engineer. 12. Maximum amperes interrupting rating (37,500) : the maximum total rms amperes that the circuit breaker will interrupt regardless of how low the voltage is. In this example, this current is 37,500 amp. At minimum operating voltage, 3.85 kv, this corresponds to 250 mva, and, for example, a t a voltage of 2.3 kv this corresponds to 150mva. The circuit breaker will not interrupt this much current a t all voltages, i.e., i t will not interrupt this much current if the product of current, voltage, and the square root of 3 is greater than the mva interrupting rating. This current limit determines the minimum kv a t which the circuit breaker will interrupt rated mva (column 4). At any voltage lower than that given in column 4, this maximum rms total interrupting current determines how much the circuit breaker will interrupt in mva. Therefore, when the voltage goes below the limit of column 4, the mva which the circuit breaker will interrupt is lower than the rnva rating given in column 10 by an amount proportional to the reduction in operating voltage below the value of column 4. 13. Rated interrupting time (8 cycles on 60-cycle basis): the maximum total time of operation from the instant the trip coil is energized until the circuit breaker has cleared the short circuit. What limits the Application of Power Circuit Breakers an on interrupting-and Momentary-duty Basis? In so far as applying power cir-

cuit breakers on an interrupting-duty basis is concerned i t can be seen from the foregoing that there are four limits, none of which should be exceeded. These must all be checked for any application. 1. Operating voltage should never at any time exceed the limit of column 3, Table 1.1, i.e., the maximum design kv. 2. Interrupting rnva should never be exceeded a t any voltage. This limit is sig’nificant only when the operating voltage is between the limits of columns 3 and 4, Table 1.1. It is not significant when the operating voltage is below the limit of column 4, Table 1.1, because maximum interrupting amperes limit the mva to values less than the rnva rating. 3. Maximum interrupting rating amperes should never be exceeded

SHORT-CIRCUIT.CURRENT CALCUUTING PROCEDURES

29

even though the product of this current times the voltages times the square root of 3 is less than the interrupting rating in mva. This figure is the controlling one in so far as interrupting duty is involved when the voltage is below that of column 4, Table 1.1 (minimum operating voltage a t rated mva). 4. Momentary current should never be exceeded a t any operating voltage. Modern power circuit breakers generally have a momeutary rating in rms amperes of 1.6 times the maximum interrupting rating in rms amperes. As a result, where there is no short-circuit-current contribution from motors, a check of the interrupting duty only is necessary. If this is within the circuit-breaker interrupting rating then the maximum Short-circuit current, including the d-c component, mill be within the momentary rating of the circuit breaker. Where there is short-circuit contribution from motors, the momentary rating of the circuit breaker may be exceeded, before the interrupting rating is exceeded in a given cirruit. Whenever there are motors to be considered in the short-circuit calculations, the momentary duty and the interrupting duty should both be checked. How to Check Momentary Duty on Power Circuit Breakers. Siuce the short-circuit current is maximum a t the first half cycle, the short-circuit current must be determined a t the first half cycle to determine the maximum momentary duty on a circuit breaker. To determine the short-circuit current a t the first half cycle, it is necessary to consider all sources of short-circuit current, that is, the generators, synchronous motors, induction motors, and utility connections. The subtransient reactances of generators, synchronous motors, and inductiou motors are employed in the reactance diagram. Since the d-r component is present a t this time, it is necessary to account for it by the use of a multiplying factor. This multiplying factor is either 1.5 or l.G, as outlined in Table 1.2. Typical circuits where the 1.5 multiplying factor can be used are shown in Fig. 1.25. The procedure is the same, regardless of the type of power circuit breaker involved. How to Check Interrupting Duty on Power Circuit Breakers. To check the interrupting duty on a power circuit breaker, the short-circuit current should be determined a t the time that the circuit-breaker contacts part. The time required for the circuit-breaker contacts to part will vary over a considerable range, because of variation in relay time and in circuitbreaker operating speed. The fewer cycles required for the circuitbreaker contacts to part, the greater will be the curreut to interrupt. Therefore, the maximum interrupting duty is imposed upon the circuit breaker when the tripping relays operate instantaneously. In all shortcircuit calculations, for the purpose of determining interrupting duties, the relays are assumed to operate instantaneously. To account for

SEPES-DIVEN SEN-RIO-EIELI', tCA

1


30

HIGH VOLTAGE INCOMING LINE

2400 4160 4800 VOLT INCOMING L I N E FROM UTILITY

(0)

$

T O P L A N T LOAD NO GENERATION IN THE P L A N T

o,:4600 A6,0 V BUS

TO P LANT L O AD NO GENERATION IN THE P L A N T

(b)

13.6 KV U

u.-L

U

USE 1.6 MULTIPLYING FACTOR NO GENERATION ON THIS BUS NO GENERATION

2400, 4160 OR

(C)

TO LOAD

FIG. 1.25 One-line diogrom of carer where the multiplying factor 1.5 may be used on circuits rated less than 5 h.

c

,,

,..

.:

.. . .


..

,

.

31

variation in the circuit-breaker operating speed, power circuit breakers have been grouped into classes, such as eight-cycle, five-cycle, three-cycle circuit breakers, etc. It is assumed for short-circuit-calculation purposes that circuit breakers of all manufacturers, in any one speed grouping, operate substantially the same with regard to contact parting time. Instead of specifying a time a t which the short-circuit current is to he calculated, it is determined by the simpler approach of specifying the generator and motor reactances and using multiplying factors. These factors are listed in Table 1.2. In industrial plants, eight-cycle circuit breakers are generally used. Normally, the induction-motor contribution has disappeared, and that of the synchronous motors has changed from the subtransient to the transient condition before the contacts of these circuit breakers part. Therefore, in calculating the interrupting duty on commonly used power circuit breakers, generator subtransient reactance and synchronous-motor transient reactance are used and induction motors are neglected. The elapsed time is so long that usually all the d-c component has disappeared. What d-c component is left is more than offset by the reduction in a-c component due to the increase in reactance of the generators. Hence, a multiplying factor of one (1) is used. In very large power systems, when symmetrical short-circuit interrupting duty is 500 mva or greater, there is an exception to this rule. In such large power systems, the ratio of reactance to resistance is usually so high that there may be considerable d-c component left when the contacts of the standard eight-cycle circuit breaker part. To account for this, the multiplying factor of 1.1is used in determining the total rms short-circuit mva that a circuit breaker may have to interrupt in these large systems. The multiplying factor of 1.1 is not applied until the symmetrical shortcircuit value reaches 500 mva. High-voltage Fuses. High-voltage fuses are either of the currentlimiting type, Fig. 1.26, which open the circuit before the first current peak, or of the non-current-limiting type, which open the circuit within one or two cycles after the inception of the short circuit. For the sake of standardization, all fuse-interrupting ratings are on the basis of maximum rms current that will flow in the first cycle after the short circuit occurs. This is the current that will flow if the fuse did not open the circuit previously, i.e., fuses are rated in terms of “available short-circuit current.” To determine the available short-circuit current a t the first cycle for the application of high-voltage fuses, use the subtransient reactances of all generators, induction motors, synchronous motors, and utility sources and allow for the maximum d-c component. The multiplying factor for allowing for d-c component is 1.6, the same as for allowing for d-c compo-

u w

TABLE 1.2

Condensed Table of Multiplying Factors and Rotating-machine Reactances

To Be Used for CaLdatina Swt-dreuit Cunanh for Circuit-breaker, Fuse, and Motor.rtartor Applicdons

1 Generators. 1

I

I

1

I

I

frequency changers

0

w

C

a

2

Interrupting duty Eight cycle or slower (general case). Rva cycle..

.......... Above 600 wlh

..............................

Any ploee where symmetricmi short-circuit kva i s loss than 500 mva

Above 600 volt,

I .O

Subtransient Subtransient

1.1

ii

s

Momentary duty

........................... ..........................

Generol GOSO.. Lar than 5 k..

Above 600 volt) 601 to 5000 volh

Near generoting station Remote from generating dolion (X/R rotio l e u thon I01

s z


1.6 1.5

High-voltaqe Fuses

5 Three-phose I n o interrupting duly

All typos, including dl wrront-limiting fuses.

All types, including dl current-limiting fuses.. Non-current-limiting lypes only..

.... Above 600 wih ... Above 600 volt'

............. 601 to 15,000 wlh

Anywhere in system

I

1

I .O

Subhqndent

1

Transient

1

Neglect

Maximum rms ampere interrupting duty Anywhere in system Remote from generoting %to. tion ( X / R mtio leu lhm 41

1.6 1 .?

i

i

i

Subtronsient Svbtronrient Svbwmrient Subwoniiont Subhmrient Subtransient

All h e p o w e r ratings..

....................

2400 and 4i60Y

Anywhere in system

Wlh

All horsepower rotingr..

....................

1.0

2400 and 4160Y

Anywhere in system

I .6

Yolh

CIrmit breaker w conladm l y p e . .

Cirwit b r w b r or contocto~lype. Clrcvit b r e e b r or contartor type..

...........

601 10 5000 volts

............ 601 to MOO volts ........... 601 lo 5000 volts

0 bywhere in system temote from gener.ting 11.lion lX/R ratio leis than 101

1.6 1.5

Subtransient Subtrmdent Subtransient Subtrmdent


R 0 m

z

Apparatus. 600 Volts and Below Interrupting or momentary duty Air circuit breakers or breaker-contactor combino. lion motor stoners.. Low-voltacp furas or fused combination motor

.................... Slarte" ...............................

8

600 volts and below Anywhere in system

I .25

Subtransient Subtianrient

600 volt* and below Anywhere in system

1 .25

Subtransient Subtransient Svbtraniient

Svbtronrienl

34

SHORT-CIRCUIT.CURRENT CALCULATING PROCEDURES

nent when determining the momentary duty on a power circuit breaker (see Table 1.2). The interrupting rating of fuses in amperes is exactly parallel, in so far as short-circuit+urent calculations are concerned, to the momentary rating of power circuit breakers. The ampere interrupting rating of high-voltage fuses is the only rating that has any physical significance. For the sake of simplicity of application in systems with power circuit breakers, some fuses are given interrupting ratings in three-phase mva. The three-phase mva interrupting rating has no physical significance, because fuses are single-phase devices, each fuse functioning only on the current which passes through it. WAVE OF AVAILABLE

THE FUSE ELEMENTS MELT BEFORE PEAK VALUE OF AVAILABLE SHORT CIRCUIT CURRENT I S REACHED

1 FIG. 1.26 Grophic sxplonotion of the current-limiting action of current-limiting fuses. See Fig. 1.27 for method of determining available short-circuit current.

SHORT-CIRCUIT-CURRENT CAKULATING PROCEDURES

35

These three-phase mva ratings have been selected so they will line u p with power-circuit-breaker ratings. For example, a high-voltage fuse rated 150 mva and a power circuit breaker rated 150 mva can he applied on the basis of the same short-circuit-current calculations. Of course, the application voltage must he factored in each case. High-voltoge M o t o r Starters. High-voltage motor starters generally employ for short-circuit protection either current-limiting fuses or power circuit breakers. The short-circuit-current calculations for applying these motor starters are the same as those for high-voltage fuses and power circuit breakers, respectively. LOW-VOLTAGE CIRCUIT PROTECTIVE EQUIPMENT (600 VOLTS A N D BELOW)

low-voltage Air Circuit Breokers. The present designs of low-voltage air circuit breakers differ from those of high-voltage power circuit breakers because they are substantially instantaneous in operation a t currents near their interrupting rating. The contacts often begin to part during the first cycle of current. Therefore, low-voltage air circuit breakers are subject to interrupting the current a t the first cycle after short circuit and withstanding the mechanical forces of that rurrent. It is necessary to calculate the current a t only one time for the application of low-voltage air circuit breakers. The current determined should be that of the first halt cycle and should be determined on exactly the same hasis as for checking the momentary duty of high-voltage power circuit breakers, except for a change in the multiplying factor as discussed in the next paragraph. The suhtransient reactances of generators, induction motors, and synrhronous motors are used, and the d-c component is considered (see Table 1.2). The multiplying factor for the d-c component is not so high in lowvoltage circuits as in some high-voltage circuits. This is due to the generally lower level of reactance-to-resistance ( X I R ) ratio in low-voltage circLits, which causes the d-c component to decay faster than in some high-voltage circuits. In rating low-voltage air circuit breakers, the average d-c component of the three phases is used, which is somewhat lower than that for the maximum phase. The generally lower ( X / R ) ratio and the use of an average d-c component for the three phases result in a considerably lower multiplying factor in low-voltage circuits. The multiplying factor has been standardized at 1.25 for the average for the three phases. This is equivalent t o a multiplier of about 1.5 to account for the d-c component in the maximum phase. Application of High-voltage Oil Circuit Breokers to 600-volt Systems. In the 192Os, 5-kv oil circuit breakers were used extensively on 600-volt

36

SHORT-CIRCUIT-CURRENT CALCULAnNG PROCEDURES

systems. The procedure for determining short-circuit currents in systems of 600 volts and below is slightly modified for checking duty on oil breakers of the 5-kv class as compared with low-voltage air circuit breakers. Both the momentary duty and interrupting duty must be checked for the oil-circuit-breaker application. To check the momentary duty, use the same procedure as for low-voltage air circuit breakers, i.e., generators, utility sources, induction motors, and synchronous motors (subtransient reactance). However, a multiplying factor of 1.5 is used instead of 1.25 as for low-voltage air circuit breakers. Oil-circuit-breaker momentary ratings are based on the maximum current through any one pole, not on the average current in the three phases which is employed in the rating of low-voltage circuit breakers. To determine the interrupting duty, use the generator subtransient reactance and utility-source reactance plus the synchronous-motor transient reactance and a multiplying factor of 1.0. Low-voltage Fuses. Several low-voltage fuses with published a-c interrupting ratings are appearing on the market. There are no industry standards to follow, but most of these seem to be following air-circuitbreaker standards, i.e., using the same rating base and same method of determining short-circuit duty as is used for low-voltage air circuit breakers. Hence, the procedure will not be repeated here except to point out that the 1.25 multiplying factor is used (see Table 1.2). So-called National Electrical Code (NEC) plug and cartridge fuses have no established a-c interrupting ratings. Many tests have been made to determine their a-c interrupting ability, but to date the industry has not applied a-c interrupting ratings. Low-voltage M o t o r Starters. Low-voltage motor starters are of two types: those using fuses and those using air circuit breakers for shortcircuit protection. Those using air circuit breakers for short-circuit protection are applied 04 exactly the same basis as low-voltage air circuit breakers in so far as short-circuit currents are concerned. Motor starters using fuses for short-circuit protection are applied on exactly the same basisas fuses in so far as short-circuit current is concerned. AVAILABLE SHORT-CIRCUIT CURRENT

In determining the short-circuit current, the impedance of the circuit protective device connected in the faulty feeder is neglected. The shortcircuit current is determined by’ assuming that the protective device is shorted out by a bar of zero impedance (Fig. 1.27). The short-circuit /


37

current which flows in such a circuit is commonly called available shortcircuit cumat. The procedure for determining the available short-circuit current is based on setting up impedance or reactance diagrams. The impedance of the short-circuit protective device that is nearest the short circuit (electrically) is omitted from the impedance diagram. Practically all protective devices are so rated and tested for shortcircuit interrupting ability; hence this procedure may be followed in short-circuit calculations. This greatly simplifies the calculations and removes the effect of impedance variations between different types and makes of devices having the same interrupting rating. I t means that one set of short-circuit-current calculations for a given set of conditions is all that is needed for applying any type of protective device, regardless of the impedance of the devices themselves.

0

GENERATOR

TRANSFORMER

MOTORS CABLE

SHORT ClRCUlTED 8 1 J UMPER OF Z E R O IMPEDANCE

CABLE SHORT

CIRCUIT

FIG. 1.27 Connections

for determining available short-circuit current for testing rhort-

circuit protective devices.


38

HOW TO MAKE A SHORT-CIRCUIT STUDY FOR DETERMINING SHORT-CIRCUIT CURRENT FORMULAS FOR SHORT-CIRCUIT STUDY'

1. Changing ohms to per cent ohms, etc.:

Per cent (%) ohms reactance

=

(90 ohms reactance

=

Per-unit

(ohms reactance) (kva.base) (1.1) (kvt)*(lO) (ohms reactance)(kva base) (kv)*(1000) (1.2)

[see Eq. (1.34)] Ohms reactance

=

( % reactance)(kv)2(10)

Per-unit ohms reactance

=

kva base per cent ohms reactance 100

(1.3) (1.4)

2. Changing per cent or per-unit ohms reactance from one kva base to another:

% ohms reactance on kva base 2 base 2 - kva X (% ohms reactance on base 1) (1.5) kva base 1 9f reactance on kva base 2 - kva base 2 X (% ohms reactance on kva base 1) (1.36) kva base 1 3. Converting utility-system reactance to per cent or per-unit ohms reactance on kva base being used in study: a. If given in per cent ohms reactance on a kva base different than that used in the study, convert according to Eq. (1.5). b. If given in short-circuit kva, convert to per-unit ohms thus:

kva base used in reactance diagram (1.6) short-circuit kva of utility system c. If given in short-circuit amperes (rms symmetrical), convert t o perunit ohms thus:

9i reactance

Yi reactance =

=

kva base used in reactance diagram (short-circuit current) ( d $ ) ( k v rating of system)

(1.7)

d. If only the kva interrupting rating of the incoming line breaker is known, * See pp. 54 to 57 for more prr-unit formulas

1 kv

= line-to-line kilovolts.

SHORTT-CIRCUIT.CURRENTCALCULATING PROCEDURES

39

9f ohms reactance -

kva base used in reactance diagram kva interrupting rating of incoming line breaker

(1.8)

4. Determining kva base of motors:

The exact kva base of a motor

=

EI 4 3

(1.9)

where E = name-plate voltage rating I = name-plate full-load current rating When motor full-load currents are not known, use the following kva bases: Induction motors: kva base = horsepower rating (1.10) 0.8-power factor synchronous motor: (1.11) kva base = 1.0 (horsepower rating) 1.0-power factor synchronous motors: (1.12) kva base = 0.80 (horsepower rating) 5. Changing voltage base when ohms are used: Ohms on basis of voltage 1 -

')* X (ohms on basis of voltage 2) (voltage 2)2

(1.13)

In Eqs. (1.1) to (1.4), ohms impedance or ohms resistance may be substituted for ohms reactance. The final product is then per-unit or per cent ohms impedance or resistance, respectively. 6 . Determining the symmetrical short-circuit kva: Symmetrical short-circuit kva

=

~

(kva base)

(1.14)

'& (kva base)

(1.15)

% X*

- y? -~

(line-to-neutral voltage)2 ohms reactance X 1000 kv2 X lo00 ohms reactance 7. Determining the symmetrical short-circuit current: (100) (kva base) Symmetrical short-circuit current = (% X*)(v%(kvt) kva base (% X*)(&)(kvt) k v t X lo00 ( d ) ( o h m s reactance) * X = reactance or impedanoe. t kv = line-&line kilovolts. = 3

(1.16)

(1.16a)

(1.17) (1.18) (1.19)

.

TABLE 1.3 Factor ( K ) to Convert Ohms to Per Cent or Per-unit Ohms for Three-phase Circuits*

0 L

Base kvo

loot Pr.

216Y/125 240 480

600 2,400 4.1 60

c*nt

1 50 Per-""it

'14 73 43.4

2.14 1.73 0.434

27.7 1.73 0.56

0.277 0.0173 0.00576

Per cent

321.5 260.4 65.21

200 Per-""it

Per cant

300

500

- __

Por-un1t

Per cent

Per-""it

Per cent

Per-""it

-0.712

3.215 2.604 0.6521

128 147 86.8

4.28 3.47 0.868

t43

4.166 2.604 0.808

0.4166 0.02604 0.00808

55.5 3.47 1.15

0.555 0.0347 0.0115

0.868 0.42 0.386

0.00868 0.0042 0.00386

1.302 0.63 0.579

0.01302 0.0063 0.00579

1.05 0.965

1.38 0.0868 0.0288 0.0217 0.0105 0.00965

30.2

6.43 5.21 1.302

071 868 217

83.3 5.21 1.72

0.833 0.0511 0.0172

I38 8.68 2.88

a1

8.68 2.17

4,800 6.900 7,200

0.435 0.210 0.193

0.00435 0.0021 0.001 93

0.651 0.289

0.00651 0.0031 5 0.00289

l1,OOO 11.500 12,000

0.0825 0.0755 0.0695

0.000825 0.000755 0.000695

0.123 0.113 0.104

0.00123 0.00113 0.00104

0.165 0.151 0.138

0.00165 0.00151 0.00138

0.247 0.226 0.208

0.00247 0.00226 0.00208

0.413 0.377 0.347

0.00413 0.00377 0.00347

12,500 13.200 13,800

0.064 0.0574 0.0525

0.00064 0.000574 0.000525

0.096 0.086 0.0787

0.00096 0.00086 0.000787

0.127 0.114 0.105

0.00127 0.00114 0.00105

0.192 0.172 0.157

0.001 92

0,00172 0.001 57

0.32 0.286 0.262

0.0032 0.00286 0.00262

23,000 37.4M) 46,000

0.0187 0.00711 0.00471

0,000187 0.000071 I 0.0000471

0.0283 0.0107 0.00708

0.000283 0.000106 0.0000708

0.0378 0.0142 0.00945

0.000378 0.000142 0.0000945

0.0567 0.0213 0.0141

0.000547 0.00021 3 0.0001 41

0.045 0.0355 0.0236

0.00045 0.000355 0.000236

0.0031 5

0.000031 5

0.0042

0.000042

0.0063

0.000063

0.0105

0.000105

69,OCU

0.0021 2 0.0000212 -

* For per-unit, K

=

0.315

-

kva base , kva base For per cent, K = kv' X 1wO kv' X 10

2.17

kv = line-to-line kilovolts

t To determine multiplying factors far any other base use figures under 100-kvs base columns multiplied by new base in kva, 100

v)

I

2 KE 2

B

-I

n

2 5 f

0

6c

R v,


41

8. Determining the asymmetrical short-circuit current: Asymmetrical short-circuit current = (symmetrical current) (multiplying factor) (1.20)

Asymmetrical short-circuit kva = (symmetrical kva) (multiplying factor) DIAGRAMS

One-line Diagram. The first step in making a short-circuit study is to prepare a one-line diagram showing all sources of short-circuit current, i.e., utility ties, generators, synchronous motors, induction motors, synchronous condensers, rotary converters, etc., and all significant circuit elements, such as transformers, cables, circuit breakers, etc. (Fig. 1.28). M a k e an Impedance or Reactance Diagram. The second step is to make an impedance or reactance diagram showing all significant reactances and resistances (Pig. 1.29). In the following pages this will be

I

C

GENERATOR

UTILITY SYSTEM TRANS

D

GENERATOR

CABLE E SHORT CIRCUIT LARGE MOTOR

CABLE J

480 VOLT MOTORS e diagram c

FIG. 1.28

, typical large industrial power system.

H

-SHORT

FIG. 1.29

INFINITE BUSES

CIRCUIT CURRENT GOES THROUGH HERE

Reactonce diagram of system shown in Fig. 1.28.

42

SHORT-ClRCUIT.CURRENT CALCULAltNG PROCEDURES

referred to as an impedance diagram, recognizing of course that only reactances will be used in many diagrams. The circuit element,s and machines considered in the impedance diagram depend upon many factors, i.e., circuit voltage, whether momentary or interrupting duty are to be checked, etc. The foregoing discussion and Table 1.2 explain when motors are to be considered and what motor reactances are to he used for checking the dut,y on a given circuit breaker or fuses of a given voltage class. There are other problems, i.e., (1) selecting the type and location of the short circuit in the system, (2) determining the specific reactance of a given circuit element or machine, and (3) deciding whether or not circuit resistance should be convidered. SELECTION OF TYPE AND LOCATION OF SHORT CIRCUIT

Three-phase Short Circuits Generally Considered. I n most industrial systems, the maximum short-circuit current is obtained when a three-phase short circuit occurs. Short-rircuit-current magnitudes are generally less for line-to-neutral or line-to-line short circuits than for the three-phase short circuits. Thus, the simple three-phase short-circuitcurrent calculations will suffice for application of short-circuit protective devices in most industrial systems. Unbalanced Short Circuits in Large Power Systems. In some very large systems where the high-voltage-system neutral is solidly grounded, maximum short-circuit current flows for a single phase-to-ground short rircuit. Such a system might be served from a large delta-Y transformer bank or directly from the plant generators. Hence the only time that single-phase short-circuit-current calculations need be made is on large high-voltage systems (2400 volts and above) with solidly grounded generator neutrals or where main transformers that supply a plant from a utility are ronnected in delta on the highvoltage side (incoming line) and in Y with solidly grounded neutrals on the low-voltage (load) side. The calculations of unbalanced short-circuit currents in large power systems can best be done by symmetrical components, see Chap. 2. Normally, generator and large delta-Y transformer secondaries are grounded through a reactor or resistor to limit the short-circuit current for a single line-to-ground short circuit on the system to letis than the value of short-circuit current for a three-phase short circuit. Bolted Short Circuits Only Are Considered. Several tests have been made to evaluate the effect of arc drop at the point of short circuit in reducing the short-circuit-current magnitude. It was felt by some engineers that the current-limiting effect of the arc was pronounced. These tests showed, however, that for circuit voltages as low as 300 volts

43


there may be no substantial difference in the current that flows for a bolted short circuit and when there is an arc of several inches of length. These test,s also confirmed modern calculating procedure as an accurate method of estimating the short-circuit-current magnitude in systems of 600 volts and less. .4rcs cannot be counted on to limit the flow of short-circuit currents even in louvoltage circuits; so short-circuit-current calculations for all circuit voltages are made on the basis of zero impedance at the point of short circuit, or, in other words, a bolted short circuit. This materially simplifies calculation because all other circuit impedances are linear in magnitude, whereas arcs have a nonlinear impedance characteristic. At What Point in the System Should the Short Circuit Be Considered to Occur? The maximum short-circuit current will flow through a cir-

cuit breaker, fuse, or motor starter when the short circuit occurs at the

4160V.

I

I

I

$?

$-

MAX.SHORT CIRCUIT DUTY ON BREAKERS ON THIS BUS $EW:RS FOR SHORT CIRCUIT

1 T ?;

A&?? Y T T - 3 &

r

+

y

* +

r-x

MAX. DUTY FOR THESE BREAKERS OCCURS FOR SHORT CIRCUIT HERE

FIG. 1.30

Location of faults for maximum Short-circuit duty on circuit breakers.

44


terminals of the circuit breaker, etc. (Fig. 1.30). These devices, if properly applied, should be capable of opening the maximum shortcircuit current that can flow through them. Therefore, only one shortcircuit location (at the terminal of the device) need be considered for checking the duty on a given circuit breaker, fuse, or motor starter. DETERMINING REACTANCES AND RESISTANCES OF CIRCUITS AND MACHINES

Typical reactances of circuit elements and machines are given at the end of this chapter. Resistances are included for certain items. These tables may be used as a basis for assigning values to the various elements of the impedance diagram. The reactances and resistances are all lineto-neutral values for one phase of a three-phase circuit. Where the reactances of a specific motor, generator, or transformer are known, these values should he used in lieu of the typical reactances in this chapter. The following is a guide to general practice in selecting and representing reactances. U s e R e a c t a n c e s of All S i g n i f i c a n t Circuit E l e m e n t s . Whether or not the reactance of a certain circuit element of a system is significant depends upon the voltage rating of the system where the short circuit occurs. In all cases, generator, motor, and transformer reactances are used. In systems rated above 600 volts, the reactances of short bus runs, current transformers, disconnecting switches, circuit breakers, and other circuit elements of only a few feet in length are so low that they may be neglected without significant error. In circuits rated 600 volts or less, the reactances of low-voltage current transformers, air circuit breakers, disconnecting switches, low-voltage bus runs, etc., may have a significant hearing on the magnitude of total shortcircuit current. As a general guide, the reactance of the low-voltage secondary-switchgear section in load-center unit substations with closely coupled transformers and secondary switchgear is not significant for all voltages of 600 volts and below. However, where there are several transformers or generators paralleled on one bus, or connections several feet long between a single transformer and its switchgear, reactances of the bus connections will generally be significant and should be considered in short-circuit calculations. I n systems of more than about 1000 kva on one bus a t 208Y/120 or 240 volts, reactance of all circuit components such as short bus runs, current transformers, circuit breakers, etc., should be included in the short-circuit study. I n systems of more than about 3000 kva on one bus a t 480 volts or 600 volts, reactances of all components such as current transformers, circuit breakers, short bus runs, etc., should be considered. It should be remembered that the lower the voltage, the more effective

SHORT-CIRCUIT-CURRENl CALCUUTING PROCEDURES

45

a small impedance is in limiting the short-circuit-current magnitude. That is why extreme care should he used to include all circuit elements in the impedance diagram, particularly for large ZORY/lZO-volt or 240-volt systems. If care is not used, the calculations will result in a value of current far higher than will actually be realized in practice. See the example outlined in Figs. 1.46 and 1.47. This often results in the adoption of low-voltage switchgear of higher interrupting rating and higher cost than are actually required. If care is used in including all reactances, the calculated reiults will be close to the short-circuit currents obtained in practice. Short-circuit calculations are of most value if they reflect accurate answers. When Is Resistance Considered? The resistance of all generators, transformers, reactors, motors, and high-capacity buses (above about 1000-amp rating) is so low, compared with their reactance, that their resistance is not considered, regardless of their voltage rating. The resistance of all other circuit elements of the high-voltage system (above 600 volts) is usually neglected, because the resistance of these parts has no significant bearing on the total magnitude of short-circuit currents. In systems of 600 volts and less the error of omitting resistances of all parts of the circuit except cables and small ampere rating buses is usually less than 5 per cent. However, the resistance of cable circuits is often the predominant part of the total impedance of a cable. When appreciable lengths of cable are involved in the circuit through which short-circuit current flows in a system of GOO volts or less, the resistance as well as the reactance of the cable circuits should be included in the GENERATOR

OF-THESE CIRCUIT ELEMENTS. IN GENERAL USF REACTANCE AND RESISTANCE OF THESE

SHORT CIRCUIT CURRENT CONSIDERING REACTANCE ONLY :20800 AMPERES

___

-. . . -. 1100 FT. 101

---(20 FT

SHORT CIRCUIT CURRENT CONSIDERING REACTANCE OF A LL PARTS PLUS RESISTANCE OF COW VOLTAGE CABLE = 11500 4MPERES.

FIG. 1.31 One-line diagram showing effect of resistance in cable circuits.

46


impedance diagram. The example of Fig. 1.31 shows the error that might result in neglecting cable resistance. I n secondary network systems of 600 volts and less, the resistance as well as the reactance of the tie-cable circuits between substation buses should be included in the impedance diagram. The example of Fig. 1.32 shows the effect of cable resistance in reducing short-circuit current in a typical industrial network.

n n SHORT CIRCUIT CURRENT USING REACTANCE ONLY = 51000 AMPERES, SHORT CIRCUIT CURRENT USING REACTANCE PLUS RESISTANCE OF T I E CIRCUIT= 41000 AMPERES.

T I E CIRCUITS 208 Y / l Z O V O L T S .

200 FT 2- 250 M,CM 3 CONO. CABLES ~~~~~T I N PARALLEL

200 F T

FIG. 1.32

One-line diogrtlm of low-voltage secondary network system showing effect of resistance of cable tie circuits.

Where to Use Exact Multiplying Factors. I n low-voltage systems having considerable lengths of cahle, the X / R ratio may be so low that the 1.25 multiplying factor would be considerably in error. Hence in these systems where resistance is considered, determine the correct X / R ratio and then use minimum multiplying factor. GUIDE FOR REPRESENTING THE REACTANCE O F A GROUP O F MOTORS

A group of motors fed from one substation or from one generating station bus may range in rating from fractional to several thousand horsepower per motor. All motors that are running at the time a short circuit occurs in the power system contribute short-circuit current and therefore should be taken into consideration. Motors Roted 600 Volts and Below. I n that portion of the power system operating at 600 volts or less, there are generally numerous small

A?


motors, i.e., under about 50 hp. I t becomes impractical to represent each small motor in the impedance diagram. These motors are constantly being turned off and on; so it is practically impossible to predict which ones will be on the line when a short circuit occurs. Furthermore, it would be impractical to obtain the characteristics of each small motor and to account for the effect of the impedance of their leads. Where more accurate data are not available, the following procedure may be used with satisfactory results for representing the combined reactance of a group of miscellaneous motors operating a t 600 volts or less. 1. In systems rated 240, 480, or 600 volts a t each generator and/or transformer bus, assume that the maximum horsepower of motors runniug a t any one time is equal to the combined kva rating of the stepdown transformer and/or generators supplying that one bus (see Figs. 1.33 and 1.34). 2. 10 systems rated 208Y/120 volts, a substantial portion of the load usually consists of lights and a lesser proportion of motor load than in 240-, 480-, or 000-volt systems. Hence in 208Y/120-volt systems where more accurate data are not available, assume a t each generator and/or transformer bus that the maximum horsepower of motors running a t

TO UTILITY SYSTEM

REbCTbNCE OF UTILITY SYSTEM REbCTbNCE OF 7 5 0 K V b TRbNSF.

QOW, OR5.,s

5.5%

0.25% OR

25 % REbCTbNCE OF EQUIVALENT MOTOR

IMPEObNCE O I b G R b M 750 K V b BASE SHORT CIRCUIT

EQUIVALENT MOTOR 750 KVb

240, 480, 600 VOLT SYSTEMS

El hKVA TO UTILITY SYSTEM

REbCTbNCE OF UTILITY SYSTEM

0.50% OR

50 % REACTbNCE OF EQUIVALENT MOTOR

SHORT CIRCUIT

FIG. 1.33

REbCTbNCE OF 7 5 0 KVb TRbNSF. EQUIVILENT MOTOR 375 K V b IMPEObNCE OIbGRbM 750 K V b BASE 2 0 8 Y / 1 2 0 VOLT SYSTEMS

Oiagromr illustrating how to include motors in low-voltage radial systems.


40

any one time is equal t,o 50 per cent of the combined rating of all stepdown trausformers and/or generators supplying power to that one bus, Fig. 1.33. For large commercial buildings the 50 per cent figure may be too low. Check carefully the mot,or load on all large 208Y/120-volt systems. I n the generalized rases referred t o in paragraphs 1 and 2 , no specific ratio of induction t o synchronous motors or no specific number of motors which prcduce unusually high short-circuit current,s has been set fort,h. T o account for these variables, a n average motor reitctance ihcluding leads is assumed t o be 25 per cent for the purpose of preparing application tables like Table 1.5 and in making short-circuit st,udies where no more accurat,e data are available. It will he noted that the average motor reactance of 25 per cent is based on the transformer or supply-generator kva rating. This figure is between the values of 28 per cent for induction mot,ors and 21 per cent for synchronous motors given in Table 1.14. Where the division between synchronous and iuduction motors is known, then more accurate calculations can be made by using the assumed motor reactances of Table 1.14. T h e reactances given in Table 1.14 are based on motor kva ratings and not supply transformer or generator ratings.

750 KVA

A 500 KVA

T

750 KVA

-480 VOLTS 500 KVA v

EQUIVALENT MOTORS WOULD BE 250 KVA AND 375 K VA FOR 280Y/120 VOLT SECONDARY SYSTEM

FIG. 1.34 rvrternr.

Diagram illustrating how lo include motors in lowvoltage secondary network


49

Although a portion of the load connected to a bus rated GOO voks or less may be heaters, lights, a-c welders, solderitig irons, appliances, arid other devices which produce no short-circuit curreiit, the total installed horsepower of motors connected t,o such a bus is geiierally much greater than the kva rating of the supply transformers and generators. Hovever, allowing for diversity, generally the total comhitied horsepower rat,ing of all mot,ors running a t one time ix-ould trot produce short-circuit currents in excess of the values obtained when using the ahore assumptions. I n systems of 000 volts or Icss, the large motors (i,e., mot,ors 011 t,he order of several hundred horsepomerj are usually few i n number and represent only a small portion of the tot,al connected horsepower; therefore, these larger motors are generally lumped in with the smaller motors and the complete group is represented as one equivalent motor i t i the impedance diagram. Synchrouous and induction motors need not be segregated when combining the motors in these low-voltage systems, because lorn-voltage air circuit breakers operr so fast that only the current flow duritig the first half cycle is considered; i.e., only suhtraiisient reactances ( X y ) of marhiiies are considered. Motors Rated above 600 Volts. High-voltage motors (rated 2200 volts and ahove) are generally larger in horsepower rating thau motors on systems operating under 600 volts. These largcr motors may have a much more significant hearing on short-circuit-current magnitudes than smaller motors, and, therefore, more exact determinatiou of the reactances of the larger motors is in order. Therefore, it is often foutid convetiient t o represent each large high-voltage motor individually in the impedance diagram. However, in large plants like steel mills, paper mills, etc., where there are numerous motors of several huridred horsepower each, it is often found desirable t o group these larger motors iii one group arid represent them by one reartaiire in the impedance diagram. Individual motors of several thousand horsepoitrer should be coiisidered individually and their reactances accurately determined hefore starting the short-circuit study. Whether considering motors individually or in groups, regardless of voltage rating of the motors, it is necessary t o obtain an equivalent kva rating of the individual or group of motors. This can be done precisely for large motors by Eq. (1.9) or can be approximated hy Eq. (l.lO), ( l , l l ) , or (1.12), when the full-load current is not known. The latter equations are used when considering a single reactance t o represent a group of miscellaneous motors.

50


I n high-voltage systems, complete motor data may not be available. Lacking these data, the connected horsepower is assumed to he equal t o the generator and/or transformer capacity supplying a given highvoltage bus. If the reactance of the leads between the transformer and/or generator bus and the motors is significant, the reactanre of these leads should be included. MAKING THE IMPEDANCE DIAGRAM

After it has been decided what elements of the one-line diagram are to be considered in the impedance diagram, the mechanirs of making the impedance diagram and of determining the short-circuit-current magnitude are as follows.

7

GENERATOR OR MOTOR OF ZERO IMPEDANCE

Treatment of Sources of Short-circuit Current. The generators and motors

are treated as if they comprised a generator of zero reactance plus an external reactor to represent the reactance of the EXTERNAL TO machine windings, Fig. 1.35. The first REPRESENT IMPEDANCE OF step in making an impedance diagram GENERATOR OR MOTOR. is torepresent every generator and motor or groups of motors and utility supply FIG. 1.35 One-line representation by a reactance connected to a zero imof generator or motor in impedance pedance bus or so-called “infinite bus,” diogmm. Fig. 1.36. This bus represents the internal voltage of the generators and motors. Completing the Impedance Diagram. The second step is to add the reactance of cables, buses, transformers, current transformers, circuit

flG. 1.36 Representation of reactances of generators, motors, and utility supply of system shown in one-line diagram form in Fig. 1.28.


51

breakers, switches, etc., in their proper location to complete the impedance diagram, top of Fig. 1.37. Choice of Ohms, Per Cent Ohms, or Per-unit Ohms Method. The next step is to decide whether to use ohms, per cent ohms, or per-unit ohms to represent the various circuit impedances in the impedance diagram. INFINITE BUS

SHORT CIRCUIT

INFINITE BUS

6.04V

STEP NO i COMBINE SERIES REACTANCES H

C+D=0.04+0.15~0.19% = 2 . O t ~0.0+0.10~ 12.10%

+1+J

STEP NO.:!

COMBINE PARALLEL REACTANCES

+

J) F,G AND I H + I = _' + L + I XI F G H + I + J

'

I

I

- _ 3 + p-j=j 2 . 5 + 0 . 2 + 0 . 0 8 3 -o,'40t

I

-=

XI

2.783 X =0.3698

STEP N 0 . 3 COMBINE SERIES REACTANCES X,,AND

X t = XI

STEP NO. 4

E

+ E = 0.36+0.04'0.40%

COMBINE PARALLEL REACTANCES X o , A . B . AND IC+D) 1 l +i

- 1 + 1 + L+-

I XR

XI A

B

C+D

'0.40 0 2 5

+I I+ 2.0

0.19

2 . 5 + 4 + 0 . 5 +5.3=12.3 X,

RESULTANT SINGLE REACTANCE

X

I ~

0.0805 % O ~ z

FIG. 1.37 Complete reaclomce diagram for system shown in Fig. 1.28. bining reactances into o single resultant value.

Steps for com-

52


Ohms are generally not used because of the difficulty of converting ohms from one voltage base to another without error and because of the very small numbers, which make accurate and easy calculation more difficult than the per cent or per-unit system. In many of the examples in this book, the assumed or given impedance or reactance data are listed in per cent, hut in the reactance dia,-rams these are converted to per-unit. N o notation will he made when that is done as it will be obvious. Equations (1.1) to (1.4) show how to convert ohms to per cent ohms, ohms to per-unit ohms. The Per-unit System for Electrical Calculations.* A per-unit system is a means of expressing numbers for ease in comparing them. A per-unit value is a ratio: a number (1.21) Per-unit = base number ~~

The base number is also called unit value since in the per-unit system it has a value of 1, or unity. Thus, base voltage is also called unit voltage. Any convenient number may be selected for the base number. For example, for the columns below, a base of 560 is used: Number 93 125 560 2053

Per-unit Volue with 560 as a Base

0.17 0.22 1 .oo 3.65

Each number in the second column is a per-unit part of the base number. In the first column, to compare the numbers, first mentally determine the ratio of one to the other. In the second column this is already accomplished. The comparison can be aided by selection of the base number which will illustrate the comparison best. In the foregoing example, if it is desired to show how much larger each uumber is when compared with the smallest number, the number 93 might have been selected as the base. This would then be obtained as follows: Per-""it Valve Number

with 93

( I ,

(I

93

I .oo

125 560 2053

I.35 6.00 22.20

Base

The value of a per-unit system is particularly useful when comparing

* From material originally Company.

prepared by H. J. Finison. iormrrly of General Ekctrir


53

numbers that are similarly related to two different base numbers. example : Core A 2300

Norm01 "0th Volts during motor starting

2020

For

Cole B 460 420

The above figures in themselves have little significance until they are compared each with its normal condition as follows: Vollr during starting per-unit of normal

0.91

0.88

Per Cent. Obviously per cent and per-unit systems are similar. The per cent system is obtained by multiplying the per-unit value arbitrarily by 100 to keep many frequently used per-unit values expressed as whole integers. By definition,

Per cent =

a number base number

x

100

(1.22)

Thus to change per cent to per-unit, divide by 100. For example, a transformer which has an impedance of 6 per cent has an impedance of 0.06 per-unit. The per cent system is somewhat more difficult to work with and more subject to possible error since it must always be remembered that the numbers have been arbitrarily multiplied by 100. For a simple example, money may draw interest a t the rate of 4 per cent per year. Early in arithmetic one learns to determine the interest by multiplying the principal by 0.04. It is thus necessary to remember to convert to the per-unit value before using the figure. In a complex calculation, this repeated conversion may invite errors. In effect it is safer and more convenient to say that interest is a t the rate of 0.04 per-unit. Impedances of electric apparatus are usually given in per cent. I t is usually convenient to convert these figures immediately to per-unit by dividing by 100 and thereafter do all calculating in terms of per-unit rather than attempt to remember always during the calculations whether a number should or should not be multiplied or divided by 100 to obtain the true value. Symbol. Just as the per cent system has a symbol (%) to designate that a given number is expressed in terms of per cent (as 6%) so also does the per-unit system have a symbol. The symbol for per-unit is (%). Thus 0.06 per-unit is written as 0.06 91. Selection of Base Number. In a per-unit system as used for expressing electrical quantities of voltage, current, and impedance, it is necessary to select numbers arbitrarily for the following: Base volts Base amperes

54


Do not then in addition arbitrarily select base ohms since it has already been fixed by the first two selections because of Ohm’s law. z = -E

I

base volts (1.23) base a m p z s Using the selected base values, all parts of an electric circuit or system may be expressed in per-unit terms as follows: volts Per-unit volts = (1.24) base volts amperes Per-unit amperes = (1.25) base amperes ohms Per-unit ohms = (1.26) base ohms In practice it is more convenient to select: Base volts Base kva The base values of other quant.ities are thus automatically fixed. Hence, for a single-phase system, base kva X 1000 Base amperes = (1.27) base volts base kva Base amperes = (1.28) base kv base volts Base ohms = (1.23) base amperes where base kva is single-phase kva and base volts is single-phase volts. For a three-phase system: base kva X 1000 Bme amperes = (1.29) X base voks base kva Base amperes = (1.30) 4 X base kv hase volts Base ohms = (1.31) X base amperes where base kva is three-phase kva, base volts is line-to-line, and hase ohms is per phase. Per-unit Ohms. In practice i t is desirable to convert directly from ohms to per-unit ohms, without first determining base ohms. By Ohm’s law, base volts Base ohms = (1.23) base amperes Base ohms

=

4


55

Substitute Eq. (1.27) (which gives the base amperes) into Eq. (1.23), to obtain base volts Base ohms = (base kva X 1000)/base volts (base volts)P Base ohms = bsse (1.32) kva x 1000 By definition: ohms Per-unit ohms = (1.26) base ohms Substitute Eq. (1.32) into Eq. (1.26) to obtain ohms Per-unit ohms = (base volts)e/(base kva X 1000) ohms X base kva X 1000 Per-unit ohms = (1.33) (base voltd2 ohms X base kva Per-unit ohms = (1.34) (base kv)2 X 1000 where base kva is single-phase kva and base kv is single-phase kv. When dealing with a three-phase system, i t is usual to select three-phase kva and line-to-line volts for the base values. Convert the above expressions to these bases to obtain ohms X base kva X 1000 X 3 Per-unit ohms = (base volts X d .3 ,)z ohms'X base kva X 1000 Per-unit ohms = (base volts)2 ohms X base kva Per-unit ohms = (1.35) (base kv)* X 1000 where ohms are per phase, kva is three-phase kva, and kv is line-to-line voltage. Usual Base Numbers for System Studies. If per cent or per-unit ohms reactance is used, the next step is to choose a kva base. In system studies it is usually desirable to select as the base voltage the nominal-system voltage or the voltage rating of the generators and supply transformers. Base kva will usually be selected as the kva rating of one of the machines or transformers in the system, or a convenient round number such as 1000, 10,000, or 100,OOO kva. After choosing the kva base, convert ohmic reactance of cables, wires, current transformers, etc., to per cent or per-unit ohms reactance on the chosen base, using Eq. (1.1) or (1.2) or Table 1.3. If ohms reactance is used, convert all per cent reactances to ohms by Eq. (1.3). Where two systems of differing voltage are interconnected through a

56


transformer, select a common kva base for both systems and the rated voltage of each system as its own base voltage. (These base voltages must have the same ratio t o each other as the turn ratio of the transformer connecting the two systems.) Base ohms and base amperes for the two systems will thus he correspondingly different. Figure 1.38 shows a typical example. Once the system values are expressed as per-unit values, the two interconnected systems may be treated as a single system and any calculations necessary carried out. Only in reconverting the per-unit values of the results to actual voltage and current values is i t necessary t o remember t h a t two different voltages actually existed in the system. Change of Base Number. Frequently the impedance of a circuit element may be expressed in terms of a particuiar base kva, and it may be desirable t o express it in terms of a different base kva. For example, the reactance of devices like transformers, generators, and motors is given in per cent on their own kva rating, and their reactances must be converted to the common base, chosen for the study by means of Eq. (1.5) or (1.36). Per-unit ohms on kva base 2 - base kva base kva 1

x

(per-unit ohms on kva base 1) (1.36)

Similarly, a machine rated a t one voltage may actually be used i n a circuit a t a different voltage. Its per-unit impedance must thus be changed to a new base voltage. GENERATOR 1000 KVA

MOTOR o(lOOO KVA)

I0;YKVA

13800 VOLTS

2300 VOLTS PRIMARY RATING

SECONDARY RATING

13200 VOLTS

2400 VOLTS

TRANSFORMER RATIO= 13 200/2400=5.5 (A)HIGH VOLTAGE SYSTEM

13 800

FIG. 1.38

BASE VOLTS

2500

BASE KVA

(A1 (El 5.5

RATIO

-

1000

I .o

41.6

EASE AMPS

233

115 5

190

BASE OHMS

6.2 5

I000

onother.

( 8 )LOW VOLTAGE SYSTEM

(5.5?

Method of converting bore volts, kva, amperes, and ohms from one value to

n Reference to Eq. (1.35) shows that per-unit ohms is inversely proportional to the square of base volts. Thus: Per-unit ohms on new base volts - (old base volt.s)* (1.37) Per-unit ohms on old base volts (new base volts)* and Per-unit ohms on new base volts = per-unit ohms on old base volts (old base volts)2 (1.38) (new base volts)2 Equations (1.37) and (1.38) may be used for per cent ohms as well as perunit ohms. Converting Ohms to a Common Voltage Base. When using ohms instead of per cent or per-unit in the impedance diagram, it is important to convert the ohmic values to a common voltage base by Eq. (1.13). For example, if the short-circuit current is being calculated in a 480-volt system (supplied by transformers rated 480-volt secondary) fed through a cable and a transformer from a 2400-volt system, the ohms impedance of the cable in the 2400-volt circuit must be multiplied by 48O2/24OO2to convert it to ohms on a 480-volt base. The transformer ratings, i.e., 480, 240, etc., and not system ratings, if different from transformer rating, are used as the voltage base for short-circuit-current calculations. Representing the Utility Supply System. The utility system must be represented by a reactance in the impedance diagram. Sometimes this utility-system reactance is available in per cent on a certain base. If so, it is merely necessary to convert this value to the common base used in the impedance diagram. To do this, use Eq. (1.5). In some cases the utility engineers will give the short-circuit kva or current that the utility system will deliver a t the plant site. In otker cases, only the interrupting capacity of the incoming-line circuit breaker is known. In these cases to convert short-circuit kva, current, or incoming-line breaker interrupting rating to per cent reactance on the kva base used in the reactance diagram, proceed as follows: If given short-circuit kva, convert to per cent by using Eq. (1.6). If per-unit is desired, use also Eq. (1.4). If given short-circuit amperes (rms symmetrical), convert to per cent by Eq. (1.7) and to per-unit by Eqs. (1.7) and (1.4). If only the kva interrupting rating of the incoming line circuit breaker is known, convert to per cent by Eq. (1.8) and to per-unit by Eqs. (1.8) and (1.4). SHORT-CIRCUIT-CURRENT CALCULATING PROCEDURES

DETERMINING THE EQUIVALENT SYSTEM IMPEDANCE OR REACTANCE

After completing the impedance diagram and inserting the values of reactance or impedance for each part of the diagram, it is necessary to reduce this network to one equivalent value. This can be done either by


58

longhand calculation or with the aid of a calculating board. Since so few engineers have access to calculating hoards and must use longhand methods, this method will be covered in sufficient detail to enable solving the short-circuit problems commonly encountered. Use of Calculating Boards. A d-c calculating board will permit accurate solution of all short-circuit problems where reactance only is considered. In most cases where resistance is a significant factor and must be considered, the d-c calculating board cannot be used readily. However, in some problems involving resistance, certain approximations can be made to obtain reasonably accurate answers on d-c calculating boards. For exact calculating-board solutions of problems factoring resistance and reactance, the a-c calculating board may he employed. A-c calculating boards have boxes to represent both the resistance and reactance of a circuit. The procedure for using calculating boards is beyond the scope of this book. Longhand Method of Combining Reactances. Longhand methods of combining reactances vary in some respects. To illustrate the principles involved, refer to Figs. 1.37 and 1.39. Arbitrary values of reactance have been assigned to the various branches. Combining the various branches of the diagram is merely a question of reducing two or more series reactances to one value and reducing two or more parallel reactances to one value until one single equivalent value is obtained. The following shows how to combine reactances and resistances. 1. Combining reactance and resistance to determine impedance, z

=

m

(1.39)

z=r+jz

wherej = 47 2. Adding series reactance of circuits where resistance is neglected add reactances arithmetically, i.e.,

x,

+ x2 + xa = x.

= equivalent reactance z,,z2, and x 3 = reactances of circuit components zs= equivalent reactance

3. Combining parallel reactances, zo = equivalent reactance

For two reactances only x, and z2 XI =

(d(z2) 22

21

+

For combining several parallel reactances 1 1 1 1 1 -=_ 2.

2,

+ - + - +X -I + E 2, 2 2

(1.40) 1

(1.41)


59

INFINITE

c. REACTANCE DIAGRAM OF CIRCUIT SHOWN IN ONE LINE DIAGRAM TO THE LEFT.

ONE-LINE DIAGRAW

P~T&

$*,

T(

* T . Pm

EQUIVALENT Y CONVERT P I T I , PITI e c , TO EQUIVALENT Y. STEP x z

c. STEP# I COMBINE SERIES REACTANCES PI~TI,RBT~,ETC.

I

a"

--&&Pa.

L

3+c*

Ct

c4

cs

DRAW NEW DIAGRAM STEP-* 3

+

COMBINE 2 C t , 3 + C+ AND THEN REPEAT STEPS 2.3 e 4 UNTIL ONE EOUIVALENT REACTANCE IS OBTAINED. STEP t t 4

FIG. 1.39 Example of the method of combining remtmces of a network-type system into a single resultant value.

MI


Some systems are such that they cannot he reduced by merely combining series and parallel rgactances. For example, take the one-line diagram of a circuit as show in upper left-hand corner of Fig. 1.39. The reactance diagram is shown the ypper righehand corner of Fig. 1.39. In addition to combining serieszind parallel reactances, it is necessary to TI, PzrT , and C1to an equivaconvert a triangle of reactances such as PI, lent Y of reactances by the formulas of Fig. 1.40. By these conversions,

\

I

B=

c=

ob

+ a c + be

a=-

b

a b + a c + bc

b:

A = ob+oc+bc

C:

a

0c A+B+C

"

A+B+C A8 A+B+C

FIG. 1.40 Formula for converting a triangle or delta of three impedances to a Y of three equivalent impedances, and vice verso.

any commonly encountered system reactance diagram can be reduced to one equivalent reactance. Combining Impedances. Sometimes i t is desirable to consider the resistance and reactance of a circuit. This involves combining impedances. The procedure for combining impedances is outlined here. The combining of parallel impedances necessitates multiplication and division of impedances (complex quantities) and is outlined here. Adding Series Impedances. When two or more impedances are in series, the resistance and reactance components are added separately to combine the series into one equivalent value. Refer to Fig. 1.41. The three series impedances are

+ + ja

z1 = TI jzl za = 72 ijxa

zz = Tp

61

SHORT-ClRCUIT+CURRENT CALCULATING PROCEDURES

EQUIVALENT IMPEDANCE

3 SERIES IMPEDANCES

FIG. 1.41

Example illustrating the combining

The equivalent impedance

+ + +

rl VZ 73 j(z1 Using the numerical values of Fig. 1.41, 2 %=

2,

1+ j 2

22

2 +j3 0.5 + j l

= = = 21 =

(1

of series impedances.

+ zz+ 4

(1.42)

+ 2 + 0.5) + j ( 2 + 3 + 1) = 3.5 + j G

The above is applicable when impedances are expressed in ohms, perunit or per cent. Combining Parallel Impedances. Parallel impedances may be reduced to one equivalent impedance as follows (see Fig. 1.42):

TWO PARALLEL IMPEDANCES

FIG. 1.42

Example illustrating L e combining

EQUIVALENT IMPEDANCE

of parallel impedances.

61

WORT-CIRCUIT-CURRENT CALCULATING PROCEDURES

(1) Reduce the per cent values of resistance and reactance in each of the given parallel circuits to a per-unit basis by dividing per cent figures by 100 or convert the per cent values to ohms. Per cent values can be used in the following if the multiplier 100 is applied properly, e.g., X

T

(Branch 1) 0.05 (Branch 2) 0.008

0.15 0.108

(2) Calculate the impedance squared z2 of each circuit 2% =

(Branch 1) rlz (Branch 2) r 2

r'

+

2 '

1

+ = ZI', e.g., 0 .052+ 0.1547-0>25 + zz2= zz2, e.g., 0.008z+ 0.108* = 0.0117 21'

(3) Obtain the ratios of r/z' of each circuit Tl 0.05 (Branch 1) -', e.g., -= 2.0 21 0.025 rz (Branch 2) -, e.g., 0.0°8 - 0.683 z'2 0.0117 ~

(4) Add the foregoing r / z z = Ga = 2.683

(5) Obtain the ratios of x/z* for each circuit 21 0.15 (Branch 1) -2 e.g., -= 6 21 0.025 XP 0.108 (Branch 2) e.g., -= 9.2 22 0.0117 (6) Add the foregoing 7 j

X/L'

(7) Ya2 = 02

(8) ra

=

=

Ba

=

15.2

+ Ba2,e.g., = 2.683' + 15.24 = 238.2

Ga e.g.,

-9

Y3'

BJ e.g.,

;2

=

~

- 0.0112

15" = 0.0642 __ 238.2 The foregoing may be tabulated for convenience in solving a number of parallel pairs of circuits: r z z4 = r' z2 r/z' 2/22 (Branch 1) 00 0 0 0 (Branch 2) 00 0 0 0 (Branch 3, etc.) ( ) ( ) ( ) ()() By addition- Go( )Bo( )

(9) xa

=

-2

Ya2

=

+

The combination of the circuits results in


63

1

Any number of parallel circuits may be accommodated by additional horizontal columns as fo branch 1 and branch 2, etc., their resultant (r/z2)’s and (x/z2)’s heling added to obtain G O and Bo. Multiplying and Dividing Impedances. Two impedmces may be multiplied as per the following equations: (21) 21

(22)

=

=

23

TI +jXl

ZP = T S

+j x , + +

= r8 jxa 2 3 = (TI jXl)(Tt = (TIT2 - 2 1 2 2 ) 13 = (nrz - XIXZ) j a = j(r1zz rczJ 23

+

+ +

jZ2) j(TIX2

1

+

TBZL)

(1.44)

Two impedances may be divided according to the following equations:

TI +j x , =-x-

r2

+j x 2

TZ TZ

- jxt

- jxt (1.45)

DETERMINING THE SHORT-CIRCUIT-CURRENT MAGNITUDE

After the reactance diagram has been reduced to a single value, the value of symmetrical short-circuit kva can be determined by Eq. (1.14), (1.15), or (1.16). To determine the symmetrical short-circuit current, use Eq. (1.17), ( l . l S ) , or (1.19). Equations (1.14) to (1.19) do not allow for any d-c component. Table 1.4 gives figures for converting kva to amperes. Apply Proper Multiplying Factor. The final step is to apply the proper multiplying factor from Table 1.2. To determine the total rms short-circuit current or kva, use Eq. (1.20).


64

TABLE 1.4 Three phore line-to-line, volh

Amperes per Kva Amperes rer phase per kvo

Amperes ier phase per kva

"0 wire

V d h

c or d-l

Amperes er kro or d-c kw

110 115 120

5.25 5.02 4.81

13.200 13,800 14,400

0.0437 0.0419 0.0401

24 48 110

180 I99 208

3.21 2.90 2.78

22,000 23,000 24,000

0.0263 0.0251 0.0241

115 120 125

8.70 8.33 8.00

220 230 240

2.63 2.51 2.41

33,000 34,500 36,000

0.0175 0.0167 0.0160

220 230 240

4.55 4.35 4.17

440 460 480

1.31 I .25

1.20

44,000 46,000 48,000

0.0131 0.0125 0.0120

250 275 300

4.00 3.64 3.33

550 575 600

1.05 1 .oo 0.962

66,000 69,000 72,000

0.00875 0.00838 0.00803

440 460 480

2.27 2.17 2.08

1,100 1,150 1,200

0.525 0.502 0.481

I I0.000 120,000

0.00525 0.00502 0.00481

550 575 600

I .82

1 I5.000

2,200 2,300 2,400

0.263 0.251 0.241

132,000 138,000 144,000

0.00437 0.0041 9 0.00401

650 750 1,200

1.54 1.33 0.833

3.300 3.450 3,600

0.175 0.167 0.160

154,000 161,000 168,000

0.00375 0.00359 0.00344

1,500 2,200 2,300

0.666 0.455 0.435

3,800 4,000 4.160

0.152 0.144 0.138

220,000 230,000 240,000

0.00263 0.00251 0.00241

2,400 3,000

0.417 0.333

6,600 6.900 7.200

0.0875 0.0838 0.0803

330,000 345,000 360,000

0.00175 0.00167 0.00160

11,000 11,500 12,000

0.0525 0.0502 0.0481

-

41.7 20.8 9.10

I .74 1.67

~

/


65

EQUIVALENT CIRCUITS

The redurtion of impedance diagrams to a single value of impedance can he greatly simplified by using equivalent circuits for duplex reactors and three-winding transformers. Equivalent Circuit for Duplex Reactors. The duplex reactor consists of two sections of winding per phase on the same core, with a t a p brought out from the junction point. The current ratings and reactances of the two sections arc generally equal. Aside from the midtap connections, whirh necessitate a total of nine leads, the construction is similar to that of the series reactor. If 1, and l2 are the self-inductances ( X , and X , are the corresponding reactances) of the individual sections, and f c is the “coupling factor” of the mutual inductance betmeen sections, then the simplified equivalent

LEX REACTOR

O N E LINE DlAGRPlM I

J

k

GENERATOR

-XI

FIG. 1.43

fc

One-line diagram and equivalent circuit for duplex reactor.

66

SHORT-CIRCUIT-CURRENT CALCUUllNG PROCEDURES

circuit for the duplex reactor is as shown in Fig. 1.43. For preliminary calculations, an average figure off. = 0.5 should give results of sufficient accuracy. Equivalent Circuit of Three-winding Transformer. When making short-circuit calculations of power systems which include three-winding transformers, there is a question on how to use the designer's reactance values. Designers give reactance values between pairs of windings. Figure 1.44A shows a three-winding transformer, and Fig. 1.44B shows its equivalent circuit. The following equations are easily derived and are the proper ones to use in short-circuit studies:

x. = x,. + 2 xs = + X2e c x,= + XAC

XIB

XBC

XdC

XBC

XAC

(1.46)

XdB

2

All reactance6 must be on same kva base. NOTE:The equivalent circuit and equations for a four-winding transformer are more complicated and will not he evident by simple analogy from Eq. (1.46).

,

(A1

FIG. 1.44 (A1

A

mi

One-line diagram and (61 equivalent circuit diagram of three-winding

transformer.

EXAMPLES OF SHORT-CIRCUIT-CURRENT CALCULATIONS'

The following examples are indicative of methods of applying the shortcircuit-current calculating procedures outlined in the foregoing. Systems 600 Volts and Below. The system shown in Fig. 1.45 involves one source of supply through a transformer from a primary system. The kva base for the short-circuit calculations is taken as the kva *NOTE:Numbers in parentheses in Figs. 1.45 and 1.47 to 1.50 refer to numbers of formulas used.

67

SHORT-CIRCUIT-CURRENT CALCUVITING PROCEDURES

INCOMING LINE

A

A

0.25 Yt

SOURCE

MOTORS

TRANSFORMER

I

750 KVA 5.5 x x (0.055%)

? T ? ?

480 VOLTS

REACTANCE DIAGRAM USE 750 KVA BASE FOR CALCULATIONS

M$ (0)

SOURCE REACTANCE ON 750 KVA BASE

0.0625 1

2

5

(1.61

1x=--XIXI% +x2-0.0625t025 0.0625XC125-0,05% T

v

750 - yj;xo,4& I

loo,ooo 750 - 0.0075%

:

5

%

(d)

18,000 AMPERES SYMMETRICAL [ 1.18)

X

o,050 18,000 X 1.25"22.500 AMPERES ASYMMETRICAL (1.201

(el FIG.1.45 Illustration of procedure for calculation of short-circuit currents in radial loadcenter system.

68

SHORT-CIRCUIT-CURREHI CALCULATING PROCEDURES

rating of the transformer. The kva of the connected motors is assumed to be 750 with an equivalent reactance of 25 per cent. Only reactances are used in these calculations. This problem is the type on which Table 1.5 is based. Large 208Y/120-volt Systems. Problems, particularly those involving secondary-network systems in the downtown area of the large cities or in large buildings, require the determination of the short-circuit current on a 208Y/120-volt basis. In these systems it is particularly important that the reactance of all circuit elements, however small, be taken into account, as they have a much more significant effect in reducing the short-circuit current a t 208Y/120 volts than a t 480 or 600 volts.

FEEDERS BREAKERS

PLAN

CHANNEL B U $ - 4 0 0 0 A

150'

n

w

I

y

Z

NETWORK TRANSFORMER

1 3 2 0 0 - 2 1 6 ~ / I 2 5 VOLTS KvA

INCOMING LINE 5 O YVA SC

II

nus o' NETWORK PROTECTOR Z500 A

L-

CIRCUIT BREAKER

.""... CD"Yl

rnY"l

4000

ELEVATION

FIG. 1.46 Arrmgement of equipment for large 208Y/120-volt spot network system.


69

The equipment for this example is arrauged as shoirn iir Fig. 1.46. The one-line diagram is shown in Fig. 1.47.4 which iurludes the hayir reartanre data on the circuit elemenk. The impedauce diagram is shown i n Fig. 1.47B. Figure 1.47C shows the condensed diagram to illustrate t,he relative distribution of reactance in the system. It will be noted t,hat the overhead bus R has 70 per cent as much impedance as the romhinatiotr of all the transformers an8,huses ahead of it,. Elimiiiatiug this item would cause a serious error in t h magnitude of short-circuit, curretit. The intermediate steps etween Figs. 1.47H and 1.47C can be worked out by followiug t h e fa oing text. The short circuit is located just ahead of the maiii 4000-amp circuit breaker as this determiires the available short-circuit, curreut, which this circuit breaker must interrupt. As pointed out previously, air circuit breakers are applied 011 the basis of availahle rurreiit, and therefore \\.heir calculat,ing the short,-rirruit duty oil them, t,he impedalire of t,he rirciiit breaker is not included. Large High-voltage Power System. T h e examplc shown in Fig. 1.48 is typical of what might, be eucouritered i n a steel mill. The kva base chosen is 100,000 kva. Precise data are available 011 large motors and are used in the short,-circuit, st,udy. Since the large mot,ors roiistitute only part of the motor load, the remaining motor load is estimated. For short circuits on the 22-kv system t,he motor load is assumed to be equal to the capacity supplying each 22-kv bus, or 62,500 k r a aiid 20,000 kva. Should more precise data be available regarding ronnevted mot,or load, these data should be used for simulating motor ront,ribution for faults on the 22-kv system. In t,his example, the connected horsepower 011 the 6.0-kv bus mas known t,o be as shown in t,he diagram. To check the momentary dut,y at F , 011the KY-kv bus, the primary system should be represented by its equivalrut, subt,raiisieiit reartaure nf 12.2 per cent. For interrupting d u t y on the 6.9-kv bus, t,he primary syst,em should be represented by a reartanre equivalent t o the iirterruptirig duty on t,he 22-kv system, or 17.5 per cent. These large complicated syst,ems should he set up 011 a calculating board to enable accurate ausivers t,o he obtained easily.

J

SHORT CIRCUITS IN SINGLE-PHASE LIGHTING A N D WELDING POWER SYSTEMS (600 VOLTS A N D LESS)

A common p r a h c e is t o use single-phase trausformers roiiuected to three-phase primary systems t,o supply single-phase loiv-voltage power for welders and for lightirrg rircuits in some of the older syst,ems. When determining the short-circuit current a t the serondaries of these transformers, it, is necessary t o use the proper impedance t o represerrt the primary system. I n three-phase short-circuit calculations, the reactance

70

SHORT-CIRCUIT-CURRENT CALCUATING PROCEDURES

FIG. 1.47

One-line diogram, reactance diagram,


71

and short-circuit-current calculation procedure for spot network system show in Fig. 1.46.

72


c:

74


of a conductor is the reactance from the center of the condurtor to the theoretical neutral. Assume that for eaeh phase the rurrent leaves on the phase conductor and returus through the neutral. In a three-phase short circuit, the three currents balance; so there is no rurrent flowing in the neutral. With single-phase line-to-line short cirruits, the eurreut leaves on one phase conductor and returns ou the other. Therefore this rurrent sees the reactance of two condurtors as beiug in series. Heure, for siuglephase tramformers conuected line-to-hie on the primary, twire the primary system impedance must be used to represent it in a true relation to the rest of the circuit. The remaining calculatious are essentially the same as for three-phase circuits using the transformer and loiv-voltagecircuit reactances. Single-phase tramformers used for supplying 120/240-volt single-phase lighting circuits usually have the midtap available for ronnerting to threemire neutral and ground by the user and are usually relatively low iu kva. These small transformers have a relatively high resistatire-t~reactance ratio compared with three-phase trausformers of a higher seroridaryvoltage rating and of larger kva rating.

7

100,000 KVA 3 PHASE SHORT CIRCUIT OUTY

BASE 500 KVA

4tL PRIMARY SYSTEM REACTANCE ON 3-PHASE BASIS.

PRIMARY SYSTEM REACTANCE ON SINGLE PWSE BASIS = 0.005X 2 * 0.01Va

i

PRIMARY SYSTEM X

:0.01%

TRANSFORMER X =O.O3Ym

TOTAL X ~ 0 . 0 4 %

1%'

o , 0 4 ~ ~ , , e o :%:26000 AMP SYMMETRICAL 0.0192 11.18 MODIFIED)

1.25 X 26000

FIG. 1.49 system.

= 32500

A M P ASYMMETRICIL K2Ol

Short-circuit-current calculating procedure for single-phase two-wire 480-volt


75 .

The most severe short-circuit condition in this case is a line-to-neutral short circuit because it involves a much higher primary-to-secondary turn ratio than does a line-to-line short circuit. Hence, this is the basis on which protective equipment should be selected. Since the reactance and resistance of the transformers are given on the basis of a full winding, it is necessary to convert to the proper values when only one-half the secondary winding is involved as is the case when a line-to-line neutral short circuit occurs. The reactance is increased by a factor of 1.2 and the resistance by a factor of 1.41. Therefore, the published reactances and resistances of these transformers are multiplied by those figures. Figure 1.49 shows a typical example where reactance only is used, as would be the case for a relatively large 480-volt transformer supplying a welder circuit. I n these calculations it is necessary to use twire the lineto-neutral reactance of the primary system. In the example of Fig. 1.50 use twice the line-to-neutral reactance of the primary. Use the proper

F+

100 000 KV4 3 PH4SE SH& ClRCUlT DUTY

1

120,240-V )IR X =:3 1.2% X ON ,. FULL ,. WINDING 84SIs

B4SE 5 0 KV4

-

PRIM4RY SYSTEM RE4CT4NCE ON 3 P H 4 K 84%

0.00198 PRlM4R"X

:0

0005~~

II 61

TR4NS X 0036% PRIM4RI SYSTEM RE4Cl4NGE ON 4 SINGLE PH45E B451S~00a)5X2iOO019~ H4LF WlNDlNG RE4CT4NCE OF TRMIYORMER42 XO0310036X TWW R 00172% 'I RESIST4NCE " .' ~144X0012~001720/1

1.25 X 10300

FIG. 1 .SO

i

I2900 4MPS ASYMHETRICbL 11.201

Short-circuit-current colculating procedure for single-phase three-wire 120/24Q.volt system.

76


reactance and resistance for line-to-neutral short circuit a t the secondary of the transformer. In both cases there is assumed to be no motor feedback.

TABLES AND CURVES FOR ESTIMATING SHORT-CIRCUIT CURRENTS To make short-circuit protective equipment application easier, particularly in circuits of 60 volts or less, many charts, tables, and curves have been prepared to eliminate the necessity for detailed calculations. Some of the more usef 1 ones are presented here. UNIT SUBSTATIONS

:o.

Standard low-voltage unit substations so widely used have standard transformer section impedance and voltage ratings. Hence, the secondary short-circuit currents available can be easily tabulated, as shown in Tables 1.5 and 1.6. The available short-circuit duty may be read directly from the table as a function of transformer kva, secondary voltage, and available primary short-circuit kva. Example of Use of Table 1.5. Assume a lonn-kva unit substation for 480-volt power service having an available SHORT CIRCUIT primary short-circuit capacity of 150,000 kva. "2 x, See 480-volt application table. Follow FIG. 1.51 0 n e - k diagram the vertical column under the 1000-kva suhshowing location of short circuit station rating down to the 150,000-kvaavailfor determinotion of short-circuit able primary three-phase short-circuit kva currents shown in Table 1.5. line in thetable. The availableshort-circuit current a t the 480-volt bus is indicated as 30,400 amp.

% ?]-'"

REDUCTION OF SHORT-CIRCUIT CURRENT DUE TO FEEDER IMPEDANCE

The unit substation application Tables 1.5 and 1.6 make it easy to determine the short-circuit current a t the main unit substation bus. By the use of the simple estimating curves the short-circuit, current at the end of the secondary feeders can he easily determined too. Henre these tables and the curves shown in Figs. 1.52 and 1.53 make it easy quickly to estimate the short-circuit current a t any point in a secondary system 600 volts and less fed by standard load-center unit substations. The curves are for 60-cycle operation. Figure 1.52 is for cable cirruits and Fig. 1.53 for bus feeders. The results are in terms of the three-phase average asymmetrical rm value during the first cycle corresponding with the basis of rating for low-


77

voltage air circuit breakers. The effect of circuit resistance both in increasing the impedanre and speeding the decay of the d-c component 'has been included. The range of operat,ing conditions encompassed is as follows: System operating voltage (nominal) : ZOSY/lZO volts, three phase, four wire; or 208 volts, three phase, three wire 480 volts, three phase, three wire; or 480Y/27' volts, three phase, four wire 600 volts, three phase, three wire Short-circuit-current magnitudes: 10,000 t o 100,000 amp Feeder-circuit construction : Three-conductor cable, No. 4 Awg to 500 MCM Busway, plug-in bus of representative designs in current ratings from 225 to 800 amp. Interlared loiv-reactance feeder bus (LVD) rated 2,000 amp, t,hrec phase (four bars per phase). y/

3

.

CABLE FLLOER LENCTM- FEET

FIG. 1.52

Chart for determining short-circuit current a t end of cable circuit consisting of three-conductor cable in conduit or interlocked-armor cable (60cycler).

NOTE:

12.2

12.3 12.4 12.5 12.6

11.9

10.0

417

150

15.9 16.5 16.7 16.9 17.1 17.2

625

500

Normol current, amp

300

750

1000

20.7 21.7 22.1 22.4 22.6 22.9

834

2080

2780

4170

I500

270

361

1 542

32.4 35.0 36.0 36.8 37.5 38.1

~

42.3 46.8 48.5 50.0 51.3 52.5

53.3 60.4 63.3 65.9 67.9 70.2

~

48.7 61.3 74.5 80.0 85.5 90.0

I

9.4 9.6 9.7 9.7 9.8 9.8

11.2 11.5 11.6 11.7 11.8 11.8 15.1 15,6 15.8 16.0 16.1 16.2

Total low-voltoge short-circuit Curlenh, thousands of amperes

1388

1 1 1 1 1 1

225

1 1 I I I 1 1

10.3 10.4 10.4 10.5 10.5

313

112.5

Substation kva rating

19.7 20.6 21.0 21.2 21.5 21.7

722

I

I

1

31.1 33.3 34.2 34.9 35.5 36.1

1203

rmal current, en

-

41.3 45.1 46.6 48.0 49.0 50. I

-

-

1804

-

Substotion kra rclting

52.2 58.3 60.8 63.0 64.8 66.7

71.2 82.5 87.5 92.0 95.9 100.0

%

impedance,

former

4.0

4.5

5.0

5.0

5.0 5.5

5.5

5.5

4.0

4.5

5.0

5.0

5.0

5.5

5.5

5.5

or different voltoge bare, multiply short-circuit current values in table by NOTE: 3. For differed wltmge hose. I tipiy 9 208 240 the ratio values in toble by the ralio naw voltoge n o r *olt.*e NOTE: 2. Motor short-circuit current contribution is 2.5 times the transformer normal NOTE: 4. Motor short-circuit current-contribution is 5.0 t i m n lhe t m n r current for 50% connected motors. former norm01 current for 100% connected moton.

~

Unlimited

50.000 100.000 150.000 250,000 500.000

.

Fi.C"it kw

short-

Available Primary threephase

SECONDARY RATING: 240 VOLTS, THREE PHASE

Available Short- circuit C u r r e n t f r o m ' t o n d a r d T h r e e - p h a s e Unit S u b s t o t i o n s

SECONDARY RATING: 2 0 8 Y / l 2 0 VOLTS, THREE PHASE

TABLE 1.5

a

$

2

5

2

'$

2

80


FIG. 1.53 Chart for determining short-circuit current (it end of feeder bur. designations refer to General Electric Company bus I60 cycles).

The type

Required Data. The basic data needed to enable the use of Figs. 1.52 and 1.53 are the following: 1. System operating voltage 2. Available short-circuit current at the source bus (average asymmetrical) 3. Length and construction of the feeder circuit 4. Connected motor load at the feeder terminal Procedure for Use of Figs. 1.52 and 1.53. The evaluation of feeder terminal short-circuit current involves only four simple steps (see Fig. 1.54): 1. Locate the magnitude of source-end short-circuit current on the proper left-hand operating voltage scale. 2. From this starting point move along to the right following along a curve or an interpolation between adjacent curves until the desired length of specific feeder construction (horizontal scales) is reached. 3. Project the latter point horizontally to the left and read the shortcircuit current contributed by the feeder on the same scale as used in 1.


81

4. Add the feeder terminal connected motor-current contribution (five

times the sum total of the motor full-load current). MODIFICATIONS FOR SPECIAL CONDITIONS

Parallel Circuit Feeders. A feeder circuit composed of two or more identical circuits in parallel can be readily treated by making a correction in the apparent length. The impedance presented by a feeder consisting of two circuits in parallel will be identical to that of a sing16 circuit of half the length; that of three circuits in parallel will be identical to that of a single circuit of one-third the length; etc. In the case of parallel circuit feeders, divide the true feeder length by the number of circuits in parallel and proceed on the basis of single-circuit data. .

I

Y

t

H k--

I .

1 i I ,/CP)OICI

I

I

l

I_

FEEDER L m m "

FIG. 1.54

I

Example rhowing how to use the charts of Fig. 1.52 and 1.53.

l

850

I

25 137.5

I I 1 50

75

100

150

1 1 I 1 333

250

200

500

Available primary Normal eurrenl. ornperes a1 240 volts

lhree-phose hoil-circull kro

104

I 1 I 1 I 1 156

208

313

417

625

833

1042

I

1388 12083

Tolo1 lox-vollage shw-circuil c ~ ~ r e nlhousandr l, of rms omperes for m e 120-volt winding short-circuited, lhe olhei opon-circuiled

- 25,000 50,000 100,000

I

6.5 9.6 12.6 6.7 10.0 13.3 6.8 10.2 13.6 6.8 10.2 13.7 6.9 10.3 13.8 6.9 10.3 13.9 6.9 10.4 14.0 - -

150.000 250,000

500,000

15.9 16.9 17.5 17.8 17.9 18.0 18.1

20.4 22.1 25.1

23.5 23.7 24.0

I 28.4 31.8 33.9 34.7

I

I

I

32.3

37.3

5.0

5.5

48.9 40.8 36.9 43.5 60.2 44.2 39.7 47.4 68.1 45.5 41.0 49.0 71.1 35.3 46.6 41.7 50.1 73.9 1 3 5 . 8 1 4 7 . 5 1 4 2 . 3 1 5 1 . 2 176.0

35.2

__

Unlimited Transformei full-winding impadançe: Per cent R . . Per cent 2..

... ....

1.4

1.4

1.2

1.2

1.21

3.0

3.0

3.0

3.5

3.5

1.21 3.5

1.21 3.5

1.0 5.5

- -

A short circuit invalving one of the secondrtry half windings (terminals Xi to X 2 ai terminals X, to X , ) , Fig. 1.51, allows eansiderahly more short-çireuit current to flow than a short circuit involving the full seeondary minding (terminals X i to X d . Consequently, the circuit-hreaker seleetions are based on the half-winding value of shortcircuit current. The eonditions on whieh the tables are hased are summsrizcd below: 1. A salid half-winding short cireuit at the tcrminals (scc Fig. 1.51). 2. Primary three-phase short-eircuit capacities vsrying from 25,000 kva to unlimited kva. For the worst case, the single-phasr short-cireuit capaeity is me-half the threephase primsry short-circuit capacity, and this value has bem used in thc celculations. This worst csse involves the assumption t h a t the primary of the transformer is connected line-to-line on the high-voltage system, not line-to-neutral. 3. The full-winding per cent impedance and per cent resistances m e given in Table 1.6. 4. The half-winding reactance was taken as 1.2 times the full-winding reactance, while the half-winding resistance was taken as 1.44 times the full-winding resistance, on full kva base. 5. The d-e offset multiplier for the first half eycle was taken as 1.25. 6. It is sssumed that the 120/240-volt units will supply lighting loads only, i.e., no motor feedbaek. 7. The only source of power connected to the secondary bus is one transformer of the capaeity indicated.


83

Feeders Consisting of Different Circuit Construction in Series. Make an independent evaluation of each common circuit construction starting at the source end. 1. Evaluate the short-circuit current a t the end of the first section of common feeder construction in the standard manner. 2. Using the answer derived from 1 as the source short-circuit-current value for section 2, proceed in the standard manner t o evaluate the shortcircuit current a t the end of the second section. 3. Using the answer derived from 2 as the source short-circuit-current value for the third section, proceed in the standard manner to evaluate the short-circuit current a t the end of the third section. Results obtained for sections beyond the first will be somewhat on the conservative side (higher than the true short-circuit-current value). This follows from the fact that the basic analysis assumes an X / R ratio of 12 a t the source end of the feeder. The true X / R ratio at the source terminals of any feeder section beyond the first will necessarily be less than 12 since no feeder construction exhibits an X / R ratio as high as 12. Interpolation for Intermediate Cable Conductor Sizes. Specific cable feeder length scales have been inscribed for conductor sizes of 500 MCM, 250 MCM, No. 2/0 Awg and No. 4 Awg. For intermediate valuesof cable size locate the horizontal scale points for the desired length of adjacent cable sizes which are charted, and interpolate between these values. For example, a No. 3/O-Awg conductor is about midway between a No. 2/0Awg and a 250-MCM. To evaluate the effect of a 100- f t run of No. 3/0Awg cable based on Fig. 1.52, locate the 100-ft point on the No. 2/0-scale and on the 250-MCM scale. A point midway between these two points will closely represent 100 ft of No. 3/O-Awg conductor. Three Single-conductor Cables in Conduit. Results obtained from the estimating curves without correction can be safely used to select protective interrupters. If desired, a closer approximation of the actual value can be obtained by increasing the apparent feeder length to account for the higher impedance of single-conductor feeder circuits. Conductor Sirs 500 M C M . . 250 M C M . . No. 2 / 0 A r g . . No. 4 Awg

Use an Appored Lenglh of

........ 130% of lhe acluol feeder Imglh ........ 120% of the o c h d feeder lenglh ..... 110% of lhe amal feeder lmglh ......... No correction

Both the 60-cycle resistance and reactance of a three-single-conductor cable feeder in conduit are greater than those of a three-conductor cable feeder in conduit or steel armor in the ratios reflected in the accompanying table:

84


Residence,

Conductor sire

I

..............

500MCM No. 2/0 A w g . . No.4Awg

I

......... .............

7%

Reactance.

Yo

I i 25

106 102

I

1-50 150

150

N OTE: Spaced open-wire circuits should be treated by conventional calculation procedures; a suitable one is given under Circuit AnalysisGeneral Case. Single-phase Circuits. Results obtained from the curves, Figs. 1.52 and 1.53, may be used with safety for the selection of protective interrupters. The true short-circuit-current value for a two-wire single-phase circuit operating at line-to-line voltage will be about 87 per cent of the t h r e e phase evaluation. Frequency. The curves, Figs. 1.52 and 1.53, are restricted t o 60-cycle operation. For operating frequencies other than 60 cycles, conventional calculations should be used, such as outlined under Circuit AnalysisGeneral Case. Note that feeder circuit resistance is not appreciably affected by frequency, while reactance varies directly with frequency.

UhIN SOURCE BUS 48O"OLTS ,.P"**E 6 0 C I C L E S SHORT CIRCUIT C W l R E N T i

4CCOOAYP

2 5 0 YCY 3IC INTERLOCKED ARMOR CABLES IN PARALLEL

FIG. 1.55 System diagram used as on example to illustrate the determination of short-circuit currenk a t the end of feeder circuits.


Example of Application-Fig. 1.55.

85

Short-circuit current at bus A ?

Source short-cirruit current = 40,000 amp Equivalent single cable feeder length = 1595 = 75 ft From curve Fig. 1.52 (4GO-volt short-circuit current scale; 250-MCM feeder Irngt,h scale) : Contribut,ion via feeder cable = 23,000 amp Motor contribution, bus A = 5 X 310 = 1,550 24,550 amp 315 Motor contribution, bus R = 5 X 03 = Short-circuit current bus A = 24,8G5 amp ~

~

Short-circuit current a t bus B? Source short-circuit current for section 2 = 24,550 amp (say 25,000) Feeder lengt,h = 75 f t From curve (4GO-volt short-circuit current scale) interpolate between the 7 5 f t point on ;To. 210 and KO.4 feeder length scales-Ko. 2 about onethird of the way from Xo. 4 to No. 2/0. Contribution via feeder cable = 11,000 amp Motor ront,ribution, bus R = 5 X G3 = 315 Short-circuit current bus R = 11,315 amp ~

CIRCUIT ANALYSIS- GENERAL CASE

The circuit, problem involved in resolving short-circuit-current magnitudes in low-voltage feeder systems is outlined in Fig. 1.56. I n general, low-voltage short-circuit current,s are expressed in terms of three-phase average asymmetrical rms amperes during the first cycle of currcnt flwv. Since main low-voltage source systems exhibit a n X / R rat,io of about, 10, it, is standard convention t o multiply the symmet,rical short,-rirruit, current, by 1.25 t o obtain the short-circuit current a t the main buses (this corresponds with a n X / R ratio of 12) (see Table 1.2). Therefore, at the main bus 1.25 E Short-circuit current = 1.25 X I symm = - X -

v5 z*

z , = 1.25 -& x

E short-circuit current

Considering the source system X / R ratio 1.25

z . = -4 x

E short-circuit current

(A +

=

12

jl) = R.

+jX.

86


+ + + + + _-

2, (obtained from reference tables) = R, j X , 2, (impedance to end of feeder) = R, R, j ( X . X , / R , ratio a t end of feeder = x. XI - x, R. Izi Rt

+XI)

M is the factor to account for d-c offset and is a direct function of the X , / R , ratio

......... ...................

XdRt ratio.. K

I 1 I I I 1 1;;s

I:*

s1:

!l

Ii6

2 1.02

I, is the local motor contribution, and the three-phase average assymmetrical rms value may be taken as five times the motor full-load rated amperes. Available short-circuit current at X = I, (three-phase avg assymmetrical rms) I,

+

61 2 -SOURCE SVSTEY IMPEDANCE Rg+ j X s OHMS/PHASE

'1

I

MAIN LOW-VOLTaGE BUS

4

FEE0ER:Zf:Rf

:'I

\J

LOAD

tjxf

OHMSIPHASE

IFROH TABLES)

VAIL4ELE SHORT ClRCUlT CURRENT DESIRED HERE IS'CURRENT CONTRIBUTION FROM SOURCE *"STEM ly*CURRENT CONTRlBUTlON FROM LCCAL YOTORS

FIG. 1.56 One-line diagram for rhortcircuit-current calculation ot the end of feeder circuits-genernl core.

07


TABLES SHOWING EFFECl OF CABLE LENGTH

Another useful way of showing the effect of length of cable in reducing short-circuit currents is given in the Tables 1.7 to 1.10. These show how much cable length is required to reduce the short-circuit current from one protective-device rating level to another for circuits GOO volts and less. Standard protective-device rating levels are: 100,000 amp rms asymmetrical 75,000 amp rms asymmetrical 50,000 amp rms asymmetrical 25,000 amp rms asymmetrical 15,000 amp rms asymmetrical 5000 amp rms asymmetrical The tables show how long a cable with a given con
I

FIG. 1.57

4

R CABLE

X

Equivalent circuit for determining cable lengths given in Tables 1.6 lo 1.9.

From the equivalent circuit per phase shown in Fig. 1.57 and using the nomenclature of Fig. 1.57, a general expression for the length of cable t o limit the short-circuit current can be derived. The equation is

L

E=

~ Z 2 1 , 2 / I ,2 R2 - X 221,

Where I J I , is large or R is small, the equation reduces to

88


In these equations R is the resistance, X is the reactance, and Z is the impedance of the cable per unit length. For any voltage not given, the length a t the new voltage is t o the length a t a given voltage as the new voltage is to the given voltage, i.e., the length is directly proportional to the voltage

where L, = length a t voltage E, Lo = length a t voltage Eo The lengths L for all conductor sizes from No. 1 Awg to 250 MCM were put in the table for comparative purposes. There are certain minimum sizes of conductors and hence certain minimum lengths of cable necessary a t various values of I, to keep the cable from being damaged before the protective circuit breaker operates. Referring to Chap. 3, i t will be noted, for example, that a t 50,000 amp (I,, Tables 1.7 to 1.10) the minimum size cable which a 50,000-amp interrupting rating low-voltage air circuit breaker will protect is No. 4/0 Awg. Hence, the only values in the right-hand column of Tables 1.7 to 1.10 that have any practical significance are the two at the bottom of the column. The values above that are of academic interest only. TABLE 1.7

Conductor size

.. .. ... ... N e . 4 A w g ... No. 2 A w g . . . No. I Awg , , . No. 110 Awg . No. 2 '0 A x g . No. 3/0 Awg . No. 4/0 Awg . 250MCM.. . . No. 14 A x g No. 12 Awg No. 10 A w g . . No. 8 Awg No. 6 Awg

Lirnitina Effect of Cable on Short-circuit Currents at 400 Volts.

Coble length 1, ft

26.9 42.6 67.5 106.5 165.0 254.0 384.5 468.0 564.0 664.0 775.0 890.0 962.0

7.3 11.4 17.9 28.0 42.6 63.7 91 .O 111.0 126.8 144.8 162.0 180.0 190.5

-

I , = avail le short-circu: I f = short-circuit current

I

::::

69.0 109.4 170.3 263.0 402.0 488.0 592.0 706.0 827.0 960.0 038.0

1:;

21.8 34.2 53.0 81.0 122.1 146.8 175.0 206.0 237.5 271.0 290.5

4.7 7.4 11.7 18.3 28.0 42.6 63.0 75.2 87.2 100.8 114.2 127.8 135.5

27.5 43.6 69.5 110.0

171.5 265.5 407.0 497.0 606.0 723.0 852.0 990.0 1072.0

9.1 14.4 22.8 36.0 56.1 86.3 131 .O 159.1 192.8 228.5 267.0 308.0 333.0

-

__

5.3 8.5 13.4 21.1 32.7 43.8 75.8 91 .4 110.7 128.8 149.1 171.2 184.1

2.4

-

3.8 5.9 9.3 14.2 21.4 31.8 38. I 44.0 50.9 55.4 64.7 69.0

-

urrent in kiloamperes a t source end of cable kiloamperes ior short circuit a t end of cable of length L

89


TABLE 1.8

limiting Effect of Cable on Short-circuit Currents at 480 Volts. Three Phase Three Single-conductor Cables in a Mmnetic Dud

I. I,

50 5

25 5 / 1 25 5 /

i

Conductor

sire

~

I1 =

50 2

5

availab short-c

TABLE 1.9

100 I5

100 5~ /

Cable length 1, fl

... 21.5 .. 34.1 ..... 54.1 ... 85.4 ... 132.5 .... 203.3 ... 308.0 ...... 374.0 .... 452.0 . 532.0 621.0 . 713.0 ..... 771 . O

=

~

-

No. I 4 Awg. , No. 12 Awg.. , No.lOAwg No. 8 A x g . . , No. 6 Axg.. . No. 4 Axg.. No. 2 Axg.. , No. I A x g No. 1/0 Awg No. 210 Awg .. . No. 3/0 Awg . ... No. 4/0 Awg , .. 250 M C M . .

I.

l 50 5

~

5.8 9.2 14.3 22.4 34.1 51.0 73.0 89.0 101.5 115.8 29.6 144.1 52.5

21.9 34.6 55.2 87.6 136.5 210.5 321.5 391.0 474.0 566.0 663.0 768.0 832.0

7.0 11.0 17.4 27.4 42.4 64.8 97.8 117.5 140.1 164.8 190.0 216.8 232.8

3.8 6.0 9.3 14.6 22.4 34.1 50.4 60.2 69.8 80.6 91.4 102.2 108.6

22.0 34.9 55.7 88.1

485.0 579.0 682.0 793.0 860.0

~

7.3 11.5 18.3 28.8 44.8 69.0 105.0 127.4 154.2 183.1 214.0 246.5 266.8

-

- short rcuit current kiloampcrcs i S O I I I ~ Pcnd of uit current in kiloarnperPs for short circuit at m i l of c

4.3 6.8 10.7 16.9 26.2 39.9 60.7 73.2 88.7 03.0 19.3 37.0 47.7

1.9 3.0 4.7 7.4 11.4 17.2 25.5 30.5 35.2 40.7 44.4 51.8 55.2

-

-

limiting Effect of Cable on Short-circuit Currents a t 240 Volts, Three Phase Three Single-conductor Cables in a Mmgnetic Duct

-

-

100

100 50

25 Conduclor *i*e

I

....__..10.8 ........ 17.1 ........ 27.0 ......... 42.7 ....__.._ 66.3 ......... 101.5 ....... 153.8 ........_ 187.0 ....... 226.0 ..... 266.0 .._....310.0 .. ... 356.0 ........_.385.0

No. 1 4 A x g No.12Axg No. 1 0 A w g No. 8 Awg No.6 A x g No. 4 A x g No. 2 A w g . . No. 1 Awg No. 110 A x g No. 2/0 A w g . . No. 3/0 A x g No. 4/0 Axg.. 250MCM

~

-

Cable length 1, ft

-

__

- -

2.9 4.6 7.2 11.2 17.1 25.5 36.5 44.5 50.8 57.9 64.8 72.1 76.3

11.0 17.3 27.6 43.8 68.3 105.3 160.8 195.5 237.0 283.0 331.5 384.0 416.0

5.5 8.7 13.7 21.2 32.4 48.9 58.8 70. I 82.4 85.0 108.4 116.4

3.5

1.9 3.0 4.7 7.3 11.2 17.1 25.2 30.1 34.9 40.3 45.7 51.1 54.3

__ 11.0 17.5 27.9 44. I 68.7 106.4 163.0 199.3 242.5 289.5 341 .o 396.5 430.0

3.7 5.8 9.2 14.4 22.4 34.5 52.5 63.7

77. I 91.6 107.0 123.3 !33.4

~

2.2 3.4 5.4 8.5 3.1 '0.0 0.4 6.6 4.4 I .5 9.7 8.5 3.9

__ I .o I .5 2.4 3.7 5.7 8.6 2.8

5.3 7.6 0.4 2.2 5.9 7.6

-- --- - - available short-circuit currcnt in kiloampercs at source end of eahle I , = short-circuit current in kiloamperrs for short circuit a t end of cahlc of length L I.

=


PO

TABLE 1.10

Limiting Effect of Cable an Short-circuit Currents ot 208 Volts, Three Phase Three Single-condudor Cables in a Magnetic Duct

25 5 Conductor sire

I I

100 25

lo:

100 50

lG

Cable length 1, h

--

..... ..... .... . ...... ...... ...... ...... ........ ...... .... .... .... .......

No. 14 Awg.. No. 12 Awg.. No. 10 Awg.. No. 8 Awg.. No. 6 Awg.. No. 4 Awg.. No. 2 Awg.. No. 1 Awg No. 110 Awg No. 2/0 Awg.. No. 310 Awg.. No. 4/0 Awg.. 250 MCM..

25 50

,

, , , ,

9.3 14.8 23.4 37.0 57.4 88.0 133.4 162.1 196.0 230.5 269.0 308.5 334.0

2.5 3.9 6.2 9.7 14.8 22.1 31.6 38.6 44.0 50.2 56.2 62.4 66.0

9.5 15.0 23.9 38.0 59.2 91.3 139.4 169.3 205.3 245.0 287.0 132.5 160.0

3.0 4.8 7.5 11.9 18.4 28.1 42.4 50.9 60.7 71.5 82.4 94.0 100.8

I.6 2.6 4.1 6.3 9.7 14.8 21 .9 26.1 10.3 14.9 19.6 i4.3 17. I

9.5 15.1 24.2 38.2 59.6 92.2 141.4 172.5 210.0 L51.0 295.5 143.5 172.0

3.1 5.0 7.9 12.5 19.5 29.9 45.5 55.3 67.0 79.4 92.7 106.8 115.5

-

-

1.9 2.9 4.6 7.3 11.3 17.3 26.3 31.7 38.4 44.7 51.8 59.5 64.0

0.8 1.3 2.0 3.2 4.9 7.5 11.0 13.2 15.3 17.7

19.2 22.4 23.9

-

REACTANCE AND RESISTANCE DATA FOR MACHINES AND CIRCUITS

When making short-circuit calculations, the most accurate reactance data available should always be used. In particular, reactauce of specific generators, larger motors, and transformers should be obtained from the manufacturer. Many short-circuit studies must he made without such specific data available, as for a proposed plant or in many older plants where the time and work required to obtain such data from the manufacturers make it impractical to do so. Since a great many short-circuit calculations fall in this category, it is desirable to use approximate reactance data. Such approximate data as are commonly used are given in Tables 1.11 to 1.31. The most applicable reactances should be selected from these tables.

91


APPROXIMATE MACHINE REACTANCESdO CYCLES

Large Induction Motors. The approximate short-circuit reactance of an induction motor (or induction generator) in per cent on its own kva base may be taken as

Per cent X : =

100 times normal stalled rotor current*

The reactance of such a machine will generally be approximately as given in Table 1.11 (in per cent on own kva base). TABLE 1.1 1 Range

M-t

15-25

TABLE 1.12

Common 20

Approximate Reactances of 60-cycle Synchronous Machines Per Cent Vdues on Moshino Kva Roting I

Salient-polo cpnerotors (without amortirre,url: 12 poles 0, leu.

.....................

..................... ..................... ...................

14polnoimne Salient-pole ganomton~(with amortiiseur): 12 pole* or In.. 14 poles or more.. Synchmnoui condenrers.

.................

Synchronwi converterd 600 v d h dc.. 250 d t s dc..

....................... .......................

I

15-35 25-45

25 35

10-25 10-35 18-35

18 14 27

17-22 28-38

10 33

10-20 15-25 25-45

15 20 30

Synchronous motor^'

.............................. ........................... .......................

6 pole 8-14 pole I 6 pole or more.

15-30 20-40 25-60

23 30 40

a Nearly all salient-pole generators built by General Electric Company since 1935 have amortisseur windings. Add transformer reactance: For compound-wound converters add 12 per cent. For shunt-wound converters add 7 per cent. These data are useful for estimating reactances of individual large motors of several hundred or several thoumnd horsepower.

* With rated voltage and frequency applied.

92


TABLE 1.13

Approximate Reactance of General Electric Company Turbine Generators, 625 to 18,750 Kva

K w ..ling 0.8 power facer

w rating

__

Volt.g* rating

1200

3600 rpm

rpm ~

625

500

240 480 600 2,400 4,160 6,900

781

625

240 480 600 2,400 4.160 6,900

875

700

937

750

240 480 600 2,400 4,160 6.900

1,250

1,000

240 480

240-4.1 60

600

2,400 4,160 6,900 1,562

1,250

240 480 600 2,400 4,160 6.900

1,857

1,500

240 480 600 2,400 4.1 60 6,900 13.800

14.5 14.5 14.5 14.5 14.0

.... .... .... .... .... .... ....

8.0 9.0 9.0 9.0 9.5 6.5 8.5 8.5 7.5 9.0 9.5 5.5

14.0

.... .... .... .... .... .... .... 15.5 15.5 15.5 15.5

.... .... .... .... .... .... .... .... 16.0 16.0 16.0 16.0 16.0 16.0

11.5 7.5 7.0 7.5 7.0 9.0 10.0 10.0 10.0 9.0 10.0

8.5 9.5 8.5 8.5 9.5 9.0 8.0 10.5 9.0 8.5 8.5 9.5 7.5

-


rABLE 1.13 Approximate Reactance of General Electric Company Turhine Generators, 625 to 18,750 Kva. (Continued) Kvn rating

0.8

X&‘

:w ‘atin(

Vdtoge ‘ding

2,500

2,000

480 600 2,400 4,160 6,900 l1.500-13.800

P.5 10.5 10.0 10.0 10.0 8.0

3.125

2,500

480 600 2,400 2.400/4.160 6.900 I 1,500 13,800

9.0 8.5 9.5 8.5 9.0

10.5

power facto,

3600 rpm

10.0

3750

3,000

480 600 2,400 2,400/4,160 6.900 11,500 13,800

9.0 10.5 9.5 10.0 9.5 10.5 10.5

4,375

3,500

480 600 2,400 2,400/4,160 6,900 11,500 13,800

8.0 9.0 9.0 9.0 9.0 10.0 10.0

5.000

4,000

480 600 2.400 2,400/4.160 6,900 1 1,500 13,800

10.5 7.5 7.0 8.5 9.0 9.5 9.5

6,250

5,000

600 2,400 2.400/4,160 6,900 I 1,500 13,800

12.0 7.5 7.5 8.0 8.5 8.5

93

94

SHORT.CIRCUIT~CURRENTCALCULATING PROCEDURES

TABLE 1.13 Approximate Reactance of General Electric Company Turbine Generators. 625 to 18.750 Kva. (Continued)

-

Kvo rotinp

0.8 power f.Ct0.

w rating

Voltage rating

X;

3600 rpm

7,500

6,000

2,400 2,400/4,160 6.900 1 1,500 13.800

9.0 10.5 9.0 9.5 9.5

9.375

7.500

2,400 2.400/4,160 6,900 11.500 13,800

9.0 10.5 9.0 9.5 9.5

12,500

10,000

2,400/4,160 6,900 11.500 13.800

10.0 8.0 9.0 8.0

i 8750

15,000

-

6,900 11,500 13.800

11.0' 11.0' 11.0'

* 0.5 psig hydrogen pressure.

TABLE 1.14

Reactances Based on Kvo of Connected Motors 1

I

Tranrient

Motor rotings ond connections

Itern

-I

reactance

I

Xi. per rent

per cent

1 28' 21 34' 27*

600 "0th or less-induction 600 volts or lewynchronous l i t e m 1 end 2 indude motor leads1 600 volh or l e u i n d u c t i o n 600 volts or les-ynchronour litems 3 and 4 indude motor leads and step-down bansformen1 Motors above 600 volt-induction Motors above 600 volt-ynchronwr Motors above 600 volh-indudon Motors o b w e 600 voltriynchromur litems 7 and 8 include step-

20 15 26 21

down transformers1

I

* Based on AIEE Standard No. 20.

1

29 35

25 31

-

95


Assumed Motor Reactances- Group of Small Motors. In many short-circuit studies, the number and size of motors, either induction or synchronous, are not known precisely. However, the short-circuit contribution from these motors must be estimated. In such cases Table 1.14 is used to account for a large number of small induction and synchronous

motors. The proportions of synchronous and induction motors (at all voltages) should be known for short-circuit investigations. Some typical ratios of total plant motor load which are usable in preliminary work are given in Table 1.15. The kva of the motors which are energized at one time varies also with the type of plant and should be investigated for the more complete studies. Approximate relations of energized to installed motors and of energized motors to source (transformer and/or generator) capacity are given in Table 1.16.

TABLE 1.15

Rotio of Induction and Synchronous Motors Motor mio, par cent Plant Induction

Sydrnnom

40 85 50 67 50

60 15 50 33 50

.............................. .............

Cement Machine shops ond IexHIe.. Rubber and rolling mills.. Paper (excluding grinder mobs). Commercial ond offiso..

............... ........ ................

TABLE 1.16

Rotio of Energized and Instolled Motors

PI.3

............. ....... ............................... .......................

Continuous PIOLOS (cement. textile). Semicontinuous (paper, reflnerier, rubberl.. Rolling mills.. Intermittent operotiom.

Installed motor k w to source kva (excluding SPW"). por cent

Energized motor kva to insbled motor k w , per cent

100 90 80 75

110

110 1 67

215 400

96


APPROXIMATE IMPEDANCE O F TRANSFORMERS

The impedance of transformers ronsidered in a short-circuit study should be obtained from the name platc or the manufacturer. However, where such data cannot be obtained, the values given in Tables 1.17 to 1.19 may be used in short-rircuit studies for estimating the short-circuit currents in the usual case. I n the usual short-circuit study, the transformer reactance and impedance may be assumed t o be the same without causing significant error for transformer banks above 300 kva. This assumption is useful because transformer name-plate data include impedance and not reactance. TABLE 1.17

Approximate Resistance, Reactance, and Impedance of Single-phase Distribution Transformers

High voltage: 2400/416OY volts and 2400/4800/8320Y "Olt. ow voltage; 120/240,240/480.600 volts60 yclos

Per cent R

High voltage; 7200/12,47OY v01b ow voltage: 120/240. 240/480,600 volts-

60 cycles

Per cent

Per cent

Per cent

X

z

R

Per. cent X

Per cent

z ~-

1.7

I .5

2.3

2.2

1.7

2.8

1.5

1.7

2.3

I.6

I .6

2.3

I. 3

2.2

2.6

1.3

2.0

2.4

75 100

I .2

2.3

2.6

1.2

3.5

3.7

1 67

1.1

3.8

4.0

1.0

3.6

3.7

250 333 500

I.o

4.7

4.8

1 .o

5.1

5.2

3 5 10 15 25 37> 50

APPROXIMATE REACTANCE A N D RESISTANCE O F CABLES

The reactance of a cable circuit is, generally speaking, a function of the spacing between conductor centers and the conductor diameter. Knowing the conductor spacing and diameter, the reactance of three-conductor


TABLE 1.18

97

Approximate Impedance of 60-cycle Power Transformers IAbovs 500 Kva) Impedance at kvo bole equal lo 55 C rating of largest Capacity winding for

Inwlotion doss, k r

Low voltage

High voltage

..... 15 or lower ........... 15 or lower 3 4 . 5 . . ......... I5 or lower 4 6 . 0 . . ......... I 5 or lower 6 9 . 0 . . . ........ I 5 o r lower 9 2 . 0 . . ......... I 5 or lower I 15.0. .......... I 5 or lower 138.0 ........... 15orlower I 5 or lower. 25..

Self-cooled or

Forced-oil

woter-cooled rating, per cent

cooled rating, per Cenl

5% 5% 6 6%

7 7% 8 856

For high-voltagr insulation elassrs intermediatr of those given, use the imppdancr of thc next higher listpd insulation class. For transforrncrs with a load-ratio control add 0.5 prr ccnt to the vaIu?s IistFd abovc crcrpt in those eases in which a IOWPY impedaner has heen sprrifirtl. Thc p ~ cr m t resistance on the hase given above rangrs from 1.0 down to 0.06. TABLE 1.19

Approximate Reactance o f Load-center-type Transformers, 60 Cycles (Three-phase) 15-kv Maximum Primary Voltage 600-volt Maximum Secondarr Voltaoe Per cent Reactance on O w n K r o Bore* 3.0 5.0 5.5

-

Kro Range

Il256-l50 225-500 750-2000

* Per cent resistance on own kva base is apptoiirnatcly 1.5 p ~ ccnt r for 150 kva snd b&w and varips from approxirnatdy 1 down to 0.8 p ~ cr m t on ratings above 150 kva

cables in nonmagnetir ducts and without maglietic binders can be determined by the formula

X

=

(

0.023 log, D 2s

+ K)

X = reactance, ohms per 1000 ft at 60 ryrlrs; S = spacing of couductors (center t o center), in. D = diameter of ronductors, in.; K = a rocffirient dependent upon ratio of iriside diameter of a ronductor to outside diameter of condurtors. For standard strand construction K = 0.25

98


This formula does not take into account any increase of reactance due t o the spiraling of the strands. Such increase is usually negligible in three-conductor cables and in large single-conductor cables, but it may amount to 1 to 2 per cent in small single-conductor cables. The effect of irregular spacing of the conductors and of magnetic binder causes an increase of reactance of single-conductor cables, compared with otherwise equivalent three-conductor cables. Cable insulation thickness varies with different types of insulation for a cable of a given voltage class. The approximate reactances of cables taking into account these variables are shown in Tables 1.20 t o 1.22. TABLE 1.20

Approximate Resistance, Reactance, and Impedance of 600-volt Cables in Magnetic Ducts per 100 Ft

Three single-conductor cables per dud. ohms per 100 fi

Three-conductor cable including interlocked armor cablo, ohms per 100 fi

Coble size

No. 14 Awg. No. 12 Awg No. 10 Awg.

.

.. .. .. N . 2 Awg . .

No. 8 Awg No. 6 Awg No. 4 Awg O

No. I Awg.. No.l/OAwg.

R'

X

R'

X

Z

0.3135 0.1972 0.1240

0.00765 0.0071 0 0.00687

0.3135 0.1972 0.1240

0.3135 0.1972 0.1240

0.00468 0.00456 0.00448

0.3 1352 0.19720 0.12410

0.0779 0.0498 0.0318

0.00638 0.00598 0.00551

0.0782 0.0500 0.0322

0.0779 0.0493 0.0312

0.00427 0.00391 0.00362

0.07460 0.04899 0.03140

0.0203 0.0163 0.0131

0.00513 0.00500 0.00495

0.0209 0.0171 0.0140

0.0197 0.0125

0.00344 0.00342 0.00340

0.02000 0.01606 0.01296

0.0157

No.Z/OAwg. No. 3/0 Awg No.d/OAwg

.

0.0106 0.00860 0.00700

0.00490 0.00486 0,00482

0.0117 0.00986 0.00850

0.0100 0.00800 0.00640

0.00336 0.00333 0.00327

0.01 054 0.00866 0.00721

25OMCM... 300 M C M . . 350 M C M . .

. .

0.00608 0.00520 0.00461

0.00480 0.00474 0.00469

0,00778 0.00704 0.00658

0.00547 0.00460 0.00400

0.00322 0.00316 0.0031 0

0.00632 0.00557

0.00510

.

0.00419 0.00359 0.00280

0.00462 0.00450 0.00438

0.00625 0.00575 0.00520

0.00354 0.00292 0.00208

0.00304 0.00295 0.00284

0.00469 0.00412 0.00346

400MCM... 500MCM... 750 MCM..

* Based on 75 C.

--

--


TABLE 1.21

99

Approrimote Resistance. Reactance, and Impedance of 5000-volt Cables in Magnetic Ducts per 100 Ft Three.conductor cable including interlocked armor cable, ohms per 100 ft

Three dngle.condudor cobler per duct, ohms per 100 ft Cable size

___.

z

R*

X

No. 14 Awg No. 10 Awwg. No. 8 Awg

0.3135 0.1240 0.0779

0.00969 O.OO8M 0.00788

No. 6 A w g . . No.4Arg.. No. 2 A x g . .

0.0498 0.0318

No. 1 Awg..

z

R'

X

0.3135 0.1240 0.0781

0.3135 0.1240 0.0779

0.006664 0.005745 0.005308

0.3291 0.1241 0.07808

0.0203

0.00748 0.00681 0.00623

0.0503 0.0325 0.0212

0.0493 0.0312 0.0197

0.004941 0.004619 0.004366

0.04944 0.03154 0.02017

0.0163 0.0131 0.0106

0.00588 0.00567 0.00545

0.0173 0.0143 0.0119

0.0157 0.0125 0.0100

0.003964 0.003792 0.003677

0.01619 0.01304 0.01061

No.3/0 A x @ . No.4/0 Awg. 250 MCM..

0.00860 0.00700 0.00609

0.00535 0.00529 0.00525

0.0101 0.00877 0.00802

0.00800 0.00640 0.00547

0.00363 1 0.003585 0.003562

0.008785 0.007535 0.006527

300 MCM..

0.00520 0.00461 0.00419

0.00519 0.00514 0.00506

0.00735 0.00690 0.00657

0.00460 0.00354

0.00351 8 0.003477 0.003436

0.005791 0.005299 0.004923

0.00359 0.00280

0.00495 0.00474

0.00611 0.00551

0.00292 0.00208

0.003344 0.003088

0.004439 0.003723

. ..

No. 110 Awg

.

No. 2/0 Awg

. .

350MCM... 400 MCM.. 500 MCM.. 750 MCM.

. .

..

Based (1

75

c.

TABLE 1.22

0.00400

--

---

Correction Factors for Nonmagnetic Ducts Single-condudor a b l e ,

Fo&r for conesting

reactancn, dl rizor

No. l 4 t o No. 8 A x g

of cable

0.8

I

1.0

I I

Focton for correcting redrlmces

No.610 No. 0 Awg

0.96

1

I

No.00to 250 MCM

0.93

1

30010 500 MCM

0.83

11

750 MCM

0.72

100


TABLE 1.22

Correction Factors for Nonmagnetic Ducts.

(Continued)

Three-conductor Cables Determine correct Z from corrected d ~ eofi X and R. N o ~orreclioni s required for interlocked armor.

Factors for correcting resistances Factor for correcting reoctancer. di sizes of coble

No. 14 to No. 00 Awg

0.87

TABLE 1.23

No. 0000 Awg to 750 MCM

1

1.0

0.98

Per Cent Reactance of Typical Three-phase Cable Circuits Per Cent Reactance of 1000 Circuit Feet on o 1000-kva nose

1 ! 1 230

System roltoge

I

460

575

~

2,400

j

4,160

6.900

~

I

1

13,800

Cqble sire. No. 4 to 1 Awg

Three single-condudor cables in iron conduit.. 98.3 Three-conductor coble in iron conduit or interlocked armored I 71.8 cable Three-conductor cable in nonmag58.5 netic duct..

..............

..................... ................

18

I

...............

.................... 68 ................ 54.8

0.358

~l

I

0.669

0.222

0.11

0.0276

9.4

0.581

0.194

0.0955

0.024

I f 0 to No. 410 Awg

I 92.5

1.075

11.5

14.7

Cable size, No.

Three single-conductor coblei in iron conduit. Three-conductor cable in iron conduit or interlocked armored cable. Three-conductor cable in nonmognetic duct..

15.74

24.6

I

1 !

23.2

14.85

0.955

0.318

17.1

10.9

0.6

0.2

0.0943

0.0237

13.72

8.8

0.52

0.173

0.0818

0.0205

Cable Sire, 250 to 750 MCM ~

Three ringlo-conductor cables in

..............

ironconduit. Three-conductor cable in iron conduit or interlocked armored coble Three-conductor cobie in nonmognetic duct..

..................... ................

85

11.3

13.63

61.4

15.4

9.85

0.538

0.179

0.0796

51

12.8

8.19

0.477

0.159

0.07

1 y

2.

0.02 0.0176

__

SHORT.CIRCUIT-CURRENT CALCULATING PROCEDURES

101

Where more precise data are not available, the values given in Tables 1.20 t o 1.23 may be used in short-circuit-current calculations without significant error. APPROXIMATE REACTANCE

OF

BUS B A R S 6 0 CYCLES

Unlike cable circuits the resistance of bus-bar circuits is so low compared with the reactance that the resistances of bus bars may he neglected in all a-c short-circuit calculations without significant error. There haye been many papers written on the subject of bus-bar reactance calculations, and a complete bibliography is included in t,hc 1945 A I E E Transactions, Vol. 64, page 385, The Design of Bus-bar Indust,rial Distribution System: An Epitomization of Available Data, by T . .J. Higgins. For practical short-circuit calculations, the reactance of bus bars may be taken from Figs. 1.58 to 1.G2 or Tables 1.24 and 1.25. TABLE 1.24

Reactance of Typical Three-phase Low-voltage Copper Busway Circuits Per cent reoctonce of 1000 circuit feet on

D

1000-kva base

System voltage Butuoy rating, amp

240

1 I 480

600

98.8 62.4

24.7 15.6

15.8 10.0

45.2 17.2 10.8 7.6

11.4 4.3 2.7 1.9

7.3 2.7 1.7 1.2

Plvg-in type:

............. ........... ............. ........... .......... .................

Upto600 60110 1000 Lox-impedance type: Upto600 60110 1000 135010 1600 2000

Although not gcnerally used in short-circuit calculations the resistance of typical copper busway circuits is giveu in Table 1.25. TABLE 1.25

Resistance of Typical Copper Busway Circuits Current Capacity of Bvsroy, Amp 250 400 600 800 1000

I350 1600 2000

Resistance. Ohms p e l 1000 Ft 0.114 0.033 0.023 0.016 0.012 0.0096 0.0073

0.0055

102

FIG. 1.58

FIG. 1.59


Chart showing reactance

spacing of rectangular bus bars

VI.


VI.

160 cycler).

spacing of rectangular bur bars

I60 cycler).


FIG. 1.60


YI.

spacing of rectangular bus bars

(60cycled.

104


FIG. 1.62

Chart showing reactance VI. spacing of channel bus bars I60 cycler).

REACTANCE A N D RESISTANCE OF OVERHEAD LINES

To assist in obtaining the conductor spacings, two typical crossarm arrangements are shown in Fig. 1.63. The arrangements used in practice will vary from system t o system, but hecause of space limitations only these two are shown. For ordinary single-phase circuits, the equivalent spacing is the distance between conduct,ors. For ordinary t,hree-phase circuits, the equivalent spacing is exprcssed by the formula + A X t( X C where A , B , and C are the distances, center t o center, of the conductors as follows:

~ - B - I -A- I

-

~

The resistance of overhead lines may not always he neglected without significant error. In general, long runs of overhead lines (several miles) at 2.4 t o 13.8 k v with small conductors 250 MCM or less have significant resistance compared to reactance; therefore resistance should be considered in short-circuit calculations for short circuits a t the ends of such long overhead lines. Resistance should he considered in all low-voltage (600 volts or less) overhead lines. Reactances and resistances may be taken from Table 1.26 for small


105

spacings (up t o 8 ft) and from Figs. 1.64 and 1.65 for spacings up t o 20 ft. Under usual application conditions, transmissiodine reactance varies over quite a narrow range. Table 1.27 includes the usual variations as well as “average ohms per mile” which are normally satisfactory for quick estimating work. Very large conductors, used to carry unusually large amounts of power for short distances, have abnormally low reactance so that this tahlr is not applicable.

L---

67“

4 - P I N CROSSARM AND SPOOL- TYPE SECONDARY RACK

6 - P I N CROSSA-M

Spocing of pins on four- and six-pin crossarms for vie in calculating line reoctance on 2400/4160-Y or 48OO-volt circuits.

FIG, 1.63

106


FIG. 1.64 Chart showing poritive-phaie-sequence reactonce of transmission lines using hard-drown stranded copper conducton (60cycle).

SHORT-CIRCUIT-CURRENT CALCUUTING PROCEDURfS

107

POSITIVE SEQUENCE 60 CYCLE

EQUIVALENTA SPACING OF CONDUCTORS IN FEET

FIG. 1.65 Chart ahowing poritive-pha**.requence ACSR conductors (60cycler).

reactance of trammission liner uting


I09

SHORT-CIRCUIT-CURRENT CALCUUTING PROCEDURES

TABLE 1.27 Approximate Equivalent Delta Spacing and Average Reactance r M i l e of Three-phase 60-cycle Transmission Lines Normal tronrmiirion Approximate equi*o1ent dell. spocing of

Line in.ul.tion class, kv

Reoctonce,

conductors. ft

5

2.5

15

3.5

23

4

34.5

4.5

46

5.5

size

ohms per mile lslrmded ~ o p p e i l

0.61 0.74

40. 4 1 0 Awg No. 6 Awg

~-

__

0.64 0.75 0.65 0.77 0.64 0.75 0.65 0.77

250 MCM No. 4 Awg 250 MCM No. 4 Awg 400 MCM No. 2 Awg 500 MCM No. 1 Awg

0.70

14 8

69 115 138 161 220 287

16 20 20 40

TABLE 1.28

System Yoiloge.. ,

0.65

..., .

411 UIUDI

sizes

0.70-0.80 both copper and oluminum

Reactance of Typical Three-phase Medium- and Low-voltage Distribution Circuits* ~

230

~

460

575

. . .... .... . .....

Equivalent delta spacing, ~ 6 ~ 1 2 ~ 6 / 11 28 1 ~1 8 1 d / I 2 / 1 8 1 3 0 1 I".

0.75

Wire sire

Per cent reoctance of I000 c i r ~ u i feet l on

__...

N0.4loN0.1Awg 180208~22345.052.156.028.833.335.8 No. I i O A w g t a 2 5 0 M C M 155 i8520238.846.550.724.8 29.732.4 300 lo 750 MCM.. 134 1631180 33.6 40.8 45.0 21.5 26.1 28.8

.....

I

D

301 3 6 1

42

1000-kvo bore

2.19 0.7620.286 0.073 2.06 0.6880.258 0.067 1.87 0.625 0.235 0.061


110

Volts (line-to-lid

*

115

230 460 575 2,300 4,160 6,900 13.800 22,000 33,000

Equivalent dslto spacing, Kv line- Equirdent delta rpmcing, in. to.li". ft

12 12 I8 I8 30 30 36 42 48 54

44 66 I10

112 I54 220

5.5 8.0 14.0 16.5 20.0 29.0

APPROXIMATE REACTANCE O F LOW-VOLTAGE ClRCUll BREAKERS A N D DISCONNECTING SWITCHES

In some low-voltage circuit calculations, the reactance of such switching equipment may be significant. The reactance of circuit breakers varies greatly, depending upon the rating and design. For approximation, however, the reactance in ohms of a circuit breaker may be taken as

0.2 continuous rating of circuit breaker in amperes The reactance of lever switches and disconnecting switches for low-volt,age circuits (600 volts and below) is of the order of magnitude ranging from 0.000050 to 0.000080ohm per pole at fiO rycles,for sizes ranging from 4000 to 400 amp, respectively, depending on the ampere rating, design, and phase spacing of the switches. APPROXIMATE REACTANCE O F CURRENT TRANSFORMERS

These data are useful o~ilyfor calculation of short-circuit currents ill circuits rat,ed 600 volts and below. The reactance of current transformers depends 011 their current rating and design arid varies over a wide range. Therefore, a sirigle value of reactance applicable to a variety of current transformers cannot be given. Current Transformers with Primary Circuits of t h e Wound Type. Approximate data on renctarice a t 60 cycles for current transformers of


Ill

type W, covering current ratings from 100 t o 800 amp based on tests at short-circuit currents, are given in Table 1.30. The values in Table 1.30 apply t o t,ransformers with a serondary burden of I volt-amp or less at 5 amp or a t normal i:urrent. For higher burdens, the impedance referred t o the primary side will be somewhat increased, but the increase is far less than that occurring a t normal currents, berause of the reduced mutual inductance between primary and secondary windings. The reactance values based on low burden are conservative fur calculations of maximum short-circuit current. TABLE 1.30

Over-all Reactance of Type W Current Transformers, Referred to Primary Winding

Approximate Values at Short-circuit Cvrrenh with D-C Component, Rms Symmetrical Component Ronging from 15,000 to 55.000 Amp Current Rating of Primary Winding, Amp

100

I50 200 250

300 400 500

600 800

Reactance (11

60 Cycles, Ohms 0.0035 0.0017

0.0010 0.00066 0.00050

0.00032 0.00022 0.00019 0.00012

These values are also representative of t,he order of magnitude of the reactance for current transformers of the following types, rated a t 5000 volts: JW1, JW4, JW6, JW14, WC12, WFI, WF6, and WF12. Reactances for other designs of current transformers of the wound primary type may be estimated by applying the folloming approximate factors t o the values of Table 1.30. Type of Current

Transformer

Foctor to Be Applied to Reactance Vduer in Table 1.30

KF85-7,500 volt

1.8

JSI-15,000 volt

0.4

Current Transformers Having a Bar-type Primary Conductor. For bar-type current transformers with currerit rat,ings from 1000 t o 4000 amp, such as t,ypes bS2-GO0 volts, WC15-5000 volts, KC60 7500 volts, the react,arice has an approximate order of magnit,ude of 0.000070 ohm a t currents within the range from 10,000 t o over 80,000 rms symmct,rical amperes, with or ait,hout d-c component,. T h e reactauce depends on the spacing bctweeu phases, since a COILsiderable amount of air flux links the primary bar conductor. The value given is t,hat for !&in. phase spacirig wit,h the t,ransformers side by side, reprcserit,ing an average value for the three phascs for t,hree-phase short circuits. Strictly speaking, the reactance in the three phascs will


I12

he unequal i n a side-by-side assembly of current transformers, but for short-circuit-current calculations an average value can ordinarily be used without serious error. To say that the reactance for bar-type transformers is equal to the air reactance of the primary conductor, considering its length, size, and shape, and the spacing between phases, is a fair approximation.

APPROXIMATE REACTANCE O F A - C REACTORS AND FEEDER REGULATORS

The reactance is proportional t o the rating. The voltage drop through the reactor at rated current and frequency divided hy the line-to-neutral voltage of the circuit gives the per-unit reactance on the current rating of the reactor. (This will also he the per-nnit reactance on the kva rating of the circuit if the rated reactor current is the same as the rated current of the circuit.) The reactanre of a given step regulator is modified by the position of the tap changer and becomes a maximum a t maximum voltage boost. It is minimum at neutral position, while at maximum buck, the impedance is higher than at neutral. TABLE 1.31

Short-circuit ImDedance of Feeder Reaulators of

I Type

No. of

phoier

core,

!-

......... I n d w . . ......... Indue..

No. of

I or 3 Ior3

Circ. "Oil,

Per cent

Kva of

.. ..

_

rd

Avg

Max

_ -~~

2400

I?

to

I0

4800

96

_____

C ~ C .k

r.g"l.tol

Min

~

,O(lCt(l"L*

Ion base of

0.65

0.85

_

1.00

_

Amp rO~Y1.101

............

Step Step. Step.

........... ...........

1 3

1 I

2400

3

3

13,800

lo

Allrolings to160omp Over 160arnp

~ _ _ _ _ O.O+

O.O+ 0.15

.... .... ....

0.6 0.7 1.0


113

REFERENCES 1. A I E C Committee Rrport. Simplified Calculation of Fault Currents, E k e . E’ng., Octobrr, 1912.pp. 509-511. 2. A I I * X Committee Ilrport. Simplifird Calmlation ai Fault Currmts. Trans. AlEE’. 1942, Vol. G I . pp. 113:3-11:35. 3. Revision Made to AIICI: Report, Simplified Calculation of Fault Currents, Efec. Emf,.. February, 194d. p. 65. 4. Darling. A. G., 4-C Short Cirrriit Caleiilating Procdure for Lon--roltage Systems, ‘I‘mns. A I E E . l!)41, Vol. GO, pp. 1121-1136. 5. Srhurig, 0. It.. Fault Voltngr Drop and ImpPdanre a t Short-circuit Ciirrmts in Low Voltngr Circuits. Trans. A I E E , 1941, Vol. 60, pp. 479-486. 6. AIEE Committw Rrport. Simplifird Calculation of Fault Currents, Trans. A I E E , 1948, Vol, 67, p. 1433.

Chapter 2

by R, H. Kaufmann

Symmetrical Components as Applied

to Short-circuit-current Calculation on Three-phase Systems The unhalanred circuit problems eucountered in short-circuit analysis can be resolved by using symmetrical-component analysis. This analysis technique is used extensively by power-system invest,igators and authors. Developed in this chapter are concepts and procedures for the application of symmetrical-romponent aualysis t,o the determination of short-circuit currents. While this procedure is built up from base fundamentals, it is aimed expressly a t the solutiori of electricalsystcm short-circuit problems. For other possihlc applications of symmetrical-component analysis such as the determination of unbalanced currents in certain circuits or machines, it is suggested that reference be made t o a more elaborate texthook* which explores the full field of application more completely.

THE USE

OF

COMPONENTS

The separation of an electrical vector quantity into components t o simplify computation procedure is familiar t o all. It has been common practire t o consider an alternating voltage or alternating current to he composed of two components a t right angles t,o each other. It should he evident that the process is not limited to two quantities, nor is it necessary t h a t the components be 90" apart. *Edith Clark?. "Cirruit Analysis of A. C. POIVPISystrrns." vol. 1, John W i k y & Sons, Inc., NPWYork, 1943. I14

SYMMETRICAL COMPONENTS FOR THREE-PHASE SYSTEMS

115

For example, take the expression

E = IZ

It is entirely valid t o express this as E

=

(I,

+ I,)Z = IIZ + I,%

provided that

I, or as

E

=

(I1

+I,

+ +IJZ 1 2

provided that

I,

=

=

I

I1Z

+ I2Z + IaZ

+ I2 + I3 = I

Thus there is 110 mystery about the use of components. It is applicable so long as the equations are linear (as they will be in electrical-cirruit work). E = IZ I = EY I, = ZJ, etc. SYMMETRICAL COMPONENTS

If the Z per phase as illustrated in Fig. 2.1 could he represented as a firm fixed value, the circuit analysis would be-simple. Since the conductors of the three phases are magnetiz I A -b cally coupled, the voltage drop in the A JWVL Nl phase depends not only on the current in the A phase but on the current in the other two phases as well. Consider the induction-motor impedance diagram of Fig. 2.2. Assume the FIG. 2.1 A simple i y m m e t r i c d rotor t o be turning a t normal speed in ryrtem, the direction produced by an impressed voltage of sequence ARC. What I Z drop will be produced in the -1phase because of a current I , alone? That is a tough one; although there are some relationships of which we are sure. Under the conditions of balanced currents of sequence ARC there will be balanred terminal voltages of sequence ABC. With normal rated voltage and light load the current will he of the order of one-fourth or one-third rated value. Under this condition all three phases appear t o have identical impedances of 1/0.25 or 1/0.333 or three or four per-unit (300 or 400 per cent). On the other hand, had the impressed voltage been applied with opposite sequence (.4CB), i t is evident that this would he equivalent t o


116

plugging. There mould he a balanced set of currents, but this time the application of rated voltage would cause currents of about six times rated value. In other words, the impedance appears to be the same in all three phases, but its value is now $6 = 0.16 perunit, or 16 per cent. The effect of mutual winding coupling alone may make the effective impedance per phase as low as 16 per cent or as high as 300 or 400 per cent. There is one significant observation. So long as the three currents are equal and separated by the same angular displacement, the effect of currents I s and I , on the voltage drop in phase A will be identical with the effect of currents I , and I , on the 'IG, Inductionvoltage drop in phase B and also with the effect motor impedance dioof currents I , and I , on the voltage drop in phase grcm. C . Thus the effective impedance will appear to be identical i n all three phases; that is, the impedance voltage drop in the A phase will bear the same relationship to the current in the A phase as the impedance drop in phase B bears to the current in phase B and as the impedance drop in phase C hears to the current in phase C . Or expressing this symbolically,

kJ! '.'

InZn _ = I,

BZB _ I_ _

ICZC

IS

I C

Thus Z , = Z B = Z c . This also identities the fact that the impedance voltage drops I a Z A , I,Z,, and I c Z c are separated by the same angles as I,, I , , and I c . These are two very important facts which emphasize the value of symmetrical components.

POSSIBLE SYMMETRICAL COMBINATIONS

There are but three possible symmetrical combinations in a three-phase system in which the three phase quantities are equal and separated by the same angle. The displacement angle must be a multiple of 120" since the three phases of a three-phase system are separated by 120". This is shown in the following three cases using currents for illustration.

Case 1. I , is 120' behind I , and Ic is 120° behind I,. Case 2 . I , is 240" behind I d and I , is 240' behind I,. Case 3. I , is 360" behind I , and Ic is 360" behind I,.


117

The vector relationships represented by these three cases of symmetrical displarement are shown in Fig. 2.3. Henceforth reference will he made to case 1 as the positive-sequence component denoted by a suhscript 1 characterized by three equal vectors 120' apart in the normal sequenre A R C ; to rase 2 as the negative-sequence component denoted by a subscript 2 charact,eriaed by three equal vectors 120" apart hut with a sequence A C B opposite normal; and to rase 3 as the zero-sequence component denoted by a subscript 0 chararterized by three equal vectors with zero angular separation (in phase with each other). Even a t the risk of unnecessary repetition, the two important properties of these three symmetrical components are repeated. The circulation of any one of the three symmktrical three-phase current patterns in a symmetrical three-phase circuit, even though the phase windings are mutually coupled, yields a balanced three-phase impedance voltage drop whose sequence pattern is identical with that of the current pattern. Likewise, the application of any one of t,he three symmetrical three-phase voltage patterns on the circuit will give rise to a balanced three-phase current whose sequence patterti is ideutical with that of the voltage. 1. Current flow of one sequence pattern produces voltage drops of the same sequence pattern only. 2. Applied voltage of one sequence pattern produces currents of the same sequence pat,tern only. 3. For each sequence pattern, the impedance can he regarded as a definite fixed quantity identical in all three phases. This then is the significance and identity of the symmetrical components (of which there are three types in t,hree-phase systems) and may be applied to voltages as well as currents.

Ic CASE I lPOSlTlVE SEOUENCEl

CASE 2 (NEGATIVE SEWENCE)

FIG. 2.3 Symmetrical patterns of current.

CASE 3 (ZERO SEOUENCE)

118

THE OPERATOR


0

I n the application of symmetrical-component analysis there will be repeated need t o shift a particular vector by multiples of 120". Particularly in analytical studies it will be advisable merely to indicate the desired operation, leaving the actual resolution until the final solution is approached. Invariably it will be found that combinations of operations appear modifying a particular vector which can be directly reduced t o much simpler form, or often simply vanish. The small letter a is used to indicate an angular advance of 120' in the vector t o which it is appended. Its use parallels the use of j as a 90" advance operator, i.e., aIb would mean a vector of the same magnitude as labut advanced 120'; while azZbwould mean a vector of the same magiiitude as Ib but advanced 240'. 0-12

O+j1.732

-02

0.5tj0.866

/

\+ /

I/

I

//'

/

f

/ //

/71:a2

, , 1.5tJO.866 /

/

/

a3

\

0.5 Ir-0.-50.866

FIG, 2.4 Functions of the 120'

operator

0.

1.5-~0.866 :I a


119

The significance of commonly encountered combinations of a operators is indicated in Fig. 2.4. For instance (az-a)Iswould indicate a vector fl times as large as laand advanced 270' in angle. Comparing the operators j and a in more detail to explain Fig. 2.4, a vector 1 t o the right on the horizontal, Fig. 2.4, when multiplied hy .i would be 0 jl. That same vector multiplied by a becomes (in terms of j ) - 0.5 j0.866; multiplied by u2 it hecomes - 0.5 - j0.866. 1-a then becomes 1 - ( - 0.5 jO.866) = 1.5 - j0.866 or an advance of 270" and times as large.

+ +

4

+

RESOLUTION OF SEQUENCE C O M P O N E N T S

It develops that any possible patt,erri of three-phase currents or threephase voltages can be resolved exactly into rombinatioris of the three types of symmetrical components. Some properties of the three symmetrical-sequence components will he of interest in showing the nature of their independence and the manner in mhirh they may he separated. Referring to Fig. 2.5 it will he seen that, if the vertor sum of the three vectors of each component is made, the answer will be zero for the positive-sequence and negative-scquence systems and 3 for the zero-sequence system. If first the B-phase quantity is advanced 120' and the C phase advanced 240" and the vector sum then evaluated, the answers will he 3 for the positive-sequence system and zero for the negative- and zerosequence systems. But if first the B-phase quantity is advanced 240" and the C-phase quantity advanced 120", the vector s u m will then be zero for

ADVANCE B 120" ,I C240*

FIG. 2.5

ADVANCE 0 240c 1200

Properties of rymmetriccll-component quantitiei.

I20


the positive-sequence system, three for the negative-sequence system, and zero for the zero-sequence system. This suggests that the three sequence components have independent degrees of freedom. Suppose that the three actual line currents I,, I,, and I , are to he resolved into three balanced-sequence components of types positive, negative, and zero. If I,, I,, and l c are added vectorially, it may be expected that whatever positive-sequence and negative-sequence component were contained therein would add u p to zero, and the answer should he three times the value of the zero-sequence component.

IA I.0

+ I B + I c 31.0 Pi + I s + I c =

=

3

If the B-phase currerit is first advanced 120' and the C-phase current 2.40' and then added, it can be expected that whatever negative-sequence and zero-sequence component were coutained therein would add u p to zero, and the sum should thus he three times the positive-sequence component.

+ a l , + a Z I c 3I., 11 + a l a + a21c I.,

1,

=

=

3

In similar fashion hy first advancing the B-phase current 240' and the C-phase rurrent 120' the sum should then he three times the negativesequeuce component in the A phase.

I,

+ a z I B+ a I c = 3I,,

1-2 =

11

+ a2Ia+ aIc 3

Sinre each of thc sequence systems is symmetrical, one can immediately identify the corresponding comporierits in the other phases. Refer to Fig. 2.3 to cherk the angular positiou of phase components. Zero sequence:

Id

=

I,,

=

Ih0 =

Positive sequenre: 1,

1.0

=

I,

+ I, + I, 3

+ a l e + a21c-

IS,= all., I C 1= aI.,

2 = =

+ + alc + + Ic

a21A I S 3 aIA a21a 3


Srgativc sequence:

I,,

=

I,

121

+ a2IB+ a l e 3

.\I1 three i.urrctits whii.11 romprise each of the three component systems now have been dekiiicd. The sum of all t,hree compotrcnt currents of each phase should equal the original actual phase current. Phase

il

:

Phase 13 : I B

+ la? f + I I I+ a l c + aIr -t I , + azIc + I , + I , + I c 3 3 3 = >SIA(aP + a + 1) + I a ( l + 1 + 1) + I c ( a + a2 + 1) !5(0 + 31, + 0) = I B = Ib,

IbO

- a?IA =

Phase C:

+ I,, + I d + a21e + IC a'IA + a l a + IC + I., + 1, + I c 3 3 3 4$IA(a + az + , I ) + IB(a* + a + 1) + I c ( 1 + 1 + 1)

Ic =

I,,

+

=

=

>$(O

+ 0 + 3Ic) = Ic

? >

1 hus a means now has been devised of separat,ing the three actual line currents (or voltages) into t,hree systems of symmetrical components, and further it, has bee11shown that the sum of the three component quantities of earh phase does exactly equal the original true line current (or voltage). Several fuiidamental equatioiis and commonly used relationships are listed i n Tahle 2.1. INDEPENDENCE

OF SEQUENCE SYSTEMS

The fact has been developed that, in symmetrical circuits, currents of one sequence produce voltages of the same sequence only and likewise

122


impressed voltages of one sequence produce currents of like sequence only. I n other words, there is no mutual coupling between scquence systems. Thus the voltage drops in impedances can be separately evaluated for earh sequence componerit of current and the resulting volt,agc drops added t o get the total voltage drop. Thus in Fig. 2.1 the t,otal impedance drop across the impedance Z in the direction of current flow is Phase A :

(IZ).

=

IdZI

+ I,,Z, + I,OZ,

Similar expressions could he written for the other two phases, hut a simpler attack is possible from concepts already acquired. The positivesequence drops will all be of equal magnitude and of positive sequence, the negative-sequence drops will all be of equal magnitude and of ncgative sequence, and the zero-sequence drops will he of cqual magnitude and of zero sequence. Therefore, Phase B :

(IZ),

=

a21.,Z,

Phase C:

(IZ), = a,I.,Z,

+ aIa2Zr+ I,,Zn + aZIa2Zz+ I,oZO

Here for the first time the advantage of the symmetrical-component approach can be appraised. For each symmetrical-compoiierit system, impedances can he regarded as having a definite fixed value identical in all three phases. The impedauce values in the three component systems may he widely different, howcver. That is, Z, may he altogether different from Z2 or Zo. Unt,il the actual currents were resolved into symmetrical componcnt,s, there seemed no alternate t o thc use of self atid mutual impedances in each phase. At this point note that under balanced-load (wnditioiis the current is entirely of positive sequence. Thus t,he usual solution of balanced operation is really a special case iiivolving only the positive-sequence system, i.e., positive-sequence voltages, poshive-sequence currents, and positive-sequence impedances. The application of these principles t o the solution of unbalanced-load problems now may he studied. It seems appropriate at this point to review some physical concepts of the three compi~neutsystems. All source machines generate only positive-sequence vokage. The winding pattern in the A phase will he repeated in the B phasc 120 electrical degrees later and i n the C phase 240 electrical degrees later. Thus identical voltages will be generated in each phase minding except that the B phase mill be 120' behind the A phase arid the C phase will he 120" behind the B phase.


TABLE 2.1

123

Fundamental Equations

With line currents I A , I*, I c known, sequence currents are 1.0

=

In

+ Is + I c = I,, = I d 3

NOTE:Voltages E., Eb, and E. generated vithin halaneed-winding rotating machines are entirely positive sequence. Commonly Used Relationships:

Negative- and zero-sequence voltages result from the impedance drop produced by the flow of negative- a n d zero-sequence components of current. Generally, positive-sequence voltages will he greatest at the source machines and diminish as one moves toward the short circuit. On the other hand, negative- and zero-sequence voltages will he greatest a t the

124

SYMMETRICAL COMPONENT5 FOR THREE-PHASE SYSTEMS

short-circuit point and diminish as one approaches the source machines, Positive-sequence voltages and currents produce (and are associated with) magnetic fields within rotating machines which rotate in the same direction as normal rotational dirertion. Negative-sequence voltages and currents produce (and are associated with) magnetic fields in rotating machines which rotate in a direction opposite to normal rotation. The latter thus produce machine torques tending to slow down a motor rotor, and the positive-sequence electrical quantities must produce a torque equal to the load torque plus that resulting from the negative-sequence current if normal running speed is to be maintained. Zero-sequence currents are in phase in all three conductors. For such currents to flow a t all it is evident that the electrical neutral must be connected to a fourth conductor or ground. Being in phase, the currents add up numerically at the neutral ronnection and become 31.0 in the neutral circuit, Zero-sequence currents produce a stationary pulsating magnetic field in the rotating machine stator winding which is predominantly of stator-leakage character, very little of which crosses the air gap to enter the rotor. Zero-sequence current will rarely be found in motors since the motor neutral is almost never grounded. PER-UNIT SYSTEM'

While symmetrical-component analysis is valid regardless of the system of units used, it will be found desirable to adopt the per-unit system. In the per-unit system, potentials are expressed as a fraction of an arbitrarily assigned line-to-neutral voltage (usually the normal operating voltage). Currents are expressed as a fraction of an arbitrarily assigned circuit current. This base current is usually selected to correspond with a convenient round-number kva such as 1000, 10,000, etc. Only two quantities can be arbitrarily assigned, i.e., base voltage and base kva or base voltage and base current. Unit values of all other quantities become fixed as soon as the first two are assigned. Unit base voltage and current are arbitrarily assigned a t some one part of the system. The values of unit voltage and current at other partsof the system become those which would result from the turn ratio of interconnecting transformers. The per-unit impedances define the fraction of base voltage which will be produced by the flow of unit base current. The value of the per-unit system is a t once apparent. The impedances of generators, motors, and transformers when expressed in per-unit on their own rating as a reference base assume almost a constant numerical * S e e Chap. 1, p. 52.

125


value throughout a wide range of physical size and voltage rating. For example, the impedance of a transformer mill be about 0.05 per-unit (6 per cent) on its own rat,ing as a base quite independent of size or voltage rating. If expressed in ohms, the numerical value of Z would vary widely wit,h 110 sigu of any common denominator. Also, in the per-unit system a particular per-unit value of current flowing into one side of a transformer comes out the other side as the same per-unit value. Refer to Chap. 1, page 54, for a complete list of equations relating ~ e r - u n i tvalues. SYSTEM APPLICATION

The approarh to circuit problems consists of writing the relations existing between geuerated voltages and impedance drops in the usual conventional manner except that three sequelice systems may he involved. In t,he simple cirruit arrangement shoivu in Fig. 2.0 it cau he seen that oue can directly evaluate (in terms of the A phase) Positive sequence:

E. Val

= =

+ + v., + +

+

r.,(zol zLl zr,) 8. - Z,I(ZC, ZL,

ZTd

Segative sequence:

0

=

v.,

=

Zero sequence:

0

=

+ + + + + I.o(Zo0 + + Zro) +

I.dZm ZL2 ZTJ Tio* -ra2(zGs zL2 zr2) ZLO

vm0= -r,,(zoot z t o+ zTd

v.0

Combined : Ti"

= v.1 =

E,

+ ve/02+ v.,o -

Iai(Zoi

+ Z L I+ Z

T -~ Za,(Zci

+ Zr.2 + Z -

T ~

+

I~O(200

ZLO

+ ZTO)

It will be useful to draw the individual sequence circuits such as indicated on Fig. 2.7. Xote that the circuit for the positive sequence is ZG "WY

Ec/

ZL

."..

".,. ..." E+T . ..... .,.A

FIG. 2.6

Typical symmetrical three-phase circuit.

1A-h

I6 -w

tc

*

vb

V0

vc


126

z L2

262

I

zT2

102-

Vaz

N

I-

SEQUENCE1

E a = 101 ( Z G I + Z ~ I + Z T ~ I ~ V O I

0 = 102 ( Z G Z t Z L 2 1 Z T Z 1 t

VO2

0 - 100 (ZGO+ Z L O t Z T O I t YO0 VA = v o l t v02+v00

=

E o - I o l ( Z G l t Z L l tZT11-102

-1w

FIG. 2.7

(ZGZ+ZLZ+ZTZI

~~tz,~+z,l

Equivalenl sequence circuits of Fig. 2.6 (in terms of the A phase)

identically that which mould be used alone for balanced-load prohlems. In the treatment of unbalanced loads, two additional circuits are involved (negative and zero sequence) which appear about the same evcept that there are no generated voltages therein and the respective sequence impedances are used. TYPE OF APPROACH

Through experience in the application of symmet,rical-component analysis, partirular types of approach, appropriate selection of reference phase, and useful equivalent circuits have been discovered vhich lead to a solution in the simplest manner. Generalized solutions of problems presented in short-circuit studies of three-phase systems (circuit-breaker selection or relay appliration) include the following forms of short circuits: 1. Three-phase 2. Line-to-line 3. Line-to-ground 4. Double line-to-ground THREE-PHASE SHORT CIRCUITS

The three-phase short-circuit condition represents a balanced threephase short circuit on the system. Only positive-sequenre quantities are involved; hence only the positive-sequence impedance system will be needed. The solution thus simplifies to an analysis of a single-rircuit


127

network involving only positive-sequence impedances and is done in the familiar conventional mariner as follows, using Figs. 2.8 and 2.9: Balanced operation [ ( l ) (2) (3) tied together] For balanced load Z, per phase, make Z, = Z, For thrce-phase short circuit, make Zx = 0 Reference phase: L4

I , = I., =

I, Ic

= =

E.

z,+ z x

a21A all

The solution becomes simply

I,

=

I,, =

~

E.

total Z I

I s = a21A I , = aIn

FIG. 2.8

Actual three-phore circuit pattern.

_ _ _ _~_ _ - _ - -

1

I

POSITIVE SEQUENCE

v,;

N+ _c

I I

zx - -v/JI.- - -2

Ib=lol

EO

2, +ZX

IS= 0216 IC =

o h

1.3

FIG. 2.9

Equivalent circuit for three-phore short-circuit analysis.

LINE-TO-LINE SHORT CIRCUITS

The generalized solution works out in the simplest manner by considering the short circuit t o exist between the B and C phases, using phase A as the reference, as illustrated in Fig. 2.10.

I28


E2F

The boundary coiiditioris a t the short-circuit point are

I, I,

v,

=0 =

-Ic

=

vc

LINE-TO- LINE SHORT CIRCUIT (SOLID1

vc

II' - 0 Ig

=-Ic

vg

=vc

SHORT CIRCUIT EQUIVALENT SEOUENCE CIRCUITS IN TERMS OF THE A PHASE

(b)

LINE-TO-LINE SHORT CIRCUIT (SHORT CIRCUIT IMPEDANCE ZF)

EcE5 :.

v b<'

VB

vc

PER PHASE

SHORT CIRCUIT EOUIVALENT SEQUENCE CIRCUITS (A PHASE REFERENCE )

FIG. 2.10

Circuits involved in line-to-line short-circuit bnolyrir,


129

So zer(i-sequenre rurrent is involved since

Ian =

I*

+ I , + 5 -- o + 3

I* - I , 3

=o

The positive- and negative-sequence currents in the A phase will he diametrirally opposite sinre

+ aIll + a l l , -- 0 + a I s - a l l , --~ (a-aa)IB 3 3 3 I,, + d l , + a I c - 0 + a l l a -_ a I= s _ (az-a)Is _ - - __ I,? = Id =

I,

3

I.,

=

3

3

(a-a’) I , 3

-Id

The solution now hinges 011 the equality of voltage on the B- and (‘-phase ronductors at the short rircuit.

V B = a2E. - a21.1Z1 - aI.*ZZ V , = aE,, - aI.lZ1 - azI.zZz To make V”

= ITc

a2E, - a21.,Z1 - aIo2Zz = aE. - aI,,Z,

- a21.,Z,

Substituting - I., for I a 2 and collecting terms

(a2-a)E. = (a2-a)I,,,ZI - ( a Z - a ) ( - I . , ) Z 2 E. = 1.1z1 IdZ9 = IG1(Z, ZJ

+ +

la, =

al., = - aIa1 -

I,

la,

=

-

aE.

z , + z, ~

+ l a z = (a‘-a) z , E.+ z* 6 ~

=

B.

- -Ic mz -

The portion of the solution which contains the circuit parameters E./(Z, Z , ) suggests an equivalent circuit in whichthe positive-sequence system Z (containing the driving voltage E . and impedance Z , ) is in series with the negative-sequence impedance system Z,. Also i t is noted that in the reference phase A the negative-sequence current is the negative of the positive-sequenre rurrent. This leads t o an equivalent circuit connertion shown in Fig. 2.11. The magnitude of total rurrent in the B- or C-phase conductor is times as much as either of the components. I n most applications, only

+

6

130

SYMMETRICAL COMPONENTS FOR THREE-PHASE SYSlEMS

the magiiit,ude is of interest,, in which case 110 attentioil need bc given the relative phase angle between this current and the refereure voltage. The same generalized siilution ran bc applied to a rase in vhich the short circuit contains impedanre. Suppose the linc-to-line impedance t o be Z F . This can be simulated by i.onsidering the systcm to he extended through an additional symmetriral branch containing an impedance ZF/" per phase. 4 solid line-to-line fault at, the end of this branch produres the efleot nf an impedance Z P ronnerted line-to-line on the basic system. The solution is as follows, using Figs. 2.8 and 2.11: Line-to-line connect,ion (line B to line C ) ( 2 ) connected to (3); (I) open For a line-to-line impcdance Z F , make ZX = Z F / ~ For a line-to-line short rircuit, make ZX = 0 Reference phase: A (Bourrdary conditions: I , = 0, I , = - I c , V ( , ) = V ( , ) ]

.

.

EG

- -j

+ z,+ z,

Resolving further, the solution hecomes simply Is

E,

v3

z, + ( Z F / 2 ) + z*+ ( Z P / 2 ) = En = 4 1P.F z1+ z, + z,

=

POSITIVE SEQUENCE

I

I I NEGATIVE SEQUENCE

-

~

Iai

I II

-

Val

- I,

zF/2

-"W&

I

N

22

va 2

zF/2

* - +v'AVP

1a2FIG. 2.1 1

Equivalent circuit for line-to-line short-circuit analysis.


131

LINE-TO-GROUND SHORT CIRCUITS

Refer to Fig. 2.12 for circuit conditions. SOTE: Circuit is symmctrical except for short-rircuit connect,ions. The simplest solution is arrived at by sclecting as the reference phase t,hat phase on which the short circuit, exist,s. IMPOHTANT SOTF:: Zero-sequence current flows through the neutral impedance Z,, but in Z , the magnitude is 31a0. Thus thc voltage drop will be three timcs as much as would be produred by Z , inserted in each phase. Since Z,is defined as the impedance per phase, the corrert value of Zo t o represent the neutral impedance Z, mill be 3 2 , . This mill he true of all circuit impedances appearing in the neutral conductor. Thrir equivalent Zomill be three times the value of Z,.

Zn

{+:%: ID+

-

IB+

L-0 SHORT CIRCUIT ON PHASE A

vc

IC+

BOUNDARY CONDITIONS vp. = o Ig = o

Ic = o

.. EOUIVALENT SEQUENCE CIRCUITS I N TERMS OF T H E A PHASE

3Zn

FIG. 2.1 2

Circuits involved in line-to-ground short-circuit analysis.

The three sequence circuits are defined in Fig. 2.12. The boundary conditions whirh must be satisficd at, the short circuit, are Solution.

v, = 0 I, I,

= =

0 0

The relat,ionships which prevail in the symmetrical part of the system are E. = I J i V-1 0 = I O Z Z l v.2 0 = I.oZ0 v 0 .

+

+ +


132

Equating these to satisfy boundary conditions,

+ V,, + Van E, - l o i Z i - Io2Zs - I,oZo + I,, + Ian a21.1 + al,, + I.u 0 + Ic2 + I,n al., + a21a2+ 1.0 0

V , = V-1 I, = IS, I C = I,, Subtracting

=

=

=

=

=

=0

IC from l e gives

+ (a-a2)ia2+ 0 = 0

l a- I , = (az-a)lal

(a2-a)I., I.,

=

(a2-a)I,z

=

1-2

Substituting t,his result into I , gives

+

IB= a21a1 al,, 4 I.u = 0 = a21,1 aI,l I,, - I., I,, = 0 (a' a 1)1., - 1-1 I,o = 0 I,, = I d

+ + + +

+

+

Thus:

I.,

=

id

=

I,,

This fact might have been evident by the geometry of line currents at the short circuit. The sum of the three component currents in the B and i n the C phase must be zero. Only if the individual component currents are equal and 120' apart in both the B and C phases could this be possible. But this would mean that in the A phase the component currents would be equal and in phase. = I,, and IaO = I., into the V Aequation gives Substituting

v, = E.

-

I,,,Z,

- I*,Z%- I.,ZO

Ea = I , I ( Z I I

+ Z , + Zu)

= 0

E.

- z , + z*+ z,

-

This suggests that the solution can he made in terms of an equivalent circuit in whirh the generated voltage E. is impressed on the three impedanre networks Z,, Z 2 , and Z o in series. It is more accurate to think of this to he in the form

Ea = I,IZI

+ I d 2 + 1,oZo

This still suggests the series ronnection of the three networks hut recognizes that the current in Z , is Ia1, in Z 2 is lo2, and in Zois Io0. Since 1-1= i 0 2 = Ia0 there is no conflict with Kirchhoff's law at the junction between the individual sequence networks. The important result is the equivalent-circuit concept by which the sequence networks ran he interconnected to yield an answer for the value of I., = I,, = 1,". The equivalent-circuit concept is helpful even when


133

the solution is to be obtained by numerical computation, hut it is of partirular importance if use is to he made of a d-e or an a-e network analyzer (calcuiating hoard). Knowing the value of I.,, the value of current in the fault (I,,) is

10 = I A = 1.1

+ 2.1 + I,,

=

31.1

whirh is t,hree times the current found directly from t,he equivalent circuit. Where t,here is impedance in the short circuit or in the neutral path, the procedure outliiied above is modified as shown in Fig. 2.13. z ..

I

:-

L-G CONNECTION THROUGH IMPEDANCE ZF

I

M 4 G h E TnE SYSTEM S EXTEhDED THRObGh Q BALAhCED C.RCJlT OF ZF PER PhASE LZFI’ZF~=ZFO=ZFI

A SOLID SHORT CIRCUIT TO GRD

NOTE: SINCE I B = I c = O ,THE INCLUSION OF ZF IN THESE PHASES PRODUCES NO ERROR, THbT I + :

FIG. 2.1 3

ZERO

BEYOND THIS IMPEDANCE RESULTS IN ZFCONNECTED LINE-TO-GROUND. LUSETHESAME PROCEDURE

External impedance in the line-to-ground connection.

_---__-__-

N

zo

SEQUENCE

VVAv

Iao FIG. 2.14

I

-1

ZF

“20

.,,* +*.,

--t

Equivalent circuit for line-to-ground short-circuit analysis.

I I

-1


134

Summarizing, the solution becomes (see Figs. 2.14 and 2.8) Line-to-neutral connection (line A to ground) (1) zonnected to ground; (2) open; (3) open For a line-to-ground impedance Z,, make Zx = Z , For a line-to-ground short-circuit, make Z , = 0 Referenee phase: A [Boundary conditions: I , = 0, ZC = 0, V ( , ) = 01

I., I A

= =

E.

+ z*+ zo + 32" + 3 2 , 3E. I d f Id + Id z,+ Z? + zo + 3z,2+ 32, I.,

=

I,o

=

ZI

=

Other Cases. The equivalent circuits by which other common circuit conditions can be evaluated are worked out in a similar manner as, for example, a double line-to-ground fault would he solved as follows using Figs. 2.8 and 2.15:

Double line-to-ground solid fault (line B t o C to ground) (2) and (3) connected to ground; ( 1 ) open; Zx = 0 Reference phase: A [Boundary conditions: I , = 0, V(,) = V(,) = 01

v,,

=

v,,

=

Vao

=

E,

-

I,,Z,= I,,

ZZO

z,+ Z" ~

Rotating-machine Characteristics. Positive-sequence currents are associated with mmf patt,erns which rotat,e at synchrouous speed in the normal rotational dirertion. The effective pi)sitive-sr(ioetice reactance is consequently influenced by time. For the first cyrle of short-circuit current, the subtransient rcact,ance of synchronous machines and the standstill reactance of induction machines apply. Within a few cycles the subt,ransient effects have decreased t o negligible proportions and the transient reactance of synrhronons marhines is i t i control while the effective impedance of induction marhines has inrreased to a value close to t,he normal running impedanre ( i t 1 the order of 100 per cent, 011 its o\vn base). During t,he next, serond or t,wo, t,wo artions are taking place in the synchronous machine. Induced field currents are decaying, and t,he

-


'

I35

FI

I

POSITIVE SEOUENCE

$+EO

I

& z

vl:

Iai+

I

- - -1 I

I ZERO SEOVENCE

VOO

I

+--A

100FIG. 2.1 5

Equivalent circuit for double line-to-ground short-circuit analysis.

effective machirie reactance is approaching the synohrooous reactance. The effective voltage ahead of synchronous reactance is approaching the value established by the steady-state field current and may he influenced by the operat,ion of a n automatic voltage regulator. Rarely nil1 it be neressary to evaluate short-circuit-current magnitudes for prolonged time intervals, but it will he well t o recognize that special treatment will be needed t o obtain correct results in such cases. Negative- and zero-sequence impedances of rotating machines car1 he considered as remaining constant regardless of the duration of shortcircuit-current flow. TRANSFORMER CHARACTERISTICS

The zero-sequence circuit produced by various transformer connections is often a source of trouble; so a considerable number of typical comhinations are defined in Fig. 2.16. The positive- and negative-sequence impedauces are equal as are those of all stationary winding circuits. There is one tricky aspect associated with Y-delta or delta-Y transformers. There is an inevitahle phase displacement hetween the highand low-tension line circuits. Standard convention has agreed that the terminals designated H , and X I shall be those which are only 30" apart. Present st,andards also st,ate that when operated with electrical sequence A B C ou t,erminals HI, H,, H 3 the high-tension system will lead the lowtension system by 30'. This displacement is the result of winding geometry and is not of the


136

CONNECTION ZERO SEQUENCE CIRCUIl

-(

ZT 4 -

x

F

3

x

2 131 ( - P H

Y SPECIAL CASE- 3 - P H CORE TVPE

---(

-7: N

WYE-WYE

WITH DELTA T E R T I b R I

131 I-pn P N

* CLOSED IF THE CORRESPONDLNG TRANSFORMER NEUTRAL I S GROUNDED. LT IS THE NORMAL TRANSFORMER AS POSITIVE SEQUENCE 2 1

FIG. 2.16 connections.

Z (SAME

Zero-sequence circuits clrrocioted with common three-phase transformer


137

nature of an impedanre voltage displacement angle. Thus if the standard transformer is operated with reversed sequence, i.e., electrical sequence .,IN' assuriated with terminals H,, H , , HI, the high-tension system will lag t,he low-tension system by 30". l3y reason of these fa&, in a Y-delta or delta-Y transformer with standard ronnertions operating with normal sequence, the positivesequenre rurrerit and voltage iii the high-tension circuit will be advanced 30" with respect to that i n the low-tension circuit, while the negativesequenre mrrent i n the high-tension -cirruit will be retarded 30", as is defined in Table 2.2. Tronsformer Zero-sequence Circuits. The zero-sequence circuit produced by various transformer couiiectious is ofben a source of trouhle; so a considerable number of typical combinations are defined in Fig. 2.1G. By first examining the zero-sequence properties of simple winding patterns, it ivill then be possible t o identify understandably the serosequetirc circuits of more complicated practical transformer connections. Delto Winding Connection. Zero-sequence current cannot flow in the circuit t o a deltt-connected 11-inding (see Fig. 2.17) sinre there is no eleetrical conuection t o ground by which it could return, even though zerosequenre current can flow within the closed delta circuit. Thus the zerosequence circuit is always interrupted at a juiirt,ion with a delta-connected minding. Y Winding Connections. Zero-sequence current cannot flow in a cirruit ronnerted to a Y-connected winding if the neutral is not grounded (see Fig. 2.18). Thus the zero-sequence circuit will be interrupted at the jurirtion with a Y-connected winding if the neutral is.uugrounded.

Iao

FIG. 2.17

FIG. 2.18

-

A circuit connecting with

(I

delta-connected transformer winding.

A circuit connecting with an ungrounded Y-connected transformer winding.


138

TABLE 2.2

Transformer Phase Shift

With standard d-lta-Y or Y-delta transformms, H I (high voltage) will hc 30" shcad of X I (law voltage) for normal phase sequence. H I will hc 30" behind X , with opposite phase sequence.

PHASE SHIFT I N

A->

OR

)-ATRANSFORHER

Standard, H , 30" ahead of X I

Many investigators pwfer to exprrss the relationship hetween high- and low-tension line currents in B slightly different manner so as to simplify the associated phase shift opcration, for example, Standard, H I 30" ahead of X I

I:, = -jr

ZC

= +jZO

o,

1;.

I.,

= -jZo1 = -jib,

I:,

= i i I b ,

1:, = -j1tIs

=

+jI-*

= +j1bx

H I 30" behind X ,

z;,

= +I,, I:, = + j ~ * ,

z:

= +I<>

Z'r2

=

-jIa9

NOTE: If currents w e not in per-unit, the transformation ratio must also he factared in.


I39

Zero-sequence current, in a circuit connect,ed t o a grounded-neutral Y-connected winding can flow if zero-sequence rurrent, in t,he secondary windings can he caused t o flow in the direct,iori iiidicat,ed by the secoridary arrows, (see~Fig.2.19).

FIG. 2.19

A circuit connecting with

0

grounded Y-connected tronrformer winding.

If the secondary currents in Fig. 2.19 cannot flow, the primary zerosequence current is limited t o the magnetizing current of the core (in t,he order of 5 per cent of rated current for 100 per cent impressed voltage). This represents a Z O of ahout, 2000 per cent on the transformer rating, which for practical purposes may he regarded as infinite. A n exception to this rulc is presented hy the thrce-phase core-type design whose construction is as indicated in Fig. 2.20. The flow of zerosequence current, in the primary windirig produces magnetic flux whii,h is in phase in the same direction in all three core legs. Since there are no external core legs between upper and lower core yokes (as would exist in a shell type of three-phase design), the zero-sequenre flux must return largely through the air. The steel tank walls provide a fairly low reluctance path forpart ofthereturn circuit, but thecrossover to t,he core yoke at both the topand bottom isdirectly through air. The magnetizing reactance c represented by this flux path FIG. 2.20 The three-phase c k - t y p e tronrusually he in the order of 30 to ,-, .. 50 per cent on the t,ransformer rating, which is low enough to have practical significance. Zero-sequence current in a circuit connected to a grounded-neutral Y-connected winding can flow if another set of transformer wiridiiigs is connected in delta as in Fig. 2.21. The closed delta provides a circuit for t h e flow of zero-sequence current. The impedance presented to the flow of current is the interminding impedance Z, (the same as the normal positive sequence ZT), Kote, however, that the zero-sequence currents are not repeated in the outgoing line circuit but are short-circuited within the delta winding.

140


111 a Y-Y-connected transformer& b&L neut,rals grounded, as in Fig. 2.22, zero-sequence current can flow if the 'reflected zero-sequence current in the other winding finds a closed circuit at some point along the connected circuit. I n this case the tramformer t,ransfers zero-sequence current from primary circuit t o a serondary circuit in the same manner that it transfers positive- or negative-sequence current. The transformer simply inbroduces a series impedance in t,he zero-sequence circuit which in magnitude is identiral with the normal positive-sequence impedance Zr. With this understanding of elemental behavior, the equivalent zerosequence circuits for the usual transformer connections can he directly resolved. Some of the more rommon ones are identified in Fig. 2.16. When drawing zero-sequence circuits for extensive systems, it is a good plan t,o designate transformers in the manner shown in Fig. 2.16, showing an interruption of the zero-sequence circuit by an open gap. By this method one ran be constantly aware that a break in the cirruit was intentional and not the result of an oversight. Circuit Resolution Example. I n Fig. 2.23 is illustrated a particular typical syst,em. The resuking composition of the positive-, negative-, and zero-sequenre circuits is also portrayed. Suppose that the immediate problem concerns the evaluation of various performance qualities on the 2.4-kv system radiating from bus L1. The first step involves a resolution of equivalent impedances by which the entire hulk system t o the left of hus L4 is expressed as a single equivalent impedanre. This would be accomplished by successively paralleling, etc., until firially a single equivalent impedance value connecting with bus L , is obtained which would then look like Fig. 2.24. In many cases

A circuit connecting with a grounded Y-connected transformer winding with a delta winding on the same core structure.

FIG. 2.21

FIG. 2.22 A circuit connecting with a grounded Y-connected transformer winding with another grounded Y winding on t h e same core structure.


141

i t vill he at onre apparent that the impedanre of transformer T swill he the major rontrolling impedance in the circuit from Ai',. I n this case it may be entirely reasonable t o consider that rated voltage is sustained on the high-tension side; or consider the short-circuit rapacity at the hightension terminals to be about equal t o the interrupting rating of the

MOT

MOT

.S. 2.23

Typical system example.


142

switching equipment on the bus M,. Obviously, some such approximation will be required in practically every problem since the actual line interconnections will otherwise extend over the entire electrical distribution system of several states. It will be of interest t o note that the zero-sequence system is quite discontinuous, which is typical of practical systems. In the present illustrative problem the zero-sequence system associated with bus L 4 is independent of that on bus M2. For comprehensive studies of extensive system networks, the equivalent sequence circuits shown in Fig. 2.23 might he set up on the d-a or a-c network analyzer. T o examine an operating characteristic at the point P I , each individual sequence circuit would be tapped at the point PI. For each sequence system the correct impedance network is that obtained from the tap lead P I and it,s own neutral bus N . The interconnections between sequence networks will he governed by the type of unbalance (see Figs. 2.8, 2.9, 2.11, 2.14, 2.15). Provision is made in the network analyzer directly to measure current in or voltage across individual branches of all three networks. Measurement of Individual Components. Useful measurement connections by which a particular sequence quantity may be independently resolved, or a particular sequence quantity excluded, are identified 011 Fig. 2.25. The circuits for obtaining 10 or Eo alone are frequent,ly used. (In applying potential transformers for measuring EO on an urigrounded neut,ral system, line-to-line rated transformers and secondary loading resistors should he used t o avoid overvoltage hazards.) The delt,a-connected current transformer circuit (which excludes l o in the output) is useful in providing internal-short-circuit protection for grounding transformers. The circuits for individually segregating the sequence quantities I , and E', are only rarely used. Possible applications would be (1) making a single-roil voltage regulator responsive t o positive-sequence voltage of

F I G . 2.24

Simplification of Fig. 2.23 for study of performance on bur 14.


143

the three-phase system, ( 2 ) providiug a protective relay which will trip if the sustained negative-sequence voltage exceeds a preassigned value. It is of interest t o note that the usual open-delta line-to-line connected potential transformer application excludes goin the secondary.

C U RR E N T

IK=CTR.VIOI NEG4TIVE SEQUENCE

ZERO SEQUENCE

VOLTAGE l i = mR A T I O I ZERO SEO'lENCE

POS!TIVE

SEQUENCE

JO 8661

NOTE

FIG. 2.25

- 8"

INTERCWNGING LINES B B C METER WILL READ Vo2

Measuring circuits for segregating specific components.

Chapter 3

by Donald Beernan and R. H. Kaufrnann

Selection of A-C Short-circuit Protective Devices and Circuit Equipment HOW TO BE SURE OF ADEQUATE SHORT-CIRCUIT PROTECTION To design an industrial power distribution system adequate from a

short-cirruit st,andpoiiit, the maximum short-circuit current at any point should he less than the short-circuit rating of the equipment applied at that piiint. When systems are so designed, it is common t o speak of them as having adequate short-circuit protert,ion. I n other cases, they are said t o have adequate int,errupting capacity (IC). Horn can one be sure that, a plant, dist,ribution system is adequate for all short-circuit eonditions? Mere are the steps t,o follow: 1. First accurately determine the available short-circuit currents a t all sigriificant poir1t.s in t,he system, using the methods outlined in Chaps. 1 atid 2. These rdrulat,ing procedures have been verified by many tests on actual systems and in short-cirruit testing lahoratories. Nariy former fallarious ideas w1iir.h led to the installation of inadequate short-circuit prot,ectire devires arid circuit elemerit,s hare beerr dispellcd. For example, the idea that, ouly about 20,000 amp maximum short-circuit rurreiit could he obtairied at -180 rolt,s has heen dispelled by actual measurements of short-circuit currents of over 100,000 amp a t this voltage. L'iitil t,he magnitude of short-cirruit currcnt is known, one cannot be sure thitt adequate short-rircuit protcrtion is provided. 2 . Iiistall only short-circuit protect,ive devires such as circuit breakers, s\ritvhes aiid fuses, and r~ombinat,ioumotor starters of known adequate I44

A-C SHORT-CIRCUIT PROTECTIVE DEVICES AND CIRCUIT EQUIPMENT

145

interrupting sating. Use circuit elements of known adequate shortcircuit-current rating. Equipment of adequate short-circuit rating can be obtained t o meet the requirements of all industries when proper consideration is given t o system design from a short-circuit-protectiorr standpoint. 3. Prepare for load growth. If the system is installed with circuit breakers that are large enough only for present requirements, the cirruit breakers will become too small from ail interrupting standpoint when capacity is added. The system should be designed and the circuit breakers should be selected on a hasis that will enable expaiisioii without exceeding the circuit-breaker interrupting ratings. Short-circuit stresses must be checked too, as the stresses increase as the square of the shortcircuit-magnitudes. Future expansion can be accomplished at practically no added expense in the initial installation by employing a modern poiver-distribution-system layout (see Chaps. 11 t o 15). Main and auxiliary switchboards in hundreds of plants in operation today were installed years ago when the plants were small. The power demand was limited then, and only small transformers were required t o supply the plant. At t h a t time the switchboards may possibly have been adeqnat,e. However, as the plants grew, more power was needed. Xew feeders were added t o carry the new load, arid new transformers were added t o the bus t o supply the added load. 111 many cases no thought was given t o t,he circuit, breakers because t,hey carried their load currents satisfactorily. Hovever, when new transformers were added, the capacit,y of the power supply inrreased. Consequently, the available shortcircuit current also iiirreased. This higher short-circuit current imposed added interrupting duty on the old circuit breakers when they were required t o clear a faulty feeder cable. Often this added short-circuit current was sufficient t o bring the total short-circuit current beyond the rating of those existing circuit breakers. However, through oversight nothing was done about it, leaving the plant vulnerable t o a major shutdown if a fault occurred which one of these old circuit breakers failed t o clear. Failure t,o consider the effect of increased short-circuit currents has heeti one of the most common causes of many of the older installations being unsafe. 4. Do not he complacent, Many systems which have been operating for years have never had a major short circuit. Operators of these systems have come to believe t,hat short circuits never occur; so they do not bother about interrupting rating of rircuit breakers. This belief compares with the assumption that fire insurance is not necessary because the factory has never burned down. The older the system grows, the weaker the insulat,ion becomes and the greater the possibility of major short cir-

146


cuits. The circuit breakers too are often inadequate in these old syst,ems. Thus, when a short circuit does occur it is almost cert,ain to cause a major shutdown with possible damage t o other propert,y as well as loss of production. 5. Use an engineering approach. If the short-circuit,-protection prohlem is approached on an engineering basis instead of depending on good luck, the plant investment can be more adequately protected and undue risks eliminated. Good luck over a period of years may give a false assurance that failures are never going t,o occur, but, good luck eventually runs out as it has in so many cases. The cost of a loss due t o a failure then is far more than it would have been to modernize the switchgear oil a planned step-by-step basis. In the engineering approach a study is made to determine t,he weak spots in t,he electric system and remedy them hefore a major shutdown occurs, with attendant financial and production loss. The engineering approach is of a prevent,ive nature, i.e., finding the weak spots and correcting them before a failure does occur. No one would t,hink of running a boiler indefinitely just hecause “ i t had never failed.” Preventive maint,enance involves continually repairing and replacing weak parts hefore they fail. The results of the failure of an inadequate circuit breaker can he as serious as a boiler failure; so the same intelligent engineering approach should be used in providing safe, adeyuat,e circuit breakers as is used with other machinery even thongh one has heen lucky enough over a period of years t o avoid the failure of an inadequate circuit breaker. Luck might change for the worse tomorrow; so it may pay real dividends not t o be complacent ahout short-circuit conditions. To have a safe power system with low maintenance cost and high service continuity, adequate circuit prot,ertive equipment is necessary throughout the ent,ire system from the place where the power system enters the plant down t o t,he smallest motor or light. An Example of W h a t Can Happen When Available Short-circuit Currents Exceed the Interrupting Rating of Short-circuit Protective Devices. An inadequate circuit breaker mas mounted in a svit,ch riiiim which was part of the distribution system. A short circuit occurred in the outgoing rable. The short-circuit duty was well above the interrupt,ing rating of the circuit breaker i n the switch house. As a result, the circuit breaker attempted to open the circuit hut did not havetheability todoso. Therefore, the circuit breaker failed, blew up, and when it did two things happened. First, the circuit breaker at the source had t o clear the fault in t,he failed circuit hreaker and thus drop all the load instead of just the one * ion. load on the fauky hranch. This meant unnecessary loss of prodwt‘ Second, a fire resulted and completely destroyed the switch house.


FIG. 3.1

147

Rerull of foilure of inodequote oil circuit breaker on heovy short circuit.

Fortunately the switrh house was isolated from nl.her I~uildit~gs, and orily the switrh house burned dn\vn. llad this fsilurc ocwrrcd i n a fiict,ory tiuilding, the damage could have been much more cxLensive. r. Ihe picture, Fig. 3.1, tells the st,nry of what happened hotter thaii ii book of mords could. The irony of this fiiilurc was that, the plant, erigineer had ri~cogriizetlt h e inadequacy of the circuit hrcakers in this swit,ch house aiid was replacing t,hem with adequat,e ones. The ot,her circuit breakers in this switi.h house had already hcen rcplaced mit,li adequat,e unes, and t.liey \wre destroyed too. One can never tell how long hia luck will last wii.h inat1t.quat.e circuit breakers or fuses. It, may rim out sooner tliaii one thinks.

SELECTION OF THE TYPE OF SHORT-CIRCUIT PROTECTIVE DEVICE r l

1 here are marly features t o he cnnsitlered in t,he sr?lcvtioii of shortcircuit protective devices t,n provide adequate short-circuit prntertinn for an industrial power syst,em. One of the most import,ant,is that t h e shorteircuit protective device be adequate for the service. The adrqiuicy of circuit breakers, fuses, or motor st.art,ers can be determined from t h e procedures outlined in Chaps. I and 2. Ariother important function of mnst short-circuit, prot ive devices is that t,hey also provide a means of switehi!ig circuits nder normal operat,irig conditions. T o m( requirements fully and eomplctcly for both circuit switching and short-

148


circuit protection, a protective and switching device should fulfill the following basic specifications: 1. It should be capable of being safely closed in on any load current or short-circuit current within the momentary rating of the device. 2. It should safely open any current that may flow through i t up t o the interrupting rating of the device. 3. It should automatically interrupt the flow of abnormal currents u p t o the interrupting rating of the device. There are two fundamental devices that are commonly used for or have as one of their functions short-circuit protection. These are: 1. Circuit breakers 2. Fuses Some motor starters are used for short-circuit protection, hut in general these have either circuit hreakers or fuses as the short-circuit protective element. A basic comparison of fuses and circuit breakers will be made and their area of application outlined. More detailed comparisons are made on the basis of syst,ems voltage classes, i.e., (1) 600 volts and below and (2) above 600 volts. CIRCUIT BREAKERS-GENERAL

M e e t s All Requirements. A modern circuit breaker meets all the basic requirements listed above. It is designed and rated to be capable of heirig safely closed in on any current within its momentary rating (some oil circuit breakers do not fully meet this requirement). It can safely open any current within its interrupting rating. When proper relays or tripping devices are applied, i t is capable of automatically opening any current which is above the pick-up setting of the tripping device and below its interrupting rating. It combines in one unit a device for safely switching the circuit under normal as well as abnormal load conditions and t o automatically open abnormal rurrents up t o its interrupting rating. Eliminates Single Phasing. Circuit breakers, in all except a few special cases where single-phase switching is used in transmission-line circuits, open all ungrounded conductors of a circuit. Therefore, the probability of single phasing of three-phase circuits is eliminated from a practical standpoint in so far as the circuit protective equipment is concerned. Adjustable Tripping Time and Pickup. The total time t o operate under various overcurrent conditions is adjustable for practically all circuit breakers. The adjustment is either in the built-in tripping devices or in the relays associated with the circuit breakers. The adjustability of time of operation makes the circuit breaker ideally suited t o selective operation as is required for circuit protective service in a system. Electrical Operation. Circuit breakers in general are suitable for elec-


I49

trical operation, which means they can be used for automatic control, remote operation, etc. Furthermore, auxiliary circuits are available on practically all electrically operated circuit breakers for the control of external auxiliary or process circuits. Wide Selection of Time-current Characteristics. Various types of relays with special characteristics to meet particular service requirements can be used with circuit breakers to broaden their scope of application. For example, time-delay overcurrent relays which match motor-heating curves can be used to enable the circuit breaker to be used for motor starting and running and short-circuit protection. Or the relays may he specially designed to protect transformers or any ot,her piece of equipment or circuit. This makes the circuit breaker and its associated relays almost universally applicable as a short-circuit protective and switching device . Repeated Operations. Circuit breakers can repeatedly open abnormal currents without destroying t,he interrupting element, Of course, inspection and some maintenance may be required after each duty cycle at or near their interrupting rating. When the circuit, openings are repeated a few cycles or seconds apart, derating factors must be applied. But fundamentally the circuit breaker does permit repeated operations without destroying itself or affecting the accuracy of operatirlg time. Same Degree of Protection after Operation. Since when a circuit breaker operates it does not destroy itself, there is little likelihood of affecting calibration of time and pickup settings; hence the same protection is afforded all the time. Minimum Temperature Effect. Most circuit-breaker time overcurrent tripping devices and relays are not appreciably affected by temperature. Hence, greater accuracy as a function of ambient t,emperature can be maintained than by devices that depend upon t,hermal conditions to activate them. Moderate Operating Speed. Circuit breakers in general are not so fast in operation a t high overcurrents as are most fuses. Wide Choice of Current Ratings. Circuit breakers are available up to 4000 amp cont,inuous current rating at GOO volts and less and up to 1200 to 5000 amp a t higher voltages. Trip-coil ratings run from 15 amp up. Interrupting levels are available from 5000 to 100,000 amp a t GOO volts or less and from 15 mva to 25,000 mva at higher voltages. Rigid Industry Standards. Circuit breakers are made under rigid industry standards which prescribe complete interrupting ratings for them and methods of test for establishing interrupting ratings. These permit the application engineer to apply them on a sound safe basis and within their rating. Proper derating factors must be applied for repetitive-duty cycles and high-altitude applications. *

* Refer to applicable NEMA standards.

150

A-C SHORT.CIRCUIT PROTECTIVE DEVICES AND ClRClllT EQUIPMENT

FUSES-GENERAL

Fuses are often considered for circuit protection because of their low first cost. Before selecting fuses in place of circuit breakers, there are certain general characteristics and limitations which must be recognized and considered as well as cost. While fuses have their proper applications, one must look rarefully a t the fuse picture in general and then more closely a t specific fuses to see how many of the hasic requirements are met. Generally Do Not M e e t All Requiremsnts. One of the first and foremost considerations is that fuses in themselves do not meet the basic requirements for a complete short-circuit protective device. Fuses alone (except t,he oil-fuse cutouts) do not incorporate any switching means to permit closing in on high currents or to switch load currents. T o meet the basic requirements it is necessary that a fuse other than a n oil-fuse cutout be used in conjunct,ion with a properly rated interrupter or safety switch. In this combination the fuse provides the ability to open ahnorma1 currents automatically. The switch should provide the ability to open load currents and moderate overcurrents which are below the blowing point of the fuse and should provide the ability t,o safely close in on short-circuit currents up to the interrupt,ing rating of the fuse. When the switch is in the closed position, it should be able to carry safely whatever current the fuse will pass. The operation of fuses in combination with interrupter switches at moderate overcurrents imposes problems not easily overcome. The fundamentals of the problem can be seen by referring to Fig. 3.2. To illustrat,e one phase of the problem, let us assume that it takes $6 see only to close and open a switch manually. Should there be a moderate overload when the switch is opened and closed rapidly, as there may well be because of connected motors, etc., the switch would have to open perhaps several times its rating because the operation took place so quickly that the fuse did not have time to melt. This area is represented by the crosshatched section of Fig. 3.2. For example, an interrupter switch might he rated to make 20,000 amp, carry 20,000 amp momentarily, and to open 100 amp. This switch, when used with a 100-amp E-rated fuse* or even a much smaller rated fuse, may not be adequate on moderate values of current,. At 1000 amp, for example, the blowing time of the fuse may be 3 see. An operator may close the switch and open it within 36 see. The fuse would not have had time to melt, and the switch vould be required to open 1000 amp, or ten times its rating. Whenever the circuit interruption takes place in two separate devices which are * E-rated fuses will carry their rated eurrmt eontinuouslv and blow in 5 to 10 min at 200 to 264 per cent of rated current.


151

interdependent on each other for complete functioning over a wide range of currents, there is always this problem of operation on moderate overloads which is much more difficult t o overcome than operation a t very high short circuits where the time for the fuse to clear is very short. As a precautionary measure for increased safety, the switch element of fused switches should be closed with a fast positive action and not opened immediately. This will give the fuse a chance to melt on moderate overcurrents before the switch is again opened. BASIC CHARACTERISTICS OF FUSES

Possible Single Phasing. Fuses are single-phase devices; therefore, one fuse may blow, leaving a multiphase circuit supplied with only singlephase power. It may not completely isolate a faulty circuit. Nonadjustable Tripping Time or Pickup. When fuses are used, their pickup setting and time-current setting are changed by changing the size or type of fuse.

AMPERES FIG. 3.2

Interrupter-switch rating and fuse time-current characteristics showing performonce on moderate overcurrent..

152


Limited Choice of Characteristics. Because fuses are thermal devices, the choice of shape of time-current characteristics is very limited for coordination purposes. Nonrepetitive Operation. Fuses, once they have operated, must be replaced. Certain types of fuses have removable links which permit salvaging part of the fuse after it has blown. Replacement cost of fuses varies widely, depending on type and size of fuse. Protection M a y Be Reduced or Lost after Operation. Because the interrupter destroys itself, care must be taken to replace a blown fuse with one of the same rating and characteristics. Otherwise, protection may he lost. There is always the danger that if no fuses are available short bars or wires may be inserted to keep power on. When this is done, all protection may be lost. Affected by Ambient Temperature. Fuses are thermal devices; therefore, their operation is subject t o variation due t o ambient temperature changes. This effect in fuses is much greater than in relays or circuit-breaker tripping elements. It is less important in high-voltage circuits. Fast Operating Speed. Fuses are generally divided into two classes: (1) non-current-limiting and (2) current-limiting. The current-limiting fuses possess two important advantages, particularly for branch-circuit protection: ( 1 ) Berause of extremely fast operation, they limit the damage due to the flow of short-circuit current. (2) They actually limit the shortcircuit-current magnitude to far less than the available short-circuit current, thereby allowing the use of smaller conductors and equipment in branch circuits. This current-limiting ability is one of the most useful characteristics of the fuses in branch-circuit protection. Non-current-limiting fuses also operate faster than circuit breakers at currents near their interrupting rating. The fast operat,ion of most types of fuses, however, makes it difficult and often impossible to coordinate them with other short-circuit protective devices located beyond the fuse in the circuit. Therefore, fuses in general are best suited for branch-circuit protection where they need not operate selectively with other protective devices between the fuse and the load. Choice of Current Ratings. Fuses are now available for low voltages (600 volts and less) up to 4000 amp. For circuits above 600 volts, the upper limit of fuse ratings is in t,he range of 100 to 400 amp. Fuses are generally limited in size hecause of thermal considerations. Large fuses may produce so much heat that ventilation and mounting become severe problems. Also, as current-limiting fuses become larger

A-C SHORT-CIRCUIT PROTECTIVE DNICES AND CIRCUIT EQUIPMENT

I53

and larger, they lose more and more of their current-limiting ability. Sinre the current-limiting ability of fuses is most useful in branch-circuit protection, the handirap of having to use small ratings to get effective rurreut-limiting artion is not so pronounred, as most branch circuits are of small rurrent rating anyway. Industry Standards. Fuses above 600 volts are made according to indnstry st,andardsesrept, that standardized levels of interrupting ratings are not set up. Low-voltage fuses have no a-c interrupting standards, although surh st,andards may be available in the future. See further disrussion nuder voltage classification. Mechanical Simplicity a t Low Current Ratings. Fuses and their associated switches for low-current circuits, i.e., about 200 amp or less, are simpler mechanically than circuit breakers. For higher current circuits t,he switrh, if built, t o have the necessary momentary and interrupting abilit,y, loses its advantage of mechanical simplicity. CIRCUIT BREAKERS VS. FUSES-GENERAL I11 selecting circuit breakers YS. fuses, the techniral ronsideratious certainly favor the rirruit breakers in most rases. Because of this, circuit breakers are generally considered the only acceptable protective devices by most engineers for all'lorations in industrial plants where switching and short-rirruit protectioii is required except for some hranch circuits and control circuits and motor starters. Fuses and switches are preferred for some hranch rircuits because of the fast operation of the fuse. Besides the technical roiisiderations, economirs is a factor. While cost is very important, it is secondary to the technical considerations noted above and secondary to select,ing the devire that has an adequate interrupting rating for t,he servire. Berause there may he in some cases a wide difference in rost between circuit hreakers and fuses, there is a tendency to get so involved in economic issues in the selection of circuit breakers vs. fuses that technical ronsiderations are lost sight of. AS a result many hazardous syst,ems are installed to save a few dollars in first cost, a saving that may soon be lost because of the poor performance and higher maintenanre of inadequate equipment, particularly in low-voltage circuits. It is for that reason and because the technical cansiderations vary somewhat with voltage that the technical considerations are reviewed in further detail as a function of voltage class. There are other factors in the selection of fuses for overcurrent protection. These factors involve mainly coordination with relay time-current characteristics or the time-current characteristics of built-in devices on circuit breakers (see Chap. 9).

I54

A-C SHORT-CIRCUIT PROTECTIVE DEVICES A N D CIRCUIT EQUIPMENT

SHORT-CIRCUIT PROTECTIVE EQUIPMENT FOR SYSTEMS 600 VOLTS AND LESS

For low-voltage systems rated 600 volts or less, there are three commonly used types of short-circuit protective devices for protecting main power circuits, secondary feeders, and branch circuits. These devices are 1. Large air circuit breakers (sometimes referred to as magnetic circuit breakers) of which the one shown in Fig. 3.3 is typical. 2. Molded-case circuit breakers of which those shown in Fig. 3.9 are typical. 3. Fused safety switches of which the one shown in Fig. 3.11 is an example of a high-quality safety switch and fuse. There are panel boards which are used for protection of small branch circuits. These are used mainly in lighting and in small power systems and employ either small molded-case-type circuit breakers or fuses as their overcurrent protective means. LARGE AIR CIRCUIT BREAKERS

Description. The large air circuit breaker, Fig. 3.3, consists of a n operating mechanism, contacts, an arc interrupter, and usually a built-in overrurrent tripping device. These circuit breakers are characterized by their sturdy construction, ample electrical clearances, availability in highcurrent-carrying and interrupting and momentary ratings. The tripping devices are adjustable as to their pickup setting and operating time, and various shapes of time-current characteristics are available. The ratings available are listed in Table 3.1. TABLE 3.1

Ratings of Low-voltage Large Air Circuit Breakers for A-C Service

1nterrvpting roting., rm, amp a.ymmetric.al

Range l r i p - d ratings,' omp

_____

240 volts and below

241-480 volts

__. ~

* Standai

and below

241-480 volts

600 volts

30-225 150-600 60+1,600 2,000-3,000 4,000

25-225 100-600 400-1.600 2,000-3,000 4,000

15-225 35-600 200-1,600 2,000-3.000 4,000

_____

~

30,000 50,000 75,000 l00,000 150,000

600 volts

~

240 voitl

25,000 35,000 60,000 75,000 100,000

15,000 25,000 50,000 75,000 I00,000

rating8 are 15, 20, 25,35, 50,70,90, 100, 125, 150, 175, 200, 225, 250,275, ., -__., .... .> - .__.)

Application. These circuit breakers are intended primarily for application in main switchboards where pou'er may be generated a t low voltage

A-C SHORT-CIRCUIT PROTECTIVE DEVICES A N D CIRCUIT EQUIPMENT

155

FIG. 3.3 Large air circuil breakers mounted in drawout metal-enclosed low-voltoge rwifchgear.

or where it may he received from the utility at low voltage and for the secondary svitchgear of load-center unit substations or in main subdistribution centers, Fig. 3.4. They are also applicable for individual branch-circuit prokction where t,he highest qualit,y device is required and where special time-current characteristics are necessary for coordination. They are particularly applicable for braneh-circuit protection for larger loads over 200 amp or for smaller loads where, as stated above, highest quality protection is desired or electrical operation is required. These circuit breakers have longer life built into them than do other types of low-voltage circuit breakers and are, therefore, suitable for many more operations, particularly where there is moderately repetitive duty imposed. Selective Tripping vs. Cascading. Large air circuit breakers may be used either in selective tripping systems or in cascade systems. Selective tripping systems, Fig. 3.5, are those in which the circuit breakers are set to trip selectively so that the one nearest the fault operates first so that only the faulty portion of the circuit is deenergized. I n this case all circuit breakers should have adequate interrupting ratings, that is, their rating should be equal to or greater than the short-circuit duty a t the

A-C SHORT-CIRCUIT PROTECTIVE DEVICES AND ClRCUlT EQUIPMENT

156

point of application. There are additional problems of selecting the time-current settings which are discussed more fully in Chap. 9. I n cascaded operation, Figs. 3.6 and 3.7, circuit breakers may he used under certain circumstances beyond their interrupting rating. This applies where the main circuit breaker (commonly referred to in application tables as the A' circuit breaker) has adequate interrupting rating, that is, its rating is equal to or greater than the short-circuit duty imposed a t the point of application. The feeder circuit breakers (commonly referred to in application tables as the B circuit breaker) in this case, Fig. 3.6, may be used to twice their interrupting rating provided that the following conditions are met. The total kva of connected synchronous motors should not exceed 25 per cent of the supply transformer or

I

/

A

GENERATOR

A

LOAD CENTER SUBSTATION

nlnUNIT

- A

3

MAIN SECONDARY BREAKER

Y

SU0

ELECTRICALLY OPERATED

- DISTRI0UTlON CENTER

MOTOR WELDER 200 HP One-line diogrom showing typical applications of large air circuit breakers. FURNACE

FIG. 3.4

LOAD

A-C SHORT-CIRCUIT PROTECnVE DEVICES AND CIRCUIT EQUIPMENT

I57

generator rating. In addition t o the main circuit breakers having adequate interrupting ratings, their instantaneous tripping attachment must be set t o operate when the current through the backed-up or B circuit breakers is not more than 80 per cent of the interrupting rating of the backed-up or B circuit breakers. This ensures that the main circuit breakers will operate whenever the short-circuit duty exceeds the interrupting rating of the B circuit breakers.

-

& 1500 KVA LOAD CENTER UNIT SUESTDTION A MDIN CIRCUIT BREAKER RATE0 DT LEAST 50.000 DMP INTERRUPTING

FEEDER CIRCUIT EREDKERS RDTED 50,000AMP INTERRUPTING

SHORT CIRCUIT DUTY DT THIS POINT 50.000 DMP RMS ASYMMETRICDL WOOD DMP FROM THE TRANSFORMER DND 9000 DMP FROM THE MOTORS

ERANCH FEEDER CIRCUIT BREDK-

+FEEDER

CABLE

/

SHORT CIRCUIT DUTY DT THIS POINT 32.000 DMP RMS DSYMMETRICDL

~UE-BU~

7 A

*

/I' ERDNCH FEEDER CIRCUIT

)BREDKERS RATED 25000 DMP INTERRUPTING

V

SHORT CIRCUIT DUTY DT THIS POINT 22000 DMP RMS DSYMMETRICDL

NOTE: SHORT CIRCUIT LEVELS DT SUB E u s s E s n REDUCED DUE FEEDER CAELE IMPEDDNCE

am

FIG. 3.5 One-line diagram showing large air circuit breakers applied in selective tripping system. Time settings of overcurrent trip elements must be properly set to obtain selectivity.

158


I

i

&I500

KVA LOAD CENTER UNIT SUBSTATION

MAGNETIC TYPE

I

I '

I

MOTOR CONTRIBUTION 9000AMP

SHORT CIRCUIT DUTY AT THIS POINT 50.000AMP RMS ASYMMETRICAL

NOTE: INSTANTANEOUS TRIP ELEMENT ON MPIN BREAKER A MUST BE SET TO TRIP AT 16400 AMP THIS IS 0.8 OF 2 0 5 0 0 AMP

20500 AMP IS THE CURRENT FLOWING FROM THE MAIN TRANSFORMER THRU BREAKER A WHEN CURRENT FLOWING THRU FEEDER BREAKER 0 I S 25.000 AMP THE R A y i N G O F B R E A K E d B

FIG. 3.6 One-line diagram showing large oir circuit breakers applied in cascade with only one source of low-voltage power.

Motor contribution must be considered. The duty including motor contribution should not exceed twice the interrupting rating of the backed-up circuit breaker. However, the motor contribution may not come through the main circuit breaker. Therefore, the main A' circuitbreaker instantaneous trip setting may be less than 80 per cent of the backed-up circuit-breaker interrupting rating because the main A' circuit breaker must trip instantaneously when the total rms asymmetrical shortcircuit current through the backed-up circuit breaker is 80 per cent or more of the interrupting rating of the backed-up B circuit breaker. For example, in Fig. 3.6 if the backed-up or B circuit breakers are rated 25,000 amp interrupting rating, the short-circuit duty a t the point of application of the B circuit breaker should not exceed 50,000 amp rms asymmetrical. This may he made up of 41,000 amp from the main source and 9000 amp from the motors. The main-source circuit breaker must trip instantaneously a t 0.8 X 20,500 or 16,400 amp rms asymmetrical. It makes no difference whether the circuit breaker is applied a t the bus or a t some point remote from the bus. When the backed-up circuit breakers are applied a t points remote from the bus, such as circuit breakers B' in Fig. 3.7, the interrupting duty a t the circuit breaker ahead,


159

B in Fig. 3.7, may be in excess of twice the interrupting rating of the backed-up B' circuit breaker, but because of cable impedance the shortcircuit current a t the point of application of the backed-up circuit breaker B' must be limited to twice its interrupting rating. Circuit breakers operating a t beyond their interrupting rating in cascade mustbe inspected after each operation and may require more than normal maintenance after interrupting currents beyond their rating even though the main circuit breaker does open. Another qnalification is that the circuit breakers must be of the same manufacture and of similar characteristics. Feeder circuit breakersshould be electrically operated because the forces incident to closing against short circuits in excess of the circuit-breaker rating may preclude successful manual closing. Circuit breakers of two widely different interrupting

I

f &I000 -UNIT

I

I

KVA LObD CENTER SUBSTATION

I

NOTE! INSTbNTANEWS T R I P E L E M E N T ON FEEDER B R E I K E R B MUST BE SET bT 12000 bMP(OQ X I 5 0 0 0 1

-knunar

I

snom CIRCUIT DUTY HERE 26000 AMP RMS ISYMUETRICbL

FIG. 3.7 One-line diagram showing large air circuit breakers in cascade applied remote from the main source of power.

160

A-C SHORT-CIRCUIT PROTEtTlVE DEVICES AND CIRCUIT EQUIPMENT

ratings cannot be cascaded. As a guide to this, refer to Table 3.2 which shows the maximum interrupting rating circuit breaker which can he used to back up any given interrupting rating feeder circuit breaker. The ratio of the columns may be more than 2 : 1. The higher interrupting duty in the main circuit breaker often comes about because of having t o select it for continuous current-carrying rating rather than interrupting rating. Regardless of the interrupting rating of the main circuit breaker, the duty cannot exceed twice the interrupting rating of the cascaded B feeder circuit breaker. TABLE 3.2

Range of Large Air Circuit Breakers Which Can Be Cascaded with Each Other Main Circuit Breaker A' Interrupting Rating, Amp Rmr

25,000 30,000

50.000 60,000 75,000 100,000 120,000 150.000

Minimum Interrupting Rating of Coscaded Feeder Circuit Breaker B, Amp Rmr

15.000 15,000 15.000 25,000 25,000 50.000 75,000 100.000

Where there are two or more sources of current to a bus with cascaded feeder rircuit breakers, the following rule applies, Fig. 3.8. All main A' circuit breakers (i.e., A : , A : , A : ) must be tripped instantaneously when the total short-circuit current through the hacked-up B circuit breaker exceeds 80 per cent of its interrupting ratings. The example in Fig. 3.8 shows what the various instantaneous overcurrent trip settings of the main circuit breakers should be for a given case. The rule is that the instantaneous setting must be proportioned t o the short-circuit current delivered through the main circuit breaker in question. The interrupting rating of the B circuit breakers is 50,000 amp. When the total current reaches 40,000 amp, the current delivered by these various sources is 6000 amp, 8000 amp, and 18,800 amp. The motor contribution is 7200 amp. All currents are rms asymmetrical. Cascaded operation is a means of lowering the cost of short-circuit protection in secondary systems. In the cascaded system, smaller feeder circuit breakers are used than in the selective system; therefore this differential favors the cascaded system from an economic standpoint. I t must be recognized, however, that the service reliability of a cascaded system is poorer than that of a selective system because in a cascaded system, whenever a feeder short circuit draws a current in excess of 80 per cent of the interrupting rating of the feeder circuit breaker, the main circuit breaker is tripped out and service on all feeders served by that main circuit breaker or breakers is lost until the service is restored by reclosing the main circuit breaker. This application has proved satisfactory from a service-reliability standpoint for many industrial processes.

A-C SHORT-CIRCUIT PROTECTIVE OEVICES AND CIRCUIT EQUIPMENT

161

However, mhere critica1 processes are iiivolved, selective tripping is generally considered essential. Selection of Large Air Circuit Breakers. As a guide t o selertion of Iarge air circuit breakers for selective or cascade service, three-phase and single-phase, see Tablc 3.3.

d:3J) 7 4 7

, 0AMP 0 0

Y

18,000AMP

TOTAL SHORT CIRCUIT CURRENTAT THIS POINT 100,000 AMP R M S ASYMMETRICAL

WHEN TOTAL CURRENT THRU BREAKER B IS 0.8 OF ITS RATING I.E. 40,000 AMP BREAKERS n'i, "'2 8 d j MUST TRIP. THE CURRENT FLOWING IN THIS CASE AND THE INSTANTANEOUS TRIP ELEMENT SETTINGSX ON BREAKERS ~ ) . n ; . eA; A R E : -

r

MOTOR CONTR IBUTION 7200 AMP THIS FAULT DRAWS 40.000 AMP R M S ASYMMETRICAL

FIG. 3.8 One-line diogram showing lorge air circuil breakers in cascade wilh more ihan m e source of power to ihe main low-volloge bur.

Air-circuit-breaker Application Tables-Cascade System and Selective System

1

per CeP

Imped. once,

300 500 750

225

112.5 150

~

5%

5 5

4

4

____________

Kvo

Tmnsformer rating, three-phore

313 417 625 834 1,388 2,080 2,780 4,170

__

amp

:".,O"t

ti""O"l

Con-

load

Norm.

__

Short-circuit current, rrni amp (overage three-phase .Zmpe.&

Recommended interrupting rating of o i r c i r w i t breaker (see flgurer above)

208Y/120 Volts. Three Phoie

I load

Norm0

1.800

load

100%

,

A"

!

1 ' 1 'I,

I

,

I

ca+

code

ti*= trip

95,900 150000 I00000 50000 100.000

A'

>

5 z5 A

2

3rn

m bJ

c

c? ; I

B 6 relec-

Recommended interrupting rating of air circuit breaker (see flgures above)

1,350' 9,750 50,000 13.550 50,000 13,400 2,700 16,100 50,000 17.900 3.600 21 500 7 5 0 0 0 11,200

8.400

alone

Trans-

ompore.)

(average three-phase

Short-circuit current, rmr amp

240 Volts, Three Phase

600 Volts ond Less Ratings required for equipment for Ironsformer and feeder circ~itl,with selection of circuit breaker 8 on basis of cascade sydsm and selective trip system. Other fadois than short-circuit duty ore important in the selection of circuit breakers for selective trionine. Refer to monvfocturer for other lirnitotions.

TABLE 3.3

-


I63

I64


Standards. The XEMA Standards that, apply to all large air circuit breakers are KO. SG3-1951. MOLDED-CASE CIRCUIT BREAKERS

Malded-case circuit breakers, Fig. 3.9, are smaller in dimension, less sturdily constructed, and do not have t,he electrical &ararrces t h a t large air circuit breakers have. They are distinguished from large air circuit hrcakcrs primarily because of the fact that t,hey are mounted in a molded plastic case. These circuit, breakers have built-in trip element,s, and in some cases they are adjustable. Also marly functions cannot, be huilt into these smaller molded-case circuit-breaker tripping elements that can be huilt into the large air circuit-breaker tripping e1ement.s. It, is not easy t o make t,hem electrically operated or t o provide large numbers of auxiliary swit,ches. Ratings Available. Ratings are available as given in Table 3.4.

FIG. 3.9

Molded-care air circvit breakers mounted in a panel board.


165

I &75D

KVA MAX

A 1:

CIRCUIT CASE MOLDED BREAKERS

LOAD CENTER SUBSTATION WITH MAGNETIC CIRCUIT BREAKERS

{ 1h 1 )

I

1

I

I

+MOLDED )

CASE BREAKERS I N PLUG-IN DEVICE

INDIVIDUAL MOLDED CASE BREAKERS

DISTRIBUTION CENTER MOLDED CASE BREAKERS

FIG. 3.10 One-line diagram showing where molded-core air circuit breakers can be applied in a low-voltage power distribution system.

Application. Because of their small size and lower cost, the moldedcase circuit breakers find application for branch-circuit, protection where the interrupting duty is within their interrupting rating, Fig. 3.10. They also find applicabion on the secondaries of some small light-duty Ioadcenter unit substations. Not Suitable for Cascade Operation. These circuit breakers are not suitable for cascade operation wit,h large air circuit breakers berause they operate so fast that the large air circuit breakers are not able to protect them (see iVEhlA Standards for Large Air Circuit Breakers, Section SG3-3.43). Xeither are they suitable for cascading vith ot.her moldedcase circuit breakers. This conclusion mas reached after exhaustive tests.

166


Selection of Interrupting Ratings. As a guide, the portion of Table 3.3 referring t o circuit hreakers for selective operation may be used mhere the continuous current is less than 600 amp and iiiterrupting duty is within the available ratings of molded-case circuit breakers. Table 3.4 gives the interrupting ratings as defined by applicable NEMA standards. TABLE 3.4

Interrupting Rotings of Molded-cose Circuit Breakers for A-C Service Interrupling iatingi, r m i amp orymmetrical

240 ~ o l t s and below

7,500 20,000 25,000

30,000 50,000

i

Range of trip-coil rrrtingr amp

241-480 volts

600 ~ 0 1 1 s

I 15,000 20,000 25,000 35,000

~-

15.000 15.000 25,000 25.000

15-100 15-100 125-225 125-225 125-600

)O, 125, 150, 175, 200, 225, 250,

Standards. At preseiit there are no applicable NEMA standards for molded-case air circuit breakers.

FUSED SWITCHES

Fused switches, Fig. 3.11, consist of an interrupter switch and a fuse mouuted on a common base and usualiy in a metal enclosure. There are many types and varieties available. There are severa1 types of fuses available. The most common variety is the standard N E C (Kational Electrical Code) cartridge fuses. These fuses practically a11 corisist,of a fusible link enclosed in a cylindrical cartridge with connectors a t each end t o slip into the fuse clips of the switch. Xew and improved designs of fuses and switches for low-voltage service have been developed recently. The fuses are mainly of the currentlimiting high interrupting capacity silver-sand type, typical of which is the General Electric type EJ-6 fuse shomn in the smitch in Fig. 3.11. Typical of the improved switches is t,he type HCI switc,h as manufactured by the Trumhull Components Department of the General Electric Company. To be specific in the follomiiig discussion of the improved types of fuses and switches, the type HCI smitch and EJ-6fuseivill beused. Ratings Availoble. There is a wide variety of lon-voltage fuses and switches available. These ratings run from as low as a few amperes up t o


167

FIG. 3.1 1 High-copocity interrupting (HCI) enclosed switch with high interrupting-rating current-limiting silver-rand fuses (EJ-6).

several hiindred amperes. Cnfortunately most, of them do not have short-cirruit rat,ings assigned. Again, t,o be specifio, t,lie type IICI switch and E.J-6 fuse will be used t,o illusirate a-c short-circuit abilities which have hem established hy test,. The preserit availahle data are listed in Table 3.5. TABLE 3.5

I n t e r r u p t i n g Ability of Type HCI Switches and C u r r e n t - l i m i t i n g Fuses (1954)

Type HCI switch

Volts

! -I,

Amperes

II

Type EJ-6 fuses

I Volts

1

I

Amperes

15-20-30 200

EJ-6

lnterrvpling ability of combination HCI rwilch and EJ-6 fuse. amp byml

100.000 100.000 100,000 l00,000

Application. All small hiRh-interrupting-ability loix~-voltagefuses are current,-limit,ing in t,heir action, hence are very fast in their operation, and from this staidpoint they are partieiilarly well suited t o branch-circuit

168

A-C SHORT-CIRCUIT PROTECTIVE DEVICES AND CIRCUIT EQUIPMEW

-L

LOAD CENTER UNIT SUBSTATION

INDIVIDUAL WALL MOUNTED H C I SWITCHES AND E J - 6 FUSES

FIG. 3.12 One-line diagram showing whsre safety switches m d fuses may be applied as the lost protective device in low-voltage distribution circuits.


169

AVAILABLE SHORT CIRCUIT CURRENTASYMMETRICAL R M S AMPERES 1.25 I SYMMETRICAL (AVERAGE FOR THREE CONDUCTORS)

AVAILABLE SHORT CIRCUIT

FIG. 3.1 3

Curves showing the current-limiting choracterirtics of type EJ-6 silver-sand current-limiting fuses (60cycler).

I70


will withstand 9000 amp rms for 0.2 cycle. So, the 30-amp fuse vill protect a wire which will be required to carry 30-amp load current. This current-limiting feature, in addition to protecting small wires in systems of high short-circuit-current capacity, can protect small switching devices. It is for this reason that the type HCI switch can he used with type EJ-6 fuses 011 circuits where the available short-cirruit-current duty is as high as 100,000 amp. The t,ype HCI switch and EJ-G fuse combination has high interrupting rating arid is current-limit,ing in its operation which enables it to beusedin many places where molded-case circuit breakers would not have adequate interrupting rating and where large air circuit breakers would be too expensive, too large, or not applicable from an engineeriug standpoint. For example, a circuit breaker for a 30-amp circuit fed from a certain lowvoltage bus may require a circuit breaker with 100,000 amp interrupting rating. The wire or cable mould have t o be of the order of 350 MCM t o withstand the short-circuit current. I n the first place, a 100,000-amp interrupting rating circuit breaker cannot be built with a 30-amp trip coil that will withstand the short-circuit forces or heating. I n the second place, any 30-amp load devire mould not have terminals that would accommodat,e 350-MCM cable, the size required to withstand 100,000 amp. The use of an EJ-G current-limiting fuse and the HCI switch rated 30 amp would provide adequate short-circuit protection, and the currentlimiting effect of the fuse mould enable a wire of smaller size t o be used. The switch and fuse comhinat,ion is not generally suitable for main feeder circuit protection because of the fact that it is difficult to make the fast current-limiting fuses operate selectively with other overcurrent protective devices that would be in the circuit between the fuse and the load. Standards. Information for standards on fuses may be obtained from the Underwriters Laboratories, Incorporated, bulletin, Standard for Fuses. Information on st,aridards for enclosed switches (safety switches) may be obtained from IJnderwriters Laboratories, Incorporated, bulletin, Standard for Enclosed Switches or NEMA Publication No. 42-78, Enclosed Switch Standards.

SHORT-CIRCUIT PROTECTIVE EQUIPMENT FOR SYSTEMS ABOVE 600 VOLTS There are in general two types of short-circuit protective equipment available for systems above GO0 volts. These are: 1. Power circuit breakers 2. Power fuses

A-C SHORT.CIRCUlT PROTECTIVE DEVICES AND CIRCUIT EQUIPMENT

171

POWER CIRCUIT BREAKERS

There are many types of power circuit breakers availahle, but basically they are divided into the oil t,ypc and the nillcss type. I n the field 2.4- t o 13.8-kv t,he oilless-type cirruit breaker, Fig. 3.14, has largely superseded t h e oil-t,ype circuit breaker. In indoor metal-enclosed switehgear of the st,ation t,ypc for circuits 13.8 Lo 34.5 kv, the air-type circuit breakers are in general superseding the oil-type vircuit breakers. I n the field above 11.A kv for outdoor switchgear, oil circuit breakers are most commonly used, Fig. 3.15. For the sake of the discussion here relative to d e c t i o n of equipment>fiom a short,-rircuit standpoint, it makes no difference whcthcr the rircirit breakers are of t,he nil or oilless type. Ratings Available. High-voltage power circuit breakers are availahle in ratings from 2.4 kv up to over 300 kv and in interruptirig ratings from 15 mva up to 25,000 mva. Complete listings of power circuit breakers can he found iii the latest copy of S E R l A Standards SG&l954. T h e circuit, breakers most comtnonly used in industrial plants are the oilless or air type, sho\rn i n Fig. 3.14. The available ratings of this type of circuit breaker are given in Table 1.1 (Chap. I).

FIG. 3.14 Typical ille err (air) power circuit breaker ar wed in metal-clad switchgear for c i t w i t s rated 2.4 to 13.8 kv.

172


FIG. 3.15 Outdoor frome-type oil circuit breaker This circuit breoker i s rated 34.5 kv.

01

used in circuits rated above 15 kv.

Application. Power circuit breakers are applicable anywhere in the syst,cms rated 2.4 kv up t,o the highest a-c voltages in use today. They combine all the essential characteristics for circuit switching and protect,ion and therefore may be used at main buses supplied by generators or transformers or i n connection with unit substations. They are also applirable at, loral switching points and for protection of primary branch circuits (see Fig. 3.16). Motor Starting or Other Repetitive Duty. Certain of the power circuit breakers, particularly the oilless type, are suitable for motor-st,arting duty within the limitations outlined by the manufacturer. It should be noted that compared with contact,ors the principal limitation of power circuit breakers for motor-starting duty is the degree of repetitive duty that can be withstood. Contactors are designed for more operations and longer life under severe operating duty cycles than are power circuit breakers.

A-C SHORT-CIPCUIT PROTECTIVE DEVICES ANO CIRCUIT EQUIPMENT

173

Selection of Interrupting Ratings. The selection of interrupting ratings of power circuit breakers for industrial applications is out,lined in Chap. 1. A detailed description of the various faetors to consider in applying oilless eircuit breakers as used in metal-clad switchgear is given there.

I

Q

GENERATOR

I

Q P

69 KV

OUTDOOR POWER C I R C U I T BREAKERS

T TYWI

GENERATOR CIRCUIT BREAKER

TRANSFORMER SECONDARY CIRCUIT BREAKER

!

' MAlN FEEDER CIRCUIT BREAKER

A AHEAD O F L I N E OF L I M I T A M P MOTOR STARTERS

LARGE OU HIGH VOLTAGE MOTORS

FIG. 3.16 One-line diogrorn rhowing where oilless power circuit breakerr in metal-clad rwitchgeclr and outdoor power cirwit brecikerr may be applied in industrial power dirtribution ryrtemr.

174


Standards. Poiver circuit breakers are eovered by NEMA Standards SG1-19% POWER FUSES A N D OIL-FUSE CUTOUTS

There are many types of power fuses available for circuits rated 2.4 kv and above. These t,ypes of fuses, generally speaking, divide t,hemselves into three categories. The first is the power fuse, typical examples of which are shown in Fig. 3.17 which are for high-rapacity power circuits. The second type that is slightly differeni, i n construct,ion i s the oil-fuse cutout, which i s really a combination of a cntout and a fuse immersed in a container of oil, Fig. 3.18. The third type of fuse is used mainly in distribntion cutouts for overhead opcir-wire outdoor distriliutioii systems of utilit.ics in urban and suburban areas, Fig. 3.119.

FIG. 3.17 Typical high-voltage (above 600 volts1 power furer: Ifeft) current-limiting nonenpulrion silver-rand type, (right] "on-current-limiting expulsion outdoor type.

176


The last type of fuse mentioned is applicable toindustrial power systems for outdoor installations only where the interrupting rating is less than the duty on the system. This fuse is not metal-enclosed and is not for indoor installation. I n general, power fuses divide thcmselves into two classes, i.e., currentlimiting and non-current-limiting. Typical of the current-limiting category are the silver-sand fuses, Fig. 3.17(left). Typical of the non-current-limiting type are the oil-fuse cutout, Fig. 3.18, the expulsion fuses, Fig. 3.17(right), as well as the “boric acid” fuses and “liquid” power fuses. A further classification is that some are expulsion type, i.e., expel hot gases when they operate. These are not suitable for indoor application because of the hazard of the expelled hot gases. Such fuses are the expulsion fuse, Fig. 3.17(right), and the “boric acid” fuse without a condenser and the “liquid fuse”. Typical of the nonexpulsion type are the silver-sand fuse, Fig. 3.17(left), and the “boric acid’’ fuse with condenser. Application- General. All types of power fuses operate faster than power circuit breakers a t or near their interrupting ratings. Because of the fast operating time of the fuses, they are generally employed as the last circuit protective device in each voltage level in a primary power system, as shown in Fig. 3.20. Typical applications are in motor starters and ahead of primaries of transformers stepping down to a lower voltage. The silver-sand fuse, Fig. 3.17(left), is often the preferred type of fuse for power circuits because of its fast operating time and currentlimiting ability. However, in some cases where coordination is required, it may be necessary to use non-current-limiting types of fuses which have longer time delay. However, when the longer time delay is obtained, the benefits of reduction of damage to the circuit through which shortcircuit current passes is lost to a large degree. Interrupter Switches and Fuses. Nonexpulsiori-type power fuses suitable for indoor use are often applied in a metal enclosure with an interrupter switch to form a switch-and-fuse cornbination for high-voltage circuits. Interrupter switches are desirable for this application because they have interrupting ratings usually in the range of 100 to 400 amp. Plain disconnecting switches are generally not satisfactory for this service because they have no interrupting ability, and therefore the combination of the plain switch and fuse cannot be used as a load-switching device. The oil-fused cutouts combine in one unit the fuse and the interrupter switching element. Interrupter slyitches and fuses and oil-fused cutouts find wide application in industrial plants as the primary swit,ching and protecting section of a load-center unit substation (see Chap. 11). Application of Fuses in O p e n Switching Structures. Open-structure switches or disconnect,ing mountings without current-interrupting ability are often used with power fuses. These can be considered for isolation purposes only. Hazards in operations are materially increased in this


177

type of appliration. That is the reason that such applications should be limited t o outdoor structures Ivhere the operator is a considerable distance from the disconnecting switch when operating it. The use of such isolatiug switches i n series with fuses in indoor metal-enclosed structures is not coilsidered safe practice bemuse of thc proximity of the operator to t,he sivitrh and the possibility of the operator inadvertently operating the switch under roiiditions i u which the switch will hare to interrupt or close in on currents ronsiderably beyond its ability. Failure may result eveti though there is a fuse in series with such switches. 33 K V

P

UTDOOR TYPE FUSE SMALL POWER

IyTy\ TRANSFORMER

I LIMITING

AHEAD OF SMALL LOAD CENTER UNIT SUBSTATIONSUSE INTERRUPTER SWITCH AND POWER FUSE OR FUSED OIL CUTOUTS. FIG. 3.20 One-line diagram rhowing where high-voltage (above 600 may be applied in industrial power distribution systems.

VOllS)

Power

178


Selecting Fuse-interrupting Rating. Fuses are generally rated in amperes interrupting ahility. CMculate the short-circuit duty in rms amperes asymmetrical at the first half cycle as outliiied in Chap. 1, and select a fuse whose interrupting rating is greater than the duty imposed. Equivalent three-phase iiiterrupt,ing ratings may also be considered. Since the ratings of fuses are not too well st,andardized, refer t o t h e fuse manufacturer for complete data before applying fuses. Standards. Power fuses are covered by S E M A Standards, Cutouts, I’orer Fuses, and Current-limit,ing Resistors, Publication S(2-1954, and AIEE Standards S o . 25. MOTOR STARTERS

There are in general three kinds of motor starters: 1 . The contactor 2. The combination motor starter 3 . The circuit breaker Contactors are in general of two types, the most common variety being t,hose which have a n interrupting rating of only ten times normal rated current. These are completely inadequate for short-circuit protection and must have addit,ional short-circuit protection provided b y either fuses or circuit breakers. When a short,-circuit protective device such as a fuse or circuit breaker

FIG. 3.21 Typical lowvoltage 1600 volts and below) combination motor starter with current-limiting silver-rand furer for short-circuit protection.


179

is used in comhinatiori with contactors, it forms what is commonly called a combination motor starter. Circuits 600 Volts and Less (Fig. 3.21). In systems 600 volts and less there are t w i types of cornhination motor starters, both employing the same type of cont,act,or. The first employs a fuse disconnecting sm.itch alirad of t,he conta,ctor, and the other a circuit, breaker, usually a moldedcase-type circuit breaker, ahead of t,he cout,actor. The select,ion between t,hese two is based mainly on the fuudamerital differelice betveen fuses and circuit, breakers as short-circuit protective devices. The fused combinatirin motor starters have an over-all interrupting ahility so that the combination motorst,arter can successfully irit,errupt an available short-circuit current equal t,o 50,000 amp rms asymmetrical wheri equipped wit,h high-interrupt,ing-capacity currentlimiting fiises. This is for a short circuit outside ” the case of t,he mot,or starter and using type E.14 fuses. The molded-case circuit-breaker comhiiiat,ion mot,or starters are limited to a maximum duty of 15,000 or 25,000 amp rms asymmet,rical. Circuits above 600 Volts (Fig. 3.22). For circuits of 2.4 kv aud up t o 5 kv, the combination motor starter commonly used consists of current-limiting silver-sand fuses and contactors with the fuses mount,ed in disconnecting-type supports and placed in a metal enclosure s o interlocked that the fuses cannot be disconneeted unless the coritactor is in the open position. In this way the disconnecting fuse mounting has no current to interrupt. Since the FIG.3.22 Typicalhighfuses are for short-circuit protection only, suit,able (2,4to 4,8 kv) running overload relays should he provided in the bin tio motor motor st.art,er. These motor starters have inter- starter using current-limrupting ratings of 150 mva at 2.4 kv and 250 mva iting rilver-rand power a t 4.16 kv. From a short-circuit standpoint they furel for short-circuit protection. may be appIied up to their momentary and int,errupting rat,iiig. Since these devices contain fuses as the short-circuit protective element, they are naturally best suited t o application as the last protective device in the circuit. When used as motor starters, they are t,he last protective device, and therefore the fast operating time of the fuse is a very dist,inct advantage in limiting damage in the motors when a failure occurs. The fast operating time of the fuse also permits low settings on other relays further back in t.he system.

180


SELECTION OF CONDUCTORS AND OTHER CIRCUIT COMPONENTS FROM A SHORT-CIRCUIT STANDPOINT The floiv of short-circuit current in an electric system imposes mechanical and thermal st,resses (heating) on all component,sof the system through which such currents flow. This includes cables, bus bars, current transformers, disconnecting switches, as \veil as circuit breakers, fuses, and motor starters. The following is intended t,o aid in the selection of circuit component,s, ot,her than circuit breakers, fuses, and motor starters, from a short-circuit,-current standpoint. POWER-CABLE SELECTION FROM A SHORT-CIRCUIT STANDPOINT

Multiple-conductor power cables possess high mechanical strength because of the compact conductor lay and the continuous concentric binding arsist,ed many times by armor or lead sheath. KOlimit on mechanical stresses in such cables has been assigned. This is not true with regard to thermal effects. In common with ot,her current-carrying parts of the electric system during short-circuit-rurrent flow, the abrupt elevation in conductor temperature will be limited only by the ability of the conductor metal to absorb the heat developed. The magnitude of the temperature increase is greater (1) as the current magnitude becomes greater (as the square of the current), (2) as the conductor cross section becomes smaller, and (3) as the duration of current flow becomes greater. Temperature limits. Power-system short-circuit-current magnitudes, feeder-conductor cross section, and short-circuit protective device interrupting time should be coordinated to avoid severe permanent damage to cable insulation during an interval of short-circuit-current flow in the system. The effect should be limited to a moderate reduction in useful cable life (possibly 1 per cent of normal life). Reasonable maximum-peak transient temperatures for various cable insulations and operating potentials have been designated and in general are approximately 150 C (see Table 3.6). At a slightly higher temperature (approximately 175 C), destructive disint,egration of organic materials may occur, accompanied by smoke and combustible vapors. At somewhat higher temperatures large quantities of combustible vapors are expelled which increases the risk of explosion and fire. It is important to note that the abnormal temperature persists much longer than t,he duration of short-circuit-current flow. For example, the flow of 20,000 amp in a KO. 4/0-Awg copper conductor will elevate the


181

copper temperature from an initial temperature of 75 C to 150 C i i i ahout 34 see. With the current then redured to zero, about 1000 see d l be required for the copper temperature to return to 75 C in a 30 C ambient. TABLE 3.6 Conductor Rated Maximum Continuous Operating Temperature and Peak Transient (Momentary) Temperature for Various Types and Operating Voltages

-

MOX

lronrienl

lollogl Cable type

d.**,

copper temp.

kv

C

Vc type V or VL, single conductor or three conductor..

Impregnated paper (slid), single conductor..

lmpregnalod paper (did), three mnductor.

Type R*.

T i p s RH

Coronol

.......

..............

...............

............................................ ............................................

.............................................

* Applies to new

I 5 8 I5

85 85 84 77

I50 145 135 120

1 5 8 I5

85 85 85 81

I50 I45 140 135

1 5 8 I5

85 85 85 81

I40 135 I30 125

I

60

5

60

8

60

15

60

I40 135 130 125

1 5 8

75 75 75

I50 145 140

1 5 8 15

80 80 80 80

I50 145 I40 I30

-

-

type R (1947 code specification). t Actual operating temperature may be lompr because of consprvative application or a favorable ambient temperature.

Conductor H e a t i n g . On the basis that all heat produced by shortcircuit-current flow is initially absorbed by the rondurtor metal (wbirh


I82

TABLE 3.7

Quick Estimating Table o f Minimum Conductor Sire' A. Low-voltoge Air-circuit-breaker Protection

Short-circuit current. ~

omp

11.25

X rymmetricall

~~

1.5 to 2 cycles

>g s*c

linrt. trip) No. 8 Awg No. 4 Awg

5,000

I0,OOO 15,000

No. 2 Awg No. I Awg

25,000

No. 1/0 Awg No. 3 / 0 A w g

35,000

5o.m

300 M C M 350 M C M

75,000

100.000

Interrupting kvo at

Short-circuit

No. No. No. No.

4 Awg

No. 2 Awg No. 1/0 A w g No. 3/0 Awg 300 M C M

1 Awg 2/0 Awg 4/0 AWQ

300 M C M 400 M C M 600 M C M 800 M C M

400 M C M 600 MCM 800 M C M I000 M C M

Duration of hort-sircuit current

current, amp

(1.0

x

symmetricoll

3,000-3.500 3,500-4.000 4.000-4.500 4.500-5.000

5,000-6.000 6.000-7.000 7,000-8,000

8,000-9.000 9,000-1 0,000 10,000- 12.500 12.500-15,000 15.000-20.000 20,000-25.000 25,000-30.000 30.000-35.000 35.000-40.000

...... 25mvo ....... 75 m w N o . 6 A w g ........................... No. 4 Awg ............. 50 mva ....... No. 4 Awg ........................... No. 4 Awg ........................... No. 2 A w g 25 mva 50 m w ....... I50 mva No. 2 Awg ........................... No. 2 Awg ............. 100 mva ....... No. 1 Awg ........................... No. 1 Awg 50 mvo .............. 250 mvo No. 1/0 Awg ...... 100 mva 150 m w ....... No. 2/0 Awr ........................... No. 3 / 0 A w r 00 m

w

150 m w 250 m w 500 m w No. 4/0 AWI 250 M C M 300 M C M 350MCM

........................... ...... 250 mva ....... 750 mvo 5 0 m r o ....... 500 mvm .......

No. No. No. No.

2 Awg 2 Awg

No. No. No. No.

1 Awg

2 Awg 2 Awg

1 Awg I / O Awg 2/0 Awg

No. 2 Awg No. 1 Awg No. 1 Awg No. 1/0 Awg No. 2/0 No. 2/0 No. 3 / 0 No. 3/0

Awg

Awg Awg Awg

No. 2/0 Awg No. 4/0 Awg 250 M C M Na 3/0 Awg No. 4/0 Awg 300 M C M 300 MCM 400 M C M

350 MCM 400 MCM 500 MCM 600 MCM

500 M C M

600 M C M 750 M C M 750 M C M

A-C SHORT-CIRCUIT PROTECTIVE DEVICES AND CIRCUIT EQVIPMENT

183

has been proved to be valid for canductor sizes of No. 8 Awg or larger*), the conductor heating is governed by the following: For Copper:

t I

duration of current flow, see rms amperes during entire interval of current flow em conductor cross sect,ion, cir mils T I= initial copper temperature, C T 2= final copper temperature, C To simplify a n application, these relationships are presented graphically in the large chart in Fig. 3.23. The permissible time for various ternperature ranges can be quickly evaluated with the aid of the auxiliary curve B , shown in Fig. 3.23. For quick estimating purposes, minimum safe conductor size is given in Table 3.7, subject to application conditions as shown. For Aluminum. The problem of joining and terminating aluminum conductors without creating local “hot spots” deserves very careful attention. There are available, however, materials and methods which laboratory tests and experience have proved to be satisfactory. In the absence of abnormal local heating, a rough approximation of permissible current duration may he made on the basis of the same limiting temperatures as for copper. (For a particular current and conductor cross section, the permissible duration of short-circuit-current flow will he 45 per cent of that for copper.) It may be more convenient to make an artificial correction in current. Consider the current to be 150 per cent of the actual value, and proceed on the chart (Fig. 3.23) as if the conductor were copper. Rms Current. Rms current as used here is defined as the root-rneansquare value for the total interval of short-circuit-current flow. The temporary d-c component encountered in a-c circuits increases the rms current, but to a lesser extent as the interval of current flow becomes longer. The appropriate factor K , by which the symmetrical current value shall be multiplied to determine the true rms current is given in chart A , Fig. 3.23, for several typical ratios of circuit 60-cycle reactance * B. W.Jones and J. A. Seott, Short-time Current Ratings for Aircraft Wire and = = =

Cable. AIEE Technical I’sper, 1946.

184



"I.

18.5

-8.

FIG. 3.23

Short-time bhort-circuit) heating limits of copper cables and correlation of current and time to elevate the copper temperature from 75 to 150 C (dlheat is oirumed to be stored in the copped.

to resistance (distribution circuits will generally fall in thc region of X / R = 10 or less). Circuit X / R ratio is generally not known and requires numerous circuit constants for an evaluation. Conservative factors ( K , ) for the more common application conditions are

K , = 1.25 Low-voltage circuit breakers tripped instantaneously Power circuit breakers, eight cycle, instantaneously tripped K I = 1 . 1 Any industrial power-distribution problem with current duraK , = 1.0 tion of 35 sec or more Short-circuit Protective-device Interrupting Time. Circuit Breakers. The minimum time duration of short-circuit-current flow in a rircuit protected by a circuit breaker tripped by an instantaneous element will vary with the type of circuit breaker used. Typical values are shown in the lower left-hand corner of the large chart in Fig. 3.23. When interruption is purposely delayed by time-delay relays or timedelay trip coils, the duration of current flow will be governed by the timedelay relay or trip coil plus the inherent delay in the circuit breaker. Fuses (Current-limiting) , Current-limiting fuses (silver-sand and National Electrical Code low voltage) tend progressively to limit the time interval of current flow to lesser values as the magnitude of current increases. As the current magnitude increases toward the maximum interrupting ability of the fuse, the magnitude of Z't approaches a fixed value (approximately) for a particular fuse ampere rating. This is equivalent to a fixed temperature rise in a particular size of conductor. Data accumulated indicate that a fuse (of the types mentioned in this paragraph) whose ampere rating is not greater than 1.5 times the conductor continuous-current rating will protect against dangerous conductor

106


temperatures for severe overcurrents up to the maximum interrupting rating of the fuse. Table 3.8 shows the wire sizes which will have less than 75 C conductor TABLE 3.8

Silver-sand Fuse Protection at High Overcurrents Based on Copper Conductor Fuse roting,

S m a l l ~ twire normally vied,

amp

RH insulation

30 60 I00 IS0 200

No. 10 Awg No. 6 Awg No. 3 Awg No. 1/0 Awg No. 3/0 Awg

Sm.lle.t

wire

protected

No. 14 No. I2 No. 10 No. 0 No. 6

Awg Awg Awg Awg Awg

temperature rise because of the flow of short-circuit current when protected by silver-sand fuses. Fuses (Nou-current-limiting). Non-current-limiting fuses accomplish current interruption at a normal current zero, and thus the current conduction time cannot be reduced below that of the first current loop of short-circuit-(.urrent flow which may be as much as about one cycle of the power frequenry. Applications should thus recognize one cycle as the minimum time of short-circuit-current flow. Application Procedure. 1. Evaluate the symmetrical short-circuit current or currents that may be critiral. 2. Define the short-circuit protertive device clearing time at this or these current magnitudes. 3. Apply the rms correction factor to allom for the d-c component for each time interval involved. 4. Make an initial check on the current-time chart for the smallest conductor size being considered (permissible time should exceed shortcircuit protective-device interrupting time). 5 . If critical, it is advisable to rorreet for the exart temperature range (see Table 3.6 and temperature-range correction curve). F. If an oversize ronduetor is considered, but the continuous-load rursent is to remain fixed, advantage can be taken of the lower initial ronductor temperature. EXAMPLES

Example 1. A transformer feeder cable is being selected to accommodate a 1000-kva 2.4-kv transformer. The rated current of the t,ransformer (240 amp) indirates a rahle conductor of 250 MCM. The trans-


I87

former iri question is good for full short-circuit current (sixteen times normal) for 5 sec. It is desired that the feeder cable have the same ability. Solution: Rms symmetrical amperes = rated current X 16 = 240 x 16 = 3900 amp. The duration of this current as defined by the conditions of the problem is 5 sec. Assume X / R ratio = 10 or less From chart A of Fig. 3.23, K 1 = 1; ( X / R ratio of 10 and time of 5 sec) Henre, the total rms amperes affecting cable heating = K , X 3900 = 1.0 X 3900 = 3900 amp On the large rhart of Fig. 3.23, locate the intersection of the horizontal 3900-amp line and the 250-MCM conductor diagonal line. The permissible time (read on the bottom scale) is indicated to be 12 sec (75 to 150 C hasis). The 250-MCM cable will adequately meet the 5-sec requirement. Example 2. Feeder circuits are t o be run from a 480-volt 60-cycle load-center unit substation at which point the short-circuit duty is 25,000 amp (20,000 symmetrical rms amperes). What is the smallest reasonable feeder conductor size based on the use of a 25,000-amp interrupting rating air circuit breaker which trips instantarieously (1.5 cycles) a t currents in excess of fifteen times the normal rating? solulion: Symmetrical current = 20,000 amp Time duration = 1.5 cycles Rms amperes = 20,000 X 1.25 = 25,000 See preceding text for explanation of 1.25 factor K , On the large rhart of Fig. 3.23, locate the intersection of the horizontal 25,000-amp line and the vertical 1.5-cycle line. The minimum size conductor (75 to 150 C basis) whose curve is above the intersection is a KO.1 Awg. Example 3. A 4-kv feeder is t o be run from a substation at which the symmetrical short-circuit current is 25,000 amp. A continuous load caparit,y of 1000 kva is desired (113 amp), and a KO.2/0-Awg coronol cable run is being considered. Line relaying is to consist of standard time-overcurrent relays on the & tap and S o . 5 time-lever setting v i t h 250/5-amp rurrent transformers. Instantaneous attachments are not planned, but could be used if set at 3000-amp line current. Solution: Symmetriral short-circuit current = 25,000 amp Case 1. No instantaneous attachment on relay Rms symmetrical short-circuit current = 25,000 amp Relay operating time = 50 cycles; (From published time-current curves) Circuit-breaker operating time = 8 cycles

188


+

Total time = 50 8 = 58 cycles Assume X / R ratio = 15 or less From chart A of Fig. 3.23, K 1 = 1 Hence, total rms amperes affecting cable heating = KI X 25,000 = 1.0 X 25,000 = 25,000 amp On the large chart of Fig. 3.23, locate the intersection of the 25,000-amp horizontal line and t,he 58-cycle vertical line. The smallest conductor whose curve lies above this intersection is a 500 MCM. Therefore, a Xo. 2/0-Amg conductor is inadequate. Case 2. Instantaneous attachment, on relay set to operate at and above 3000 line amperes. Two point,s must, he checked: (1) a current of 3000 amp and time delay of overcurrent relay (just below the operating current of the instantaneous element) and (2) the maximum current with instantaneous relay operation. 1. From published relay data, the relay time a t 3000 line amperes is 66 cycles, circuit-breaker time is 8 cycles, making a total time of 66 8 = 74 cycles. From chart A of Fig. 3.23 for X / R ratio of 10 and time of 7 1 cycles, K I = 1. Total rms amperes affecting cable heating = KI X 3000 = 1.0 X 3000 = 3000 amp. The intersection of 3000 amp and 71 cycles on the large chart of Fig. 3.23 s h o m that a Xo, 2/0-Awg conductor is amply large to carry the 3000 amp for 74 cycles. 2. At the maximum current, instantaneous relay operation will be obtained. The total current duration will be the relay time ?,g cycle plus circuitbreaker time 8 cycles, or 836 cycles. For 8>i-cycle time interval, K , = 1.1. Total rms current affecting cable heating = K , X 25,000 = 1.1 X 25,000 = 27,500 amp. The intersection of the 27,500-amp horizontal line and the 84g-cycle vertical line on the large chart of Fig. 3.23 indicates a No. 4/0-Awg conductor (75 to 150 C basis) and shows that point 2 cont,rols the cable size. However, a No. 4/0-Awg conduct,or mould operate at less than rated temperature. A specific check may show that a KO. 3/0-Awg conductor is adequate. Rated conductor temperature coronol cable = 80 C (see Table 3.6), ambient temperature = 40 C. Xormal temperature rise produced by rated current = 80 - 40 = 40 C. Rated continuous current for No. 3/0-Awg coronol cable = 185 amp. The temperature rise will be roughly proportional to the square of the current.

+

r

I

A-C SHORTT-CIRCUITPROTECTIVE DEVICES AND CIRCUIT EQUIPMENT

T i m e - seconds

189

190


Hence, the normal conductor temperature of a No. 3/0-Awg conductor operating a t 143 amp would he expected to be

(g)’

(full-load rise)

+ ambient

=

=

63.8 C, or 64 C

The maximum momentary temperature for coronol at 5 kv is 145 C (see Table 3.6). From detail chart B , Fig. 3.23, the correction factor K for an initial conductor temperature of 64 C and final of 145 C is K = 1.13. From the large chart of Fig. 3.23, the permissible time for 27,500 amp in No. 3/0-Awg conductor (75 to 150 C basis) is 6.7 cycles. The permissible time corrected t o a 64 to 145 C basis is K X 6.7 = 1.13 X 6.7 = 7.6 cycles. Therefore, a No. 4/O-Awg conductor is the correct selection since a No. 3/O-Awg conductor would fail t o meet the 8.5-cycle requirement. FUSING CURRENT TIME FOR COPPER CONDUCTORS

The fusing current time curves for copper conductors are shown in Fig. 3.21. The curves are based on the folloiving assumptions: 1. Radiatiou may be neglected because of the short time involved. 2. Resistance of 1 cu cm of copper at 0 C is 1.589 microhms. 3. Temperature-resistance coefficient of copper a t 0 C is 1/234. 4. Melting point of copper is 1083 C. 5. Ambient temperature is 40 C. Data are an adaptation from the eight,h edition of “Standard Handbook for Elect,rical Engineers.”* * A . E. Knowlton (editor-in-chief), “Standard Handhook for Electrical Engineers,” 8th ed., Chap. 4, McGraw-Hill Book Company, h e . , S e w York, 1949.

Chapter 4

by W. R. Crites and Maynord N. Halberg*

Voltage-Standard Ratings, A llowable Variations, Reduction of Variations, Calculation of Drops The purpose of any industrial power system is to maintain voltage a t the terminals of power-using equipment. This voltage should bewithin acceptable limits-equal to the rated voltage of this equipment. The standard voltage ratings for utilization equipment are discussed in this chapter, along with the standard voltage ratings for power generation and distribution equipment. KOpractical power system can maintain voltage a t rated value a t the utilization equipment a t all times. The voltage variations allowable and the methods which can be used in the design of a power system to keep the variations within acceptable limits are discussed. I t is necessary to calculate the voltage drop in the power system for steady-state conditions and during the starting of the larger motors to determine whether or not the voltage mill remain within acceptable limits. Methods of calculating these voltage drops are presented in this chapter.

* The following men, formerly in Industrial Pawcr Enginwring. General Electric Company, made substantial contributions to the material in this chapter: W. K. Boice, General Electric Company, l e w Haven. Conn.; D. F. Capehart. General Electric Company, Cincinnati, Ohio; J. R. Eliason, General Electric Company, Sehenectady, N.Y. 191

192

V O L T A G F S T A N D A R D RATINGS, VARIATIONS, CALCULATION OF DROPS

VOLTAGE DESIGNATIONS *

It is necessary t o have a n understanding of the voltage names of systems and t,he voltage rat,ings of various pieces of apparatus used in the system before start,ing a discussion on system-voltage problems so t h a t the proper voltage identification can be used throughout. It is also necessary t,o know v h y the voltage designat,ions are applied t o help in understairding the system-voltage disussion in the following sections. The volt,age-identification structure is summarized iu Table 4.1. For each of the nominal syst,em voltages listed, t,he table gives voltage ratings of generators, transformers, motors, and (in some cases) lamps. T o illustrate the use of Table 4.1, consider a 13,800-volt system. The generators would be rated 13,800 volts. Transformers stepping power down from transmission voltage would have secondary windings (I?, Fig. 4.1) rated 13,800 volts. Transformers steppiug power down t o utilization vokage in load-center substations would have primary mindings (C, Fig. 4.1) rated 13,800 volts. Motors connerted directly to the 13,800-volt bus would lie rated 13,200 volts. From the foregoing summary and Table 4.1 it is evident that care must tie exercised in using the proper voltage ident,ifiration for each piece of equipmelit as well as for the system. Some fundamental rules are as follo\vs : 1. When speaking of equipment, the rated voltage is used, aud it is the voltage to which the operating characteristics are referred. 2. When speaking of systems, rat.ed voltage is not an applicable term because various piwes of equipment in a given system often have different voltage ratings. Therefore, t,he term n o m i n a l s y s t e m vollage is used for convenient designation of systems and circuits t o define the voltage class. The problem of proper identification would be easier if all apparatus of a given voltage class had the same vokage rating. Then, of course, tem voltage could have that same value. Possibly if the industry were starting over again, vokage ident,iticatioii mould be made that simple. But, as syst,ems grew, voltages were ini,hed up t o compensate for t,he voltage drop between source arid load. As a result, of t,hese changes that have taken pla(.e over a period of years, transformer arid generator voltage rat,ings are generally higher than utilization-eiiuipment vnltagc rat,ings. There is logic in this in that the voltage rating of transformers, for example, is t,heir no-load rating. Since most plants are supplied by transformers, the concept has beeri acceptcd that, supply equipment will have a higher voltage rating than utilization equipment,. This means that in a 480-volt system, for cxam* For a iiirthrr rrpansion of t h i s srihjpet F W l < I ~ ~ I - X 1 5 MKPport., A l’refrrrrd Voltage I h t i n g s of :\(: Systrrris and Equipmmt, N I X l’uhliration lo. R-6. S E M A I’ulilirstion l o . 117, \lay, 1‘JIU.

VOLTAGE-STANDARD

RATINGS. VARIATIONS, CALCULATION OF DROPS

193.

ple, transforniers or geiierators supplying motors ivoiild have a ratiiig of 480 volts whereas t,he motors irould have a ratiiig of 440 volts. Part of this ditrereiicc: is compeiisated for by voltage drop iii the traiisformer aiid in the distributioii system betiveeii the traiisformers aiid the motors. Therefore, in general, the voltage at the motors is reasoiiably iiear thc iiame-platc ratiiig iii the average system. I n older types of distrihiitioii systems it i m s commoii prartire to use step-doivii trmçformers irith a Iower primary voltage ratiiig thaii thc transformers which ivould siipply that systcm. For example, the ti'aiisformer steppiiig dowi from the iitility voltage ofteii hnil a ratiiig of 2400 volts oii the secoiidary, aiid the traiisformer steppiiig doi\-ii to the utilizatioii voltage of 480 or 240 volts had a ratiiig of 2300 volts oii tlie primar?.. Becausc of the desigii of preseiit-day systems n-itli smaller drgi'ers of volt,age drop, aiid judirioiis m e of taps i i i traiisformers, the prartirc is, as INCOMING

4

MASTER U N I T SUBSTATION ( P R I M A R Y SUBSTATIONI

\

1

( A I P R I M A R I WINDING

u

(IF USEDI

m l SECONDARY

WINDING

X P L A N T P R I M A R Y D I S T R I B U T I D N VDLTAGE

LOAD C E N T E R U N I T SUBSTATION (SECONDARY SUBSTATION IN FACTORYI PRIMARY WINDING WINDING

FIG. 4.1

Typicol industrial plont power ryrtern

VOLTAGE-STANDARD RATINGS, VARIATIONS, CALCULATION OF DROPS

194

evident from Table 4.1, t o use the same voltage rating for all traiisformer windings connected t o a given system voltage. This is true whether the transformers are stepping down to this system or steppiug down from this system. TABLE 4.1 No min0 I

Genordor

system

rated

*olt.ge

voltage

Boric Pattern of Voltage Identification Transformer secondory rated voltage

Transformer primary rated voltage

120 or 120/240 I20 or 120/240 I20 or 120/240 240 or 120/240 240 or 120/240 240 or 120/240 208Y/120 208Y/120 208Y/120

120 240 I20

Motor and control rated rottoget

230 115

Three-phase Systems

208Y/120' 240 480* 600 2,400' 4.160' 4,800 6,900* 12,000 13,200 13.800' 23,000 34,500 46,000 69,000 1 1 5,000

20sY/l20 240 480 600 2.400 4,160 4,800 6,900 12.500 13.800 13.800

........ ........ ........

208Y/120 240 480 600 2,400 4,160 4.800 6,900 12,000 13,200 13,800

........ ........ .... .... .......

208 or 120 240 480

600 2,400 4,160 4.800 6,900 12,000 13.200 13.800 22,900 34,400 43,800 67,000

L.mp rated YoltDge

118or120

I

220 or 208 208.118. or 120 236 220 440 165 2,300 4,000 4,600 6,600 11,000 13,200 13.200

I I0,OOO

* In ~ P I Vinstallations, or W ~ P ~ P Ya srlwtion P ~ oi voltngr can l i p ~ n a d rthrsr . i ~ r cprcferrrd s y s t m valtagrs. t Specifying t h e w valiirs for motor voltsgcs is itnportarrt: For instnnw. motors to opprste on -IltiO-. GWC-, or 18,800-volt systrins should Iw rntcil 4000. (i(iO0. or 1:1,200 volts, resp2ctively. The one-line diagram (Fig. 4.1) shows a t y p i i d method of distributing power in industrial plants and will be used as referenre to identify some portions of the systems and equipment referred to. RATED VOLTAGES OF TRANSFORMERS

Transformer voltage ratings are hased on the no-load values, and the ratio of primary to secondary rated wltages is equal t o the turn ratio. The transformers have a voltage rating for each xindiiig. These are

VOLTAGSSTANDARD RATINGS, VARIATIONS, CALCULATION OF DROPS

195

the voltages a t which characteristics are measured. What then are standard transformer voltage ratings for industrial plants? First, consider primary or master unit substations and transformers which step down from some voltage above 15 kv to plant primary distribution voltage, which is generally below 15 kv (see Fig. 4.1, top substation). The standard primary-winding ( A , Fig. 4.1) voltage ratings of this class of substation and transformers are 110, 67, 43.8, 34.4, and 22.9 kv. These are the actual transformer-minding ratings. They are derived from the old rating structure based on secondary ratings in multiples of 115 volts. When secondary ratings were boosted to multiples of 120 volts, the high side rating was raised to maintain the same turn ratio. For instance, 33,000-2300 volts was once a standard rating. Thc corresponding present-day transformer would be rated 34,400-2400. The familiar designations 115, 69, 46, 34.5, and 23 kv refer to the classes of insulation used with these transformers. Secondary-winding ( B , Fig. 4.1) voltage ratings of this class of industrial substat,ions and transformers are 13.8, 13.2, 12, 6.9, 4.8, 4.16, and 2.4 k v . S e x t consider transformers in load-center unit substations (see Fig. 4.1, bottom substation) used in t,he industrial plants for stepping down from plant primary distribution voltage to utilization voltage. As stat,ed above, the plant, primary voltage is usually less than 15 kv. Therefore, the list belox includes only voltages below 15 kv. The primary-winding (C, Fig. 4.1) voltage ratings of load-center unit substations are 13.8, 13.2, 12, 6.9, 4.8, 4.16, and 2.4 kv. . Note that the primary voltage rating of this class of transformers (bottom, Fig. 4.1) i s the same as the secondary voltage rat,ing of the primary substation transformers (top, Fig. 4.1). The voltage ratings of secondary substations in the plant which supply motors and other utilization equipment are divided into two classesthose for serving utilization equipment above 600 volts and those for serving utilization equipment below BOO volts. Standard rat,ings are listed in Table 4.2. TABLE 4.2

Transformer Secondary Voltage Ratings ( I ) , Fig. 4.1) Supplying Utilizoti0n Equipment Roted

Above 600 Volts, Kv 6.9 4.8 4.16 2.4

Supplying Utilization Equipment Roled 600 Volt, or 0e1ox. Volt. 600 IY or delta1 400 IY or delta1

240 208Y/l20

All standard unit substation transformers have taps in the primery winding to allow compensation for voltages that vary from the transformer rating. The most common are four 255 per cent taps, two above

196

VOLTAGkSTANDARD RATINGS, VARIATIONS. CALCULATION OF DROPS

aiid two below normal, giving a total adjustment of plus or minus 5 per cent,. With these t,aps in the primary winding, a transformer actually has five different ratios. I t vould he very cumbersome to refer to all five of these ratios in all discussions; therefore, when in the following discussion a transformer is referred t o as having, for example, a rating of 2400-480 volt,s, the discussion will apply equally well whether the transformer is operated 00 the cenher t a p or other taps. Regardless of the tap used, the t,raiisformerwill still be referred to as a 2100-480-volt transformer, Comhined light arid power systems are frequently used where motors are supplied a t 180 volts, for example, and lights are supplied at 120 volts from the same 480-volt system, using dry-type transformers. The standard primary volt,age ratings for t,hese light,ing transformers are 600 volts, 480 volts, arid 240 volts, aiid the standard secondary vohage ratings are 208Y/120 volts and 120/240 volts. Two rated kva 5 per cent below normal t,aps are provided in these transformers t,o allow for operation of 120-volt lamps near t,heir rated voltage when the voltage on the 480-volt system is below 480 volts as it normally vill be.

TRANSFORMER VOLTAGE REPRESENTATIONS

Transformer voltage designations become rather complex. For illstance, windings may have series-parallel connections. Or they may be designed for connection line-to-neutral on higher rated volt,age systems, such as 3400-volt transformers which are suitable for line-toneutral operation OIL 4160-volt systems. These and other complex arrangements make exact identification desirable. These variables in t,ratisformer voltage ratings have long been expressed by various symbolic met,hods. Such methods are essential because t o fully describe the \\-indings of transformers often would require a fairly lengthy paragraph. However, t o bc of any value a transformer rating so expressed should meao the same to everyone. To further a consistent use of symbols, hot,h KERIA and ASh standards have been established t,o rci~ommenda standard transformer “shorthand.” Four symbols are used: the dash (-), t,he slant (/), the X, and the Y. In general terms, their uses are as follows: Dash (-). Used to separate the voltage ratings of separate windings in a specific transformer. Slant ( I ) . Used t o separate voltages t o be applied to or obtained from the same windiug. X. Used to designate separate vokagcs obtainable by reconnection of the coils of a winding in series or multiple combinations. Y. t!sed t o designat,e a winding t,hat is Y-connected. The absence of


I97

this symbol in a three-phase transformer rating indicates that the winding is delta-connected. The use of the dash, slant, and Y can he easily illustrated by the voltage rating of the transformer for a typical load-center suhstation. 4160-480Y/277: Note that this meaus the 4160-volt high-voltage winding is delta-corinected while the 480-volt winding is Y-couiiected with t,he neutral brought out. A three-winding t,ransformer might have this voltage rating: 13,800-2400-480Y/277. In three-phase transformers the slaut is ofteri used to indicate wiiidiiigs connectable either in delta or Y. For iiistauce, a 2400/4lCiOY windiug can be couiiected either for 2400 volts deka or 4160 volts Y. Xote that the delta voltage is expressed first. When a Y-connected winding has the neutral brought out it is siguified like this: 2OSY/lZO. Here the line voltage is expressed first, fol1oir.d by the line-to-neutral voltage. If the neutral is brought out with reduced insulation, that fact is shoivu by 208 Grd Y/120. Another use of the slant is to indicate taps, especially 011 single-phase transformers. For instance, a 240-volt wiuditig with a midtap is expressed 240/120. When a single-phase t,ransformer with a series-multiple winding is vound to be suitable for three-wire service on the series conoectioii, it is designat,ed 120/240. When a winding has several taps close to the rated volt,age, it is cust,omary to specify them as illustrated in t,his specific case: four 255 per ceut rated kva taps, t x o above and tI5-o below rated voltage. The X symbol is used to separate t,he volt,ages obtainable in a seriesmukiple minding not, suitable for three-wire operation. For example, a minding rated 120 X 240 can be connected with t,he coils in parallel to obt,ain 120 volts or Tr.it,h the coils iu series for 240 rolt,s. RATED VOLTAGES OF GENERATORS

Siiice the generator is a source of elect,ric poir-er aud is ofteu i u parallel wit,h primary substation transformers (see Fig. 4,1), its voltage aud ('oiiscquently its rat,itig is in practically all cases the same as the transformer in a giveu voltage class. Listed in Table 4.3 are the three-phase generator ratings that, are recommended by the latest EEI-SE5I.i report. TABLE 4.3 208Y/120 "Olt. 240 volts 480 volts 600 volts

* Ratings

Generator Voltage Ratings* 2,400 volts 4,160 volts 4.800 volts 6,900 volts

13.800 "011. 14,400 volts

of 11,500 and 12,500 volts are n s ~ dfor genrrators on smnr rstablislird hut are, not rrrommmdrd for nmv systim~s. Thc corrcsponrling trnnsfornii,r rating is 12,000 w i t s and transformcr taps sllon for paralirl oprration. systems

198


The 14,400-volt rating has been adopted largely in large generating stations where the input is transformed up to higher voltage in a unit transformer generator arrangement (see Fig. 4.2).

2

mTwI

4 I

FIG. 4.2 HIGH VOLTAGE BUS

Unit transformer generator arronge-

merit,

RATED VOLTAGES OF MOTORS

A t the other end of the system are the motors, and their rat,ings reflect the fact that voltage at utilizatioii equipment is somewhat loirer t,haii a t the sources of power because of voltage drop. Single-phase motors are usually rated at 115 or 230 volts. The standard voltage rat,ings of polyphase motors are given in Table 4.1. TABLE 4.4

M o t o r Voltage Ratings

110 "0111

550 "011.

6,600 Volt.

208 volt. 220 wit.

2,300 ~011s 4,000 ~ o l t i 4,600 volts

I1.000 volt,

440 rolls

13,200 volts

hlot,or-cotit,rol equipment has the same voltage rating as the associated motor. RATED VOLTAGES OF LAMPS

Inrandescent lamps are standardized at 120 volts. Higher voltages have not in general heeo found sat,isfactory. Fluorescent lamps offer a wider range of operation and are commotily rat,rd a t 118, 208, 230, and 265 volt,s (for line-t,o-neut,ral on 480-volt systems). OTHER APPARATUS

Some other types of equipment such as capacitors and industrial heating equipment have compromised between the extremes of generator


I99

rating and motor rating in a given voltage class. For instance, industrial heating devices are rated at 115,230, 4G0, and 575 volts. Capacitors are rated at 230, 460, 575, 2400, 4800, 7200, 12,470, and 13,800 volts. NOMNAL SYSTEM VOLTAGES

The choice of the numerical value t o represent nominal system voltage is purely arhitrary and does not attempt to indicate an average system voltage. It is merely a name. However, it is very desirable that a consistent practire in designating nominal voltages be followed. When used properly, the nominal voltage should give a good picture of the voltage struct,ure of a system with a minimum of misunderstandings. The standard values for nominal system voltage correspond t,o the ratings of source equipment. TABLE 4.5

Standard Nominal System Voltages Singlo Phase

120 120/240 240 Three Phore

208Y/l20

240 480 600 2,400 4,160

4,800 6,900 12,000

13,200

34,500

46,COO 69.000 115,000

13,800 23,000

Table 4.5 is not complete but is representative of industrial practice. To repeat, it is extremely important to identify properly the voltage rating of each piece of apparatus in a system as well as to identify the nominal system voltage. The voltage ratings of the various pieces of apparatus, as ran he seen from the foregoing, may be different even though the apparatus is for use on the same given voltage class system. Therefore, correct identification of each piece is of paramount importance. For example, if one is buying equipment to supply a 180-volt system, the secondaries of t,he transformers should he specified as -180-volt rating. The motors and control should be specified as 440-volt rating. The system nominal voltage is referred to as 480 volts. Other apparatus on this system may have different voltage ratings. For example, capacitors would be rated 460 volts; heating equipment would be rated 460 volts. It is also important to remember that transformer and generator voltage

200

VOLTAGE- STANDARD RATINGS, VARIATIONS, CALCULATION OF DROPS

ratings are always higher than utilization-device ratings. This is logical because the transformer voltage ratings are the no-load voltage ratings, and as load is applied to the system the voltage drops to near the nameplate rating of the lower rated utilization apparatus.

VOLTAGE SPREAD AND FLICKER REQUIREMENTS* STEADY-STATE VOLTAGE REQUIREMENTS

An ideal electric power system is one which will supply constant frequency and volt,age at rated name-plate value t o every piece of apparatus in the system. I n modern power systems, frequency is a minor problem. It is impractical, however, t o design a power system which will deliver absolutely constant rated name-plate voltage to every piece of apparatus. Since this cannot he attained, what are the proper limits of voltage variation in a n industrial plant? These should be determined by the characteristics of the utilization apparatus. First, certain definitions are essential to underst,arid clearly the discussion of this problem. Voltage Spread. Voltage spread is the difference between the maximum and minimum voltages which appear at any location in a system under riormal operating conditions. Voltage spread is not intended to cover momentary voltage changes uf a transitory nat,ure such as those due t o switching surges, motor starting, welders, etc. The first part of this discussion is primarily concerned with voltage spread a t utiliaatiori equipment. This is the diKercnce between the maximum and minimum voltages a t the terminals of the utilization equipment under normal system operating conditions (Fig. 4.3). Maximum values usually appear during light load and minimum values a t full load on the electric system. Another important type of voltage spread is primary or supply voltage spread which is the difference between the maximum and the niinimum voltage a t the service entrance or plant primary bus of a particular plant under normal operating conditions. Voltage Zone. Voltage zone is the envelope of all voltage spreads for a particular voltage class of system. For any specific voltage class designated by a nominal system voltage there inherently exists an appreciable range of operat,ing voltages between the systems having the highest and lowest voltages for this class. Countrywide, this zoue is larger thaii the voltage spread at, ariy one location because of recognized differences in practices of different companies.

* The data in this sretion arc l a r ~ c l yadapted from an AIEE Industrial Power System Coinmittre 1Lpurt. Industrid Voltag- Ilrquirpmeats, Elec. Eng., vol. 6 i , 1948, pp. 358-374.

3.3 7.

z

5 ,

PRIMARY SYSTEM

LONGEST SECONDARY FEEDER

_ _ _ _ _ _ _ _ _ _ _ _NO_ LOAD _ _ _VOLTAGE _~ _ - - - - - - - - - - - - - - 480 2500-PRIMARY VOLTAGE SPREAD. NO LOAD TO F U L L LOAO AT PLANT SERVICE ENTRANCE

:z:

s

r

2400-

-

0

0 Y

Y

> E

2 N 0

1

SPREAD IN SECONDARY SYSTEM

v) Y

k9 2

E >

TRANSFORMER VOLTAGE DROP

Q

2300-

(L

P

FEEDER VOLTAGE DROP

T. Y < 0

NO LOAD VOLTAGE

5

> 0 PRIMARY VOLTAGE SPREIO, NO LOAD TO F U L L LOAD AT PLANT SERVICE LNTRANCE

> 2200-

*

I

I

E

L

I

0

MINIMUM FULL LOAD VOLTAGE

TRANSFORMER VOLTAEE DROP

C U R V E A - T R A N S F O R M E R OPERATING ON HIGHEST TAPRATIO 2 5 2 0 - 4 8 0 VOLTS AT NO LOAD. CURVE 8

TRANSFORMER OPERATING O N LOWEST TAPRATIO 2260-480 VOLTS AT NO LOAD.

FIG. 4.3

Examples of voltage zone, spread, and drop.

VOLTS

VOLTAGFSTANDARD RATINGS, VARIATIONS, CALCULATION OF DROPS

202

difference in voltage in various parts of the power system. The other cause is primary voltage spread a t the service entrance of the plant.

EFFECT O F VOLTAGE DROP

To show the effect of voltage drop in a plant it will be assumed that the primary voltage is maintained a t a constant value regardless of plant load. The simple circuit shown in Fig. 4.4 will be used as an illustration. The primary voltage is assumed to be of such magnitude that the secondary voltage on the transformer is 480 volts a t no load. Referring to Fig. 4.5, a t extremely light load there is essentially no voltage drop through the transformer or in any of the secondary circuits connected to the transformer. Consequently, the voltage is substantially the same throughout the plant, and any lights or other incidental load connected a t this time is subject to practically the no-load voltage. It is particularly significant a t this point to recognize that transformer voltage ratings are the no-load SECONDARY BUS

TRANSFORMER CIRCUIT

FIG. 4.4

480

.-

Typical industrial plant power circuit,

400 VOLTS

ZERO VOLTAGE DROP A

2

2

470-

:rp 4 6 0 ~

_ _ _ _ _ ~. TRANS FA NO

y)

3

9

]----____________________

----

460

TOTAL VOLTAGE

450

FIG. 4.6 spread.

l,z

LOAD VOLTAGE-480 VOLTS VOLTAGE DROP VOLTAGE DROP IN THRU 15 VOLTS TRANSFORMERSECONDARV FEEDER-IOVOLTS

sE~!~48oro~?p~"2~Ts G?!E?

IN BRANCH DROP CIRCUIT-

--- -___

_________________-__

Full-load voltage conditions for circuit shown in Fig. 4.4.

5 VOLTS

_

A

No primary voltage


203

ratios. For example, a transformer rated 4160-450 volts will produce 480 volts a t no load with 4160 volts applied to the primary. When load is connected to the transformer, current flows, and this causes a voltage drop in the secondary circuits as shown in Fig. 4.6. At t,he secondary bus the voltage drop caused by the current flowing through the transformer is assumed to be 15 volts. With constant primary voltage the secondary bus voltage varies from 450 volts a t no load to 465 voks at full load--the voltage spread a t this point is 15 volts. There are assumed additional drops of 10 volts in the secondary feeder and 5 volts in the branch circuit, making a total drop to load A of 30 volts. If the lowest voltage in the plant exists a t load A , then the maximum voltage spread is 30 volts (450 a t no load to 450 volts a t full load, or 30 volts). In designing an industrial power system the voltage spread should be kept t o a minimum consistent with reasonable first cost. If the spread is too great,, the voltage may be too high a t light load, causing equipment operating during that period to burn out, or voltage may he too low a t full load a t much of the utilization apparatus, impairing the performance and reducing the production obtained from the equipment, The second cause of voltage spread is the primary voltage spread a t the plant service connection. This may be caused by voltage drop in the primary system, or it may be due to regulation of the primary system by voltage regulators. To show the effect of primary voltage variation, assume that the primary voltage drops as load comes on in the plant. The transformer taps have been selected so that the no-load voltage is 450 volts as in Fig. 4.5. When load comes on the power syst,em,the same voltage drop occurs as in Fig. 4.6, but in addition, the primary system voltage is assumed t,o drop sufficiently to cause an additional 10-volt drop in the vokage at the secondary of the transformer. This primary voltage spread adds to the total voltage spread in the plant, making the spread 480 to 440 volts or a total of 40 volts as is shown in Fig. 4.7 instead of only 30 volts as shown in Fig. 4.8 where there was no primary voltage variation. The primary voltage spread may not always be in the direction shown in Fig. 4.7. The primary voltage may rise when the load comes on because of voltage regulators in the primary feeder circuit or because of other voltage regulators in the primary power system. This voltage rise of the primary reduces the voltage spread in the plant, as shown in Fig. 4.5. Very weak primary systems with a high drop or regulated primary systems whose load cycle does not coincide with the load cycle of the plant may cause excessive voltage spread in the plant-beyond the limits shown in Table 4.9. This is illustrated in Fig. 4.9. Automatic voltage regulation is required in such cases to bring the voltage spread within the limits shown in Table 4.9. Changing transformer taps to increase the vo1t:ige a t full load will not solve the problem because that will increase the no-load voltage beyond 450 volts.

VOLTAGLSTANDARD RATINGS, VARIATIONS, CALCULATION OF DROPS

204

480-

----

470-

VOLTAGE DROP THRU TRANSFORMER 15 VOLTS

y1

9 450

VOLTAGE DROP IN SEOWDARI FEEDER VOLTAGE DROP IN

-

TOTAL VOLTAGE SPREAD 480 TO 440 VOLTS 140 VOLTS1

_________________________________

440

I

-______

---

FIG. 4.7 Full-load voltage conditions for circuit shown in Fig, 4.4 with 10 volts (on 480volt baris) primary voltage spread. Primary voltage varies from maximum at no load to minimum a t full load.

VOLTAGE DROP I N SECONDARY FEEDER10 VOLTS

FIG. 4.8 Full-load voltoge condition3 for circuit shown in Fig. 4.4 with 10 volt. (on 480volt basis) primary voltage spread. Primory voltage varier from minimum at no load to maximum at full load.

_____

470

;1

LOAD VOLTAGE

4SO

PRIMARY VOLTAGE SPREAD

440

___---

430

VOLTAGE DROP THRU TRANSFORMER

420

-

480 VOLTS

~

460;

G

_________NO

2s__vw3

. 410 . J

TOTAL VOLTAGE SPREAD 4 8 0 TO 410 VOLTS 170 VOLTS)

- 40

VOLTS

VOLTAGE DROP IN SECONDARY FEEDER lo

VOLTAGE DROP

V O L T A G b S T A N D A R D RATINGS, VARIATIONS, CALCULATION OF DROPS

205

EFFECT OF VOLTAGE SPREAD O N UTlLlZATlON EQUIPMEN?

G e n e r a l Effects. Whenever the voltage a t the terminals of a utilization device varies from name-plate rating of the de\.ice, something is sacrificed either in life or performanre of t,he equipment. The effert,may be minor or serious, depending upon the chararteristirs of the device, how the device is applied, and the amount the voltage deviates from the device rating. KESIA Standards provide for rert,ain tolerances whirh may he taken advantage of without seriously affertiiig the performanre of the apparatus. However, with usbge of electrir pover for precise operations, there is often a major sacrifire in produrtion for volt,age variations of considerably less than given in t,he NERlA Standards. So that the plant engineer can better judge the effect 11f vokage variation on t,he electric equipment in his plant, the rharacteristics of many commonly used derires are given here. I t is these rhararteristirs rvhirh have been used as a st,arting point for establishing the desired voltage spread of Tables 4.8 and 4.9. Effect on Induction Motors. Induction motors are the most rommoir utilization derires in industrial plants. Thr variatioii i n rharactrristiw as a function of voltage for the widely used inductiotr motors is shoivn i n Table 4.G. The material in this section deals only n-ith the cffert 011 motor chararterist,ies of rhaiiges in voltage magnitude. The effect, of unbalanced voltages is also very importatit and shonld he rotrsiderrd. The rurrent may hecomc esressive for only a small voltage iuihalanre. The XEBIA St,andards should be consulted for detailed information on this subject,. Principal Effects of l o w Voltage on Induction Motors. The most significant effects of too lox voltage are reduction in starting torque a t i d increased full-load t,emperature rise. The redurtion of st,arting torque may be significant i n mot,or applications driving high-inertia rqnipmeirt. The lower torqne i d 1 result, in longer armleration periods. Torque mot,ors are also very materially affected hy redured voltage as thi. torque decreases as the square of the voltage; thus a t 10 per reut helow normal voltage, the torque is redured 19 per cent. The increased heating at low voltage aiid full load rediirrs thr lifr of the insulat,ion. Principal Effects of High Voltage on Induction Motors. The most, significant efferts of too high voltage are inrreased tnr(lue, inr,rrasrd starting rurrent, and decreased p o r e r factor. The increased torque may muse rouplings to shear off or damage t o driven equipment. Increased starting curretit raiiscs greater voltage drop in the power system, henre increases light, flirker. Uecreased p o ~ v z r factor is particularly disadvantageous where power-fartor peualty rlanses

206

VOLTAGE-STANDARD RATiNGS, VARIATIONS, CALCULATION OF DROPS

TABLE 4.6

General Effect of Voltage Variation on Induction-motor Characteristics

I

Voltage Variotion

90% voltage

Starting and maximum running torque... Synchronous speed.. Per cent d i p . . Full-load speed.

................. Decrease 19%

Functionof voltage

(Voltage)’

.......... No change Cons1.nt ............... Increase 23% 1 (voltagel~ .............. Decreore 136% ISyn. speod--.llpl Efficiency: Full load.. ................ Decrease 2 points .............. 9% load. .................. Proclicolly no change .............. )i l o a d . . ................. Increase 1 to 2 point$ .............. Power faclor; Full land.. Increase 1 point load.. Increase 2 lo 3 point! load. Incrcoie 4 lo 5 points Full-load ~ u r r e n t . Increase I1 Yo Starting w r r e n l . . Decrease 10 to 12% Voitoge Temperature rise, full load. Increose 6 to 7 C Maximum torque capocity.. IV0ltogeJ~ Decrease 19% Magnetic n0ire.m load in parliculor.. Decrease slightly

I( 36

................

................. .................. ............. ............. ..... .... .....................

110% voltage

InCreOle 21 N o change

Decrease 17%

Increase 1 % Small increo*e Procticdiy no change Decrease 1 to 2 points

.............. Decrease 3 points .............. Decrease 4 points .............. Decrease 5 lo 6 points .............. Decrease7% Inc,eo.e 10 to 12% .............. Decrease I lo 2 C

..............

Increa3e 21 %

Increase slightly

This table s h o w gencral effcets, which will vary somewhat for specific ratings.

are applied by the utilities. The higher the motor voltage rises, the lower the power fartor mill become. This may result in a greater penalty and hence a higher power bill. While the temperature rise at full load on standard motors decreases slightly for moderate overvoltages, the temperature rise may increase on certain types of sperial motors a t even very small overvoltages. Overvoltages of 10 t o 1.5 per cent have caused numerous burnouts on special four-speed grinder motors. Motors rated for intermittent load are also materially affected by overvoltagcs. While marry drive applications are not seriously affected by voltage deviations as much as plus or minus 10 per cent from rated voltage, there are import,ant applications that are. Effect on Synchronous Motors. The effect of voltage variation on the performance of synchronous motors is similar t o that on induction motors. However, while t,he starting torque varies as the square of the voltage, the maximum or pull-out torque varies directly with the voltage. From the above discussions it will be noted that, in general, voltages slightly in excess of motor name-plate rating have less detrimental effect

V O L T A G k S T A N D A R D RATINGS, VARIATIONS, CALCULATION OF DROPS

207

on motor performance than voltage helow the name-plate rating. This is one of the bases on which the voltage spreads in Table 4.9 mere determined. A s a n example, the figures show a recommended spread of 420 t o 180 volts for the 480-volt nominal system voltage, which is approximately 4 per cent below and 9 per cent above the 440-volt motor rating. Effect on Incandescent lamps. The light output and life of incandescent filament lamps are critically affected by the impressed voltage. I n Table 4.7 is shown the relationship of lamp life arid output t o voltage for a vokage range from 80 t o 120 per cent of rated voltage. I n general it may be said that for incandescent filament lamps a 1 per cent deviation from rated voltage causes a change of 3 t o 335 per cent in light output. It can be seen from Table 4.7 that a 10 per cent reduction in lamp voltage results in a 30 per cent reduction in light output. In other words, when the voltage is 10 per cent low, the investment in the lighting system is working at only 70 per cent efficiency-thus, 30 per

! i! 3 a

2

0 c

3

PER CENT NORMAL VOLTS

9 a FIG. 4.10 Characteristics of large gar-filled incandescent type C lampr. average of many lampr.

There are the

208

V O L T A G b S T A N D A R D RATINGS, VARIATIONS, CALCULATION OF DROPS

cent of the investment is lost. With an overvoltage of 10 per cent the lamp-life is reduced t o less than oue-third~-t,hus lamp-replacement costs are three times as great as a t normal voltage. Other dat,a arc shown in Fig. 4.10, from which it, should be noted that the lumens per watt., or lamp efficieilcy, rises sharply at voltages above 100 per cent. I n some cases, operating eronnmies result from hurriing lamps at higher efficiency and short life, or vice versa. TABLE 4.7

Effect of Voltage Variations on Gar-filled Incandescent-lamp Choracteristics

Socket voltage

Per cent rated voltage

Per cent rated lighl output

Per
~

96.0 102.0 108.0 110.4 112.8

80 85 90 92 94

47 58 70 75 81

1900 850 400 300 225

115.2 117.6 120.0 122.4 124.8

96 98 100 102 104

87 93 I00 105 115

170 130 100 75 60

127.2 129.6 132.0 138.0 144.0

106 108 110

I20 I30

45 35

I40

115

I60 185

30 15 10

I20

~* Throrrticnl lifv in thc nhsrrrcc of any mcrhanicnl hrcakagc.. In onlinary sprvire, mcchanird hrrakage r c d n r r s t h p liip expectanrc a t tlrr lo\ver roltagrs. Effect on Fluorescent l a m p s . The changes in lamp characteristics iI-ith rariatioii in cirruit, voltage arc given in Fig. 4.1 1. IIIgeneral, 1 per cent variatiim i n line voltage n-ill changc t,he lumeir oudput only about 1 per cent. Toltage is a factor in starting reliahility, and voltages l o w r than recommeiided may result in unsatisfactory starting. It will be noted that the ores-all efficiency (if the fluoresrerrt, lamp decreases if the line volt,age is raised above normal. The increased line volt,age causes the choke t,o pass more current to the lamp. This loivers the resistance of the arc. column, rcsulting in a lower voltage drop i n the lamp itself. The input, Ti-atts t o the lamp are slightly increased, and t,herefore the lumen output increases over a cert,aiii range. In this condition, however, the higher currcnt density priiduces the short ultraviolet radiation less effirieutly; wilserpently t,he luminous efficiency of the lamp decreases.

VOLTAGGSTANDARD RATINGS, VARIATIONS, CALCULATION OF DROPS

209

Fluorcsreiit lamps are f a r less af'ectrd hy circuit voltage variatioli tllan filament lamps, from t,he standpoint of life. The life of preheat-type lamps should he quite satisfartmy throrlghollt, the range of published voltage fur the various Iiallasts; tlicsr volt:tg(. ranges, iii general, are 110 ti] 125, 100 t o 2l(i, 220 to 250. :ind 240 t o 280. There may be some derrease i n life performalire \\-3ir11 o p w i t c d a t maximum vokage as compared with that, at miiiimiim vdt:igr. I I ~ i ~ e v t ~ r there are a numher of other fartors, SWIM: of whidi arc i ~ i r p r ~ ~ d i r ~tlr;lt, tal~l~ affect life. I I atill There is ?onsideral)le differenre i l l this rrsprct het ~ C ~ slimlillp regular preheat-type lamps. The iiistairt-start rathodr whivh is ~ l s r di l l all slimline and instalit-start lamps van Iw o p e m t d ovrr ii ividr ~'angr,if current, from 120 to 430 ma, with rrlatiwly littlr d f c r t 1111 life. Ballasts also affert life. Even though they mtsrt sprvitivatioll r ~ y ~ i ~ i r c , . meiits, they have maiiufactnriiig toleranre and t h i w a r c drsigit rliffwvnws b e t w e n types. DECREASE0 UNCERTAIN OPERATION EXCESSIVE

LIGHT OUTPUT ANP STARTING AND MAY RESULT AT UNDER VOLTAGE.

INFERIOR LAMP PERFORYANCE AN0 DANGER OF OVERHEATING AUXILIARY MAY RESULT AT EXCESSIVE OVER VOLTAGE.

/ i

x

\

I

RECOMMENDED OPERATING RANGE BEST PERFORMANCE

I

I

L I N E VOLTAGES

FIG. 4.1 1

Characteristics of fluorescent lornpr OI function of voltage applied to bollort.

VOLTAGE-STANDARD RATINGS, VARIATIONS. CALCULATION OF DROPS

210

Fluorescent lamps also differ from filament lamps in that the frequency of start,irig is a factor iii the life obt,ained. Rated life is usually based on 3 hr of operation per start,. For 10 hr operation per start, the lamp's life is increased approximat,ely 35 per cent. Therefore, ally data 011 life vs. circuit voltage for the normal range in operat,iiig voltage ivould have little significaiice. At voltages below the lower limit, insufficient preheat current for proper cathode emission prior t o starting may result in short life. At voltages heyoiid the upper limit, the overcurrent operat,ioii may rcsult in unsatisfartory lamp life. Effect on Mercury Lamps. The effect of voltage variation on mercury lamps is shown in Fig. 4.12. Effect on Resistance Heating Devices. The energy input and therefore the heat output of resistaiice heaters varies in general with the square of the impressed voltage. Thus a 10 per cent drop in voltage will cause a drop of 19 per cent in heat, output. This, however, holds true only for an operating range over which the resistance remains constant. M a n y healing devices are conservat,ively designed arid if thermostatically controlled may operate satisfactorily even if the voltage varies 10 per cent or more. However, in many rases the designer must confine his heating units into a miiiimum of space and must, therefore, operate them near maximum rating. Also the temperature requirements for many heating applicatioiis IiecessiMe the operation of the heating units a t maximum temperature. h drop i n voltage meaiis a drop in heat input, varying with the square of the voltage, and a loss in production. On the other hand, excessive voltage will increase the temperature of the heating units and therefore will reduce their life. This condition applies especially t o furnaces operating at high temperatures near the maximum permissible for

I

I

I

I I

I I

I

/I/

I

I

I I

I

I

I

u OC 60

40

U

0

10

I

FIG. 4.12

-I

70

60

P R I M V0LTIT.F

I

rn CCYI

I

1

I

I

I

90 wo 110 iao 130 0 s TIIANSFORMER TAP SETTING

Choracterirticr of mercury type H 400-watt lamps.

I

140

I

VOLTAG&STANDARD

RATINGS, VARIATIONS, CALCULATtON

OF

DROPS

211

the type of heating unit used. To assure uniform high production and the best operating conditions, the voltage should be maintained mithiu a spread of plus or minus 5 per cent of rated voltage. Effect on Infrared Heating Processes. Although the filaments of the lamps used in these installations are of the resistance type, the energy output does not vary with the square of the voltage because the resistance varies a t the same time. The radiated energy vs. voltage is shown in Fig. 4.13 for the rating of 115 volts used on industrial infrared lamps. The wattage input is nearly proportional t o the energy output for a voltage range of 50 t o 150 per cent of rated voltage. The change in wattage and radiated energy is only 7 per cent for a 5 per cent change in voltage. However, this might he more harmful thau a larger change in typical resistance heaters employing thermostatic controls, if the product dryiiig is very sensitive t o temperature differences. For the usual paint-drying applications, no voltage coutrols are required with infrarcd lamps. Uniformity of product speed in the oven is the usual objective for coiiveyerized operations. Differences in heating requirements are rea,dily accomplished by connecting the infrared lamps to a number of circuits, so that some of the lamps can be switched on and off in accord with t,he exact, heat,ing needs. I n t,he cases vhere lamp sivitching cannot rompensat,e for the volt,age variat,ions, it may be necessary to use a voltage regulator to maintain conveyer speed and product quality. Effect on Electronic Equipment. The current-carrying ability or emissiou of all elect,ronic tubes is affect,ed seriously by voltage deviation from rating. Figure 4.14 shows typical emission curves plotted agairist cathode heater voltage. Curve 1, entitled Oxide Coated, applies t o most of the thyratrons, pliotrons, and rereiving tubes. Curve 2 for thoriated tungsten applies t o the small transmitter tubes and some of the hattery-

Radiant-energy output of General Electric Company industrial infrared lamps a function of impressed voltoge.

FIG. 4.13 QI


212

heated tubes. Curve 3, Pure Tungsten, applies t o the oscillator tube such as used in high-frequeiicy induction and dielectric heaters. The rathode-life curve f o r pure tungsten indicates that, the life is redured by half for esrh 5 per rent iiiwease iii cathode volt,age. This redured life is due t o the higher rate of evaporation of the rathode material. At voltages below rating, the loss d emission has very serious sec-

40LL

20 0

30

40

FIG. 4.14

Calculated values of electronic-tube emission and life


213

ondary effects. In a vacuum tube surh as the pliotron and kenet,ron a small loss of eniission below that needed means rednced ont,put and sometimes excessive tube heating which is reflerted in a shorter life. However, for gas-filled tubes such as thyratrons and phanotrons i n ivhirh t,he rurrent, is not limited by the tube spare rharge, if insuffirient emission is available t o carry the load current, the gas molecules bombard the rathode surfare and may destroy t,he tube in a matter of minnt,es. Therefore, it is extremely important that the rathode voltage be kept up near rating on these tubes for sat,isfactory service. I n addition t o the above factors there are ot,her important things to be taken into ronsiderat,ion. If the volt,sge is too high, the evaporated material from the cathode may contaminate the grid or anode and cause grid current and arc-back, making the tuhe iuoperativc. If the rathode voltage is too low in the gas-filled tuhe, the snrfare callnot be activated properly and loses its emitting effiiknry very quirkly. This permits bombardment, as explained above, and destruct,ion of the cathode. T o permit the voltage t o fall helow rather than t o rise slightly above rating i s serious. Standard industrial t,uhes are desigued t o operate \vith a voltage tolerance of plus or minus 5 per cent. Iloivever, if a closer tolerance than this can be maiutained, thc user will he amply repaid i n increased tube life and reliable operatioil. If voltage sij-ings must he tolerated, it is more desirable t,hat,t,he minimum s\ving be t o not less thau 95 per cent, of rating even hhuugh the average voltage may he slightly above rating. While this prartice \\-ill, of course, give somewhat redured tube life, it is preferablc t o low xwltage rr-hich rauees rapid tube drterioration. While t,he effect of voltage change is most, important on the tube cathode, it is also undesirable ill ot,her parts of the ririwit. Electrotiic circuits, as all other electric cirruits, lost power mparity rapidly if the voltage is decreased from rating. Although critiml circuits normally contain voltage-regulator tubes and other mealis t o hold a constant reference vokage in spit,e of line-voltage variat,ions, economic reasons prevent voltage regulation on t,he majority of rirruits, and henre thcir funrtion will naturally be impaired by excessive voltage variation. This is especially true when magnetic sat,uration is part of the roiitrol function. Effect on Solenoid-operated Devices. I n this group fall solenoids, brakes, valves, and rlutrhes. The pull of the a-c solenoid varies approximately as t,he square of the voltage. There is some deviation from this law, depending upon which part of the brake-horsepower cnrve the solenoid is working. The temperature rise, too, varies approximately a s t h r square of the vokage. I n general, solenoids are liberally designed and standard rommerrial solenoids are designed to operate satisfartorily on 10 per cent overvoltage

214

V O L T A G F S T A N D A R D RATINGS, VARIATIONS, CALCULATION OF DROPS

and 15 per cent undervoltage. Since an a-c solenoid has an inrush current of approximately ten times the sustained value when sealed, the branch circuit sJpplying it should be of ample capacity to prevent an excessive voltage drop. Effect on Capacitors. The corrective capacity of capacitors varies with the square of the impressed voltage. A drop of 10 per cent in the supply voltage, therefore, reduces the corrective capacity by almost 20 per cent, and where the user has made a sizable investment in capacitors for power-factor correction, he loses the benefit of 20 per cent of this investment. RECOMMENDED VOLTAGE SPREAD AT UTILIZATION EQUIPMEN1

Rased on the foregoing effects of voltage variation on utilization equipment and an extensive poll of industrial plant operating engineers, the AIEE Committee on Industrial Power Applications established the recommended voltage spreads at the terminals of devices in industrial plants. These are shown in Tables 4.8 and 4.9.* TABLE 4.8

Recommended Voltage Spread a t the Terminals of Utilization Devices in Industrial Distribution Systems 600 Volts and Below Commonly "red ulilizolion-device

Nominal Iyllem volt.ge

480

A00

Recommended limib of volloge at terminals of ulilizolion devices

"Oltage rating.

!

440,* 460 550,* 575

!

420-480 525-600

Drsigriations for nominal system voltages are those commonly used in industrial plants. * ThPse are standard polyphase-motor voltage ratings. t Polyphase power loads may not operate satisfactorily a t this l o m ~ rlimit

In designing industrial power distribution systems, the system design engineer should design for voltage spreads not in excess of those mentioned in Tables 4.8 and 4.9. If anything, it would be desirable to design for closer limits to allow for critical utilization apparatus that may be developed and widely used in the future. The history of electricity in industrial plants has been to extend its use to more and more functions. As

* Thcse rwommcndstions are in iuhstantial agreement with thP recommmdations of the joint EM-SEMA Committce whirh puhlishrd their findings in a report, Prcferrpd Voltage Ratings of AC Systems and Equipmcnt.


TABLE 4.9

21s

Recommended Voltoge Spreod a t the Terminols of Motors Served ot Primory Voltoge Recommended limits af voltage at terminalr of high-voltage moiors Motor-nome-plote *oltoge rating

Nominal syitem dtage

2400 2400 4160 4800 6900

*

~

1

2300" 4600 6600

1I I

2160 2250 3920

4500 6470

2380 2480 4320 5000 71 30

I'rmrnt standard rnot,or voltagc rating.

well as driving the utilization equipment, it is alço used for a11 types of rritical proccss control systems; therefore, its role is hecorniiig exceedingly important, and t o fulFiI1 this role effectively, good voltage must he rnaintaiiied iii industrial plants. L I G H T FLICKER V O L T A G E REQUIREMENTS

Relatively slom chaiiges in voltage are associated mith voltage spreads as discussrd iii tlie foregoiiig. There are, however, maiiy types of voltage changes 1rhii.h are of a traiisient nature aiid last only a feiv cycles. Thcse are commiiiily referred to as voltage flicker, aiid its primary effect is to cause flicker iii t h r light ciiitput of lamps. The arnount of voltage variatioii as a fiiiirtioii of frequency of variation which can be xvithstood on iiicaiidesrent larnps aiid not cause ohjei:tionahle psychological effects is shown iii Fig. 4.15. These curves were preseiited in the General Electric Review, hugust, 1925. Fluoresceiit lamps are less suhject t o flicker over a range of voltage that is beloiv that whirh mil1 piit them out. Iii industrial plants, voltage flicker i s caiised primarily hy the followiiig types of load: repetitive motor starting, large rei,iprocatiiig cornpressors, punch presses, etc., which dram a fluctiiating load; resistarice wcldcrs; aiid arc furnaces. T o elimiiiatc objcctionable light flicker, the design of the systcm should be siich that the lirnits of Fig. 4.15 are adhered to. Wider lirnits may be iiscd uiider certaiii coiiditioiis without cornplaiiit from the personnel orrupyiiig tlie affei,tcd arca. Ho!rcv&, this subject is so cornplicated aiid involved t h a t general guides other than Fig. 4.15 would probably not be of much use.

VOLTAGE-STANDARD RATINGS. VARIATIONS. CALCULATION OF DROPS

216

FLICKER OF INCANDESCENT LAMPS CAUSED 81 RECURRENT VOLTAGE DIPS

I

5 0 Y

w 3

5 0

' t

t-

z Y

0

,'w

a 0 DllO

10

PL"

"0"I

82 6 YlUUlLI

DlPI

J

2

I

TIME

FIG, 4.15

30

12

L

PLI1 SECOND

IFCOYDL

BETWEEN DIPS

Relation of magnitude of voltage dips to frequency of dips for incandescent

IWlPS.

METHODS O F REDUCING VOLTAGE SPREAD AND FLICKER REDUCING VOLTAGE SPREAD (See Fig. 4.271

\Vitlr recommended values of voltage spread established by the N E E Industrial Power Systems Committee a i d EEI-SEA\Z.%, it is possible to

study specitiv syst,ems to see hon. they romparc with these rcquiremerits. Where voltage spreads arc found t,o be heyorid t,hose limits, there are four 11-aysof reducing the voltage spread. 1 . Carry the power further a t a higher voltage and a t a lesser dist,aim at 1o\vcr voltage, i.e., use the load-center power system. 2. 1tediii.p the impedance of the systrm. 3. Use regiilat,iiig equipment t o rompelisate for volt,age drop. 4. Use s\~iti~Iied capacitors. llaintaiiiiiig the volt,age at an average desirable I e i d also requires the judicious use of traiisformer ratios and taps. Traiisformer taps (for changing a t no load oilly) do trot, reduce the spread but affect only t,he general voltage level arid particularly the light load voltage in the plallt.


217

Load-center Distribution Systems. The load-ceiiter distribiitioii system is 11011- almost uiriversally used i n industry for, among othcr reasons, it provides Ion- voltage drop, henre small voltage spread, berausc the power is carried right t o the load i.etiter at high Iwltage. Refer to Chap. 11 for a one-liiie diagram of a typical load-retiter system. Table 4.10 illustrates the advantage of higher voltage distribution. I t is obvious from this table that the tiig gaiii is made by going from voltages iii the (i00-volt class to voltages i l l the 2.4- to 13.8-kv class for rarryitig poll-er from the source to the load ceuter. T o illust,rate furthcr, supposc that the voltage drop i n a 480-volt system Tx-ith long serondary feeders is 20 per relit total in the secondary feeders oiily. Should this power he carried a t 4160 iiistead of 480 volts, the percentage voltage drop x~ould have been only slightly ovcr one-quarter of 1 prr rrnt. Siirrc the loadcenter system minimizes the length of low-\.oltage feeders, it minimizes one of the chief causes of voltage drop atid herive redures voltage spread.

TABLE 4.10

Per Cent Voltoge Drop as a Function of Circuit Voltage for a Feeder of a Given Cross Section circcuit V0ltoge 240 480 2,400 4,160

Relative Per Cent Voltage Drop

13.800

0.12

400 100 4

I .33

Some examples will serve t,o illustrate the better voltage conditions in the load-cetit,er system. The average 480-volt load-renter substatioii is rated 750 h a . With ail average load density of 10 va per sq ft, this substatioir will servc a i l arca of 73,000 s q ft,. Ideally, the load area would be a square, with the substatioii esartly i i i the renter; then the longest feeder length ivould tie about l(i5.ft. Rut it i d 1 he assumed t,hat t,he area is somewhat rcctaiigular atid that the suhstatioii rannot he lorated exactly at the center. The artual length of the longest feeder might then he ahout 200 ft. Figure 4.16 rontains charts showing the voltage profiles for this 480-volt suhstatioii. The trairsformcr taps should lie set for 480 secondary volts when the primary voltage is at its maximum atid with no load on the substation. The highest, voltage that is eticoi~titered by ally equipment served hy this substatioii is 480 volts. At maximum load, voltage drop has its maximum effect. A 4 pcr rcnt voltage reduction i u the primary system is assumed, to illustrate the Ion--voltage rondition. This could he due to a dcrreasc i n the power-vompauy supply voltage with inrreased load on its system. h drop of 15 volts due t,o traiisformer react,ancc can he experted. Assuming the 200-ft feeder t o ronsist of a 250-MCM cahle per phase and to he fully loaded a t 80 per rent power factor, i t mill~iutro-

V O L T A G b S T A N D A R D RATINGS, VARIATIONS, CALCULATION

218

OF DROPS

duce about another 7 volts drop. h final 5 volts may be lost in the branch circuit. The result is a minimum voltage at the end of the branch circuit of 433 volts. In this system, voltage varies between the limit,s of 433 volts and 480 volts---a voltage spread that should, in general, be satisfactory. The old-type system often uses a suhstatioii as large as 3000 kva at 480 volts. And not heing of unit substation construction, it has to be located at one edge of the load area-probably with the t,rausformers outdoors. With the same load density as before, 10 va per sq f t , the 3000-kva snbstation must supply an area'of 300,000 sq ft,. I n this substation the longest feeder will probably be ahout 900 ft. The corresponding feeder voltage drop will be 29 volts. Here the voltage spread is from 411 to 480 volts. Such a spread is well heyond the recommeuded limits. A full load voltage of 411 i s too low t o be coiisidered good practice; 420 i s the recommended minimum voltage for 440-volt motors. 7 5 0 KVA SU0STATlDN

460

2440 420

400 NO LOAD VOLTAGE CONDITIONS WITH PRIMARY

VOLTAGE a T M A X I M U M

2 440->

420

--

SECONDARY FEEDER/ VOLTAGE DROP- 7 VOLTS

-VOLTAGE DROP-

VOLTAGE-STANDARD RATINGS. VARIATIONS. CALCULATION O F DROPS

219

60

48 w 0

a

3 0 >

ABLE 51ZE

k

3

U0.4 vo. I 000

U

n U w

z

FULL LOAL AMP

90

I40 210 500 MCM 4 0

201

-

(

-

7 24l W

z

J

5 SECONOARY FEEDER LENGTH(FEET1 FIG. 4.17 Chart showing length of three-conductor 600-volt cable in iron conduit to produce 2 3 per cent voltage drop a t the most unfavorable power factor and full load on the cable.

Tolerable Secondary-feeder Voltage Drop. Figure 4.17 offers a guide as to about how far fully loaded cables for circuits 600 volts and less can be run and not encounter voltage-drop troubles in a n average industrial plant. Thcrc are many variables which can alter the maximum feeder length materially, such as power factor of load, primary voltage drop, load per feeder, etc. The chart of Fig. 4.17 is based on representative conditions, i.e., primary voltage drop 5 per cent, transformer drop 355 per cent, branch-circuit feeder drop 156 per cent. The remainder is the secondary-feeder drop of 2>5 per cent, the basis of Fig. 4.17. The allowable spread a t 480 volts is 480 t,o $20, or 60 volts or 1255 per c e n the sum of the percentages just ment,ioned. Secondary-feeder drops greater than 235 per cent should be cherked under conditions expected at the plant, t,o see if t,hey can be tolerated without c-using undesirably wide voltage spreads. Looking at this another way, 480-volt secondary feeders longer than 250 f t for small cable sizes and 400 f t for larger cable sizes should be

220

V O L T A G G S T A N D A R D RATINGS. VARIATIONS, CALCULATION OF DROPS

avoided from a voltage-drop st,andpoint. If longer feeders must he used, check the voltage drop. The tolerable secondary feeder lengths are somewhat longer for 600-volt cirruits, i.e., abont 300 and 500 ft, respectively, aud considerably shorter a t 240 and 208 volts, i.e., about 125 arid 200 f t , respectively, a t 240 volts and 100 and 175 ft,, respectively, a t 208 volts. Reducing Impedance. Sirirc volt,age drop is a product of current t,imes impedance, anything t,hat, is done t,o reduce the impedance of a circuit will reduce its voltage drop. The following are some suggestions: 1. t-se closely spaced cotidurtors, i.e., use cable instead (if open wire with widely spaced (.onductors. 2. I'se interleaved huses, that is, bnses wit,h several cotiduvtors per phase arratiged 8 , B , C; -4. B , C ; -4, R, C , etr., instead of liuses with all conductors of one phase widely separated from the other phases. 3 . Use t x o smaller vahles in parallel instead of one larger cahle. 4. Use standard-rea(.tattce instead of high-rea(.tanre transformers. €Iigh-reart,ance transformers reduce short-circuit rurrciits but increase voltage drop, particularly lrith poor power-factor loads. A compromise is necessary hecause a lover t,hari standard-reactanre transformer, n-hilc reducing voltage drop, may invreasc short-circuit, rurrents so high as t o require unreasonable switchgear for protection of the circuits fed by the transformer. 5 . Keep feeders-particularly low-voltage feeders- as short as possible. 6. Use series caparitors t o neutralize the i n d u h v e reactance of a rirwit. There are few if any general applira.tions of the series capacitors for this purpose in indust,rial plants except in coiiriect,ion wit,h resistanre welders as n o k d later in this chapter. In a few rases they have heen used in connection with motors t o maintain sufficient vokage a t the motor terminals when starting a large motor oii a soft, system or t o neutralize system impedance t,o maintain good voltage on lights, etc. U s e Regulating Equipment. Even where the plaiit power system uses a load-center system t o rarry t.he power the great,est practical distance at high voltage, where impedances have been kept, t o a minimum and lorn-voltage feeder lengths as short as possible, it, may not be able to meet the required voltage spreads because of too much voltage variation in the primary supply system or because of a process requiring unusually close voltage spread. In some old plants, low-voltage feeders may he large and long, giving excessive voleage drop i n the secoiidary system. Marry plants operate at 240 or 208Y/120 volts and have excessive voltage spread that would be reduced t o tolerable limits if 480 volts were used instead. However, the change may not he practical or ecoriomical a t t,he moment. I n such cases voltage-regulating equipment provides the answer t o the problem.


221

Voltage Regulation of M a i n Power Source. Where a transmission or disl,ribution h i e supplies a plant with power whioh has a voltage spryad

greater than a’.iout 5 per rent,, it may tie difficult t o mairrtairi the desired voltage spread even wiLh the best designed plant power system. In these cases sonic form of voltage regulatioii is often required. If the supply is at, high v d t a g e arid must, he stepped down t,o below 15 kv (commonly -Ll(i0 or 13,800 volts) for distritiut,ion, regulation rain tie built int,o the transformer. This regulation is accomplished by automatic t a p changing which xi11 operat,e under load (load-rat,io (:oiitrol). Usually t,here are 32 (76 per cent) steps to enable close volt,age control over a range of plus or minus 10 per cent.. 1,oad-ratio control for plus or niiriirs 10 per cent range is a very low cost iii the over-a11 plant costs, arid yet because the load-ratio corit,rol provides good voltage, it will improve prodii(tiiin and quality of maliufartured goods. Iienc,e, the dividend from t,his small investment \rill often repay the investmerit many t,imes over earh year. It is strongly rei~ommcrided tha.t load-ratio control tie cmsidered itt every transformix stepping down from voltages ahorc 15 k r to plant primary voltage iii the raiige of 2:I to 13.8 ku. The systems aliove 1.5 kv arc not dn-ays regulated to suit, t,he industrial plant but for most, r f i r i i ~ n ot x r - a l l opcratioli of t,lie poiver sys-

FIG. 4.18

A typical outdoor packoged substation in which bod-ratio control con be

incorporated.

222

VOLTAGLSTANDARD RATINGS, 'VARIATIONS, CALCULATION OF DROPS

tem. When load-ratio control is installed, both the utility and industrial plant can operate their systems independently and to their own best advautage without interference voltage wise. Figure 4.18 illustrates an outdoor substation, typical of those whose transformers can include load-ratio control. Voltage Regulators. If power is supplied by the utility at below 15 kv, the only transformation required is at the individual load-center substations. Load-ratio control in each industrial load-center unit substation is uneconomical and even may he impractical. Hence, where the primary-voltage spread is wide enough to require voltage regulation, separate voltage regulators should be installed in the primary supply, Fig. 4.19. For this service either three-phase step voltage regulators (Fig. 4.20) or induction voltage regulators (Fig. 4.21) can be used. Their standard range of voltage regulation is plus or minus 10 per cent. The question is sometimes raised as to whether two induction regulators should be connected in open delta. This is slightly less expensive than three regulators to regulate three-phase circuits. However, the opendelta connection creates an unbalanced voltage condition that should be avoided. The voltage unbalance is small but may be enough to increase

STEP OR INDUCTION VOLTAGE REGULATOR REGULATOR HOLDS CONSTANT VOLTAGE HERE \

NOTE

:

v

v

t Y

THE BY-PASS PERMITS MAINTENANCE O F SERVICE TO PLANT WHERE REGULATOR IS BEING MAINTAINED

-

-

FIG. 4.19 One-line dioorom rhowino the a d , .i c o t i o n of steD or induction voltom r e d o tori for holding constant voltage on the plant primary bur for plants served at primary voltage.

VOLTAGE-STANDARD

RATINGS, VARIATIONS, CALCULATION OF DROPS

223

FIG. 4.20 Typicol three-phore step voltage regulator roted 13,200 volts, 208 kvo, plus or minus 10 per cent voltage regulotion.

FIG. 4.21 Typicol induction voltage regulator rated 225 kvo, 4330 volts, plus or minus 10 per cent voltoge regulation.

224


the heating appreciahly in fully loaded polyphase motors. For this reason, best prartice avoids the open-delta conneition in favor of three-phase regulation. It is rerommended that serious consideration be given the addition of t,hese regulat,ors i n the plant supply lines whenever the expected voltage spread in the primary supply lines exceeds ahout 5 per cent. Regulators may he hypassed for maintenance and a t the same time maint,ain unregulat,ed service t o the plant. Itegulat,ors, like any other piece of apparatus, must be given consideration from a short-circuit, standpoint. Feeder Voltage Regulation. trt,ilit,iesoften regulate individual feeders at distribution voltage (2100 or 416F volts, for example) t o compensate

480 VOLTS

SECONDARY FEEDER INDUCTION VOLTAGE REGULATOR FEEDERS TO MOTORS, ETC LIGHTING FEEDER

LIGHTING LOAD 120 VOLTS FIG. 4.22 One-line diagram showing the opplication of air-cooled induction voltage regulators for secondary feeder regulation.

VOLTAGE-STANDARD


225

for the voltage drop i n that, feeder. The itidustrial plaut does iiot often ry, since \.oltage drops in individual primary ferde usually small, less illan 1 or 2 per cent. Thus, regulation of voltage a t the main h s is more rwmmotily used. While there seems t o he little, jirstifiratioii for irrdividuul primary-feeder regulation, there may be many appliratioris for individual seroiidaryfeeder reeulation. For examuk, ._ . ,tlie voltage spread may I)<> satisfartory for t,he majorit,y of utilization e q u i p ment, such asmotors, welders, etr., but not, considered good cliough for lights. I n such cases, t,he lighting feeder may be regirlat,ed and the rest irnregulated, Fig. .4.2%. For such applications, air~ooledregirlalorslikethat shown in Fig. 4.23 may be used. I n other cases, individual loads a t GOO volts or less may require voltage regulation t,o obtain the desired performance from the equipment,. Rirh loads might he heating unit,s, process cont,rol, infrared ovens, hluepririt machines, lights, radio arid television transmitt,ers, brooders, etc. Where these loads are served at, utilizat,ion voltage, aii iridrictimi regulator like that, of Fig. 4.23 may be used. lnductrol P o w e r Pock. A iie\\- dcvelopment is a regulating loi\~-voltage subst,atioII known as the Inductrol Power Pack. It, is a itiiit made up primarilyof an indurtioii voltage regulator arid a dry-t,ype transformer. The transformer is rat,ed 480 or 600 volts on t,he primary aiid %08Y/lZOor FIG, 4.23 A modern induction voltage 120/240 volts on the secondary. A regulator for circuitr 600 volts and leis. Typical of either single 01 three phase. primary switching- arid protective device arid secondary terminals complete the package. This unit may be used for supplying regulat,ed lighting power from general-purpose 480or 600-volt feeders or for supplying any other loads with regulated 120volt power from 480- or 600-volt power systems. Shunt C a p a c i t o r s . Refer t o Chap. 8 for a comp1et.e discussion of the application of shunt capacitors t o improve voltage conditions. Autotransformers. I n some cases where the general voltage level is

226

VOLTAGE-STANDARD


lorn and transformt?r taps cannot he used t,o correct for it, autotransformers may he used t o provide a permanent boost i n voltage. T h e autotrausformer does riot reduce t,he spread. A t,ypiral appliration xvouid lie in the case of a, 208Y/120-voll system supplying 220-nolt, mot,ors. The volt,ege may be proper lor the lights hut not high eiiimgh for t,he 220-volt motors. An autolrmisformer could he used l o step 208 volt,s up t o 220 volts for the motors only. Generator Voltage Regulators. Where power is generated by t h e plant,’s oxvti geiir.ralors, the voltage on t,he powerhouse bns can be held constant or exwi varied with load to compensate for voltage drop as load comPs on. Problems of voltage rcgulat,ion where industrial generators are operalnd iri parallel with utility systems are referred to in Chap. 15. REDUCING FLICKER (See Fig. 4.27)

Reduction of flicker is often a much more difficult prohlem than the reduction of voltage spread previously referred to. Flicker due t o reciprocating motor-driven loads such as compressors, purich presses, et,c., can often be reduced by increasing the inertia of the met:lranical system to smooth out the pulsations. Where this is not ive, t,wo t,liings may be done. One is to separate flicker-producing load from the lights or critical load, i.e., use separate supply circuits. The nther is to use a voltage stabilizer, Fig. 4.24, to feed the critical load. Sometimes the critical load is fed through a motor-generator set t o provide good voltage for tliet load. This, lioxvevrr, is more expensive over all thaii voltage stabilizers and in gcmral offers no advantage in this

FIG. 4.24

Typicol voltage stabilizer.

VOLTAGE-STANDARD RATINGS, VARIATIONS, CALCULATION

OF DROPS

227

application. Voltage regulators previously discussed are not fast enough to correct flicker, but for single-phase circuits, and in small sizes, automatic voltage stabilizers are available to hold voltage mit,hin very close limits. A typical model is designed to maintain an output voltage of 115 volts with maximum variation of plus or minus 1 per cent, even though the supply volt,age may vary between 95 and 130 volts. The volt,age stabilizer has no moving parts and no electronic tubes; its operation is obt,ained from the properly coordinated characteristics of reactors and caparitors. Series Capacitors. Series capacitors can be of value in reducing voltage flicker. C:hapt,er 8 contains a complete discussion of the application of series capacitors.

SELECTION OF TRANSFORMER TAPS

All modern transformers in ratings above 100 kva and most or those helow that kva rating have taps in the windings to change the turn ratio. The taps do not materially affect the voltage drop through the transformer; they merely change the turn ratio, hence the no-load voltage ratio. For example, a standard transformer rated 2400-480 volts may have four 2>5 per cent taps in the 2400-volt winding. The standard for these taps in transformers used in industrial systems is to have two 256 pcr cent, taps above 2400 volts and two 24i per cent taps below 2400volts. The no-load ratios of such a transformer would be as given in Table 4.11. TABLE 4.11

No-load Voltoge Ratios of Standard Transformer Rated 2400-480 Volts 2520-480 “0th 2460-480 volts 2400-480 volts 2340- 480 volts 2280-480 volts

5% obove tap 236% obove top Norrnol rating top

2>P%

below top

5% below tap

These taps do not improve voltage regulation but are only for changing the general vokage level iq the plant. If a 2400-480-volt transformer is connected to a system whose maximum voltage is 2520 volts, then the 2520-480-volt tap could be used which would provide a maximum of 480 volts no load on the system, as shown by curve A , Fig. 4.25. If, for example, another system had a maximum no-load voltage of 2400 volts, then the 240&480-volt t a p could be used to provide 480 volts no load in the plant. This would be as shown in curve B , Fig. 4.25. Similarly if a plant had a maximum voltage of 2280 vo!ts, then the 2280-480-volt tap could be used to provide a maximum of 480 volts no load in the plant, as shown in curve C , Fig. 4.25. It will be noted that in all cases the secondary no-load voltage is 480 volts; so the secondary system does not know

2600-

-

-

-

-

U

c

2

2400

>

-

480 VOLTS ---- ---_ _ _MAX ____--

-

> (r

440 V MIN

-?

B

40 VOLTS

-

U

I

(L

a

4 8 0 VOLTS MAX

2300-

-

SPREAD

________

480 VOLTS MAX

---------

440V MIN

VOLTAGkSTANDARD RATINGS, VARIATIONS, CALCULATION OF DROPS

229

no-load voltage. By using the next tap u p on the transformer, that is, the one rated 2460-480 volts, the turn ratio of the transformer has now been changed so that the no-load voltage is 472 volts, as shown in curve B , Fig. 4.26. The voltage spread will be substant,ially the same, i.e., 40 volts, so that the minimum voltage is now 432 volts, which is well above the recommended minimum for plant distribution systems. By judicious selection of the transformer t a p t,he voltage within the plant can he kept Tyithin acceptable 1imit.s provided that the primary voltage does not vary more than about 5 per cent and that the plant distribution system is designed along modern lines with the load-center system using short secondary feeders and transformers not larger than about 1500 kva a t 480 volts or proportional sizes a t other secondary volt,ages. Changing taps cannot, correct conditions where voltage spread is t,oo great. For example, suppose a plant suffered from low voltage at remote points and had a large volt,age spread. T o be specific, suppose the spread was 80 volts and the minimum voltage at the remote end was 400 volts, then the maximum voltage would be 480 volts. If taps are changed to raise the general voltaga level, the spread will not change but the 400-volt

I 8 5-VOLTS - -- --4---MAX - --- -_-

g a

I-

J

0 5

>

a a

I LL

2400

40 VOLTS SPREAD

440 V MIN

P

FIG. 4.26 Voltage profile showing that rotisfactory voltages con be obtained without excessive no-load voltage by proper election of taps on transformer.

230


Volloqe Correction For TypicoI Feeder Circuits

FIG. 4.27

Summary of methods of improving


LOW LOAD VOLTAGE Feeder VOltoge Condition

Ci.<"iI Loading

1 . tow

Normal

2. High leeden drop

Normal

3. High feeder

Normal

iee

1A1 (81 ICI

(81 ID1 ICI

(El I81

drop.

4. High feeder drop.

Overload

IEI IF1

HIGH LOAD VOLTAGE I . Normal drop.

2. Normol

NO load

i I

drop.

3. vo1t.g.

Volt.ge r e g d o t o r Tronlformer ,ap setting

IBI 1A1

Voltage 'egulalor

101

Tranrformer top 'elting

(A1

"No b o d " Leoding Aulornotic switching o f capociExcept shunt ot no lood tors
rile

IVolt.ge

regulator is Only P'"
I0l"tiO"J

Correcl by m e o n r of

Flicker

I . RerirtanPOI or seom.

Series r a p o d o r with welder to reduce dernond by power.farlor correction Series c"Po
2. Flmh.

rssislmnre

Separote welder supply r i r 4 l Volloge stabilizer l o r lighting circuit Separate welder supply c i r w i t

welders.

3. Motor loads. such 01: sow mill^, Rubber milli. Grinders. 4. Arc furnorer.

Voltoge Ifobiliier for lighting circuil Series coparilor in line to "e",,.lize re.artance

Sep..ate

motor '"pply Ci.CYil

Voltage slmbilirer for lighting circuit Sante 0 s for lmotonl

voltoge conditions in an indurlriol p l a n t

IF1

IHJ

LIGHTING FLICKER Lood Causing

IGI

'y'lem

231


232

minimum may he raised t o 420 volts. At the same time the maximum is raised to 500 volts, too high for generally sat,isfactory performance. Conversely if the maximum voltage is too high and a wide spread exists, the chatrgiug of taps, to reduce t,he maximum voltage, reduces the minimum vokagc still further. CALCULATIONS O F VOLTAGE DROP CALCULATION OF STEADY-STATE VOLTAGE DROPS

Steady-state voltage drops are duc t,o current flowing through a n impedauce. T o calwlate the steady-state voltage drop, the circuit impedance, circuit current, and power factor of that curreut relative t o some voltage must he known. I n this discussiou the power factor will be that of the load. Rigorous methods of calculating voltage drop can he very involved and complicated, particularly in cases \\.here the sending-end voltage only is kiron-u and t,he current and poiver factor of the load vary with variation of receiver-end voltage. For the purpose of ordinary use in industrialplant problems, approximate methods are generally satisfartory. Two methods of determining voltage drops are described. The first is hy calculation using either the sending- or receiving-end voltage, the magnitude and power factor of the load current, and the total impcdance of the rircuit. The second method involves using charts of voltage drop vs. load for the various circuit, components. Voltage Drop by Formula. Thc voltage drop in a power system may he calculated by selecting the formula which is most snitahle as t o accuracy desired and the voltage n-hieh is known, such as the receiver- or seuding-end voltage of the circuit. 111all the following formulas except Eq. (4.8) the voltages are line-toueutral voltage drops. T o obtain t,he liue-to-line voltage drop in a threephase system, multiply the line-to-neutral voltage drop by For single-phase syst,ems t,he line-to-line voltage drop is obtaiued by multiplying the line-to-neutral voltage drop by 2. I t is possible under some (~ouditionsto oht,aiu an answer with a negative sign from t,he folLo\yiiig formulas. I n such ('ases t h e auswer should he interpreted as showiug that the receiver voltage is higher thao the sending-end voltage. These cases will be rare, however, since the great majorky of systems will have receiver or load voltages which are lo\\-er than the source or sendiug-end voltage.

4.

Nomenclature for Formulas e = line-to-ucutral voltage drop es = line-to-neutral voltage at source end

en

=

line-to-neutral voltagc at load end


233

angle whose cosine is the load power factor line current = resistance of the circuit, ohms = reactance of the circuit, ohms (By convent,ion, inductive reactance is positive and capacitive reactance is negative.) cos 8 = load power fact,or in decimals sin 8 = load reactive factor i n decimals (By convention, sin B is positive for lagging power-factor loads and negative for leading power-factor loads.) Approximate data on circuit and transformer impedances may be obtained from Chap. 1 and trigonometric functions from the Appendix. Exact Formulas. If eR is known, 8 I R X

= =

Line-to-neutral voltage drop = d ( e B cos 8

+ I E ) ? + (ee sin 8 + I X ) p- eR

-

If es is known,

Line-to-neutral voltage drop = es I R cos 8 I X sin B - .\/es* - ( I X cos 8 - I t l sin 8)'

+

+

(4.1)

(4.2)

The voltage drop can also he obtained hy a proportional method. Both the voltage drop and phase shift due t o voltage drop can be obtained by

where all quantities are expressed vertorially and Z, is the equivalent load impedance and Zsis the system impedance including ZL. Voltage drop

= es

- eR (numerically)

(4.4)

I n Using Eq. (4.3) it should he noted that the load impedance is assumed to he constant, whereas all other formulas are based on the load current remaining constant. A p p r o x i m a t e Formulas. I n practical cases, the results of these approximate formulas are suffiriently accurate where a slide rule is used. If e R is knomn, Line-to-neutral voltaee droD Y

-

I ( R cos

'+

sin

( I X cos 8 - I R sin S ) l + 2(en I R cos 8 I X sin 8 )

+

+

If es is known, Line-to-neutral voltage drop =

I R cos 8

- I R sin 8)* + I X sin 8 + ( I X cos 8 2e.

234


Most Commonly Used Approximate Formula.

Where either e R or es

is known, then Line-to-neutral voltage drop = I ( R cos 0

+X

sin 0 )

(4.7)

Equation (2.7) can he converted as follows to calculate the per cent voltage drop : Per cent voltage drop

=

+

kva (R cos 0 X sin 0 ) 10 (kv)*

(4.8)

where kva is three-phase kva and kv is line-to-line kilovolts. For singlephase circuits the per cent drop is twice this value. From the vector diagram in Fig. 4.28 it can he seen that, whilc Eqs. (4.7) and (4.8) are approximate, they are close enough for practical purposes. In practical cases the angle between e,; and ey will he small. In these formulas the error diminishes as the angle between e R and es approaches zero and is exact if that angle is zero. The latter condition will exist when the power factor of an inductive load is the same as the power factor of the inductive circuit through which load current is causing the voltage drop. In Fig. 4.28, 0 is the power factor of the load. Effect of Nonlinear Loads. The error caused by variation of load cur-'-_ rent and power factor with voltage applied to the load is not taken into consideration in any of the foregoing formulas. If this error is significant, it may he compensated for by using the cut-and-try method; that is, first assume a given load or receiver-end voltage eR in the formulas. Then if the value obtained by subtracting the calculated voltage drop from the sending-end voltage is considerably different from the assumed receivingend voltage, make another try. Generally such refinement is not necessary when the total plant voltage drops are less than 10 per cent. SENDING END OR BUS VOLTAGE

\ R E C E I V E R OR LOAD VOLTAGE

F

CALCULATED VOLTAGE DROP

ERROR

ACTUAL VOLTAGE DROP

FIG. 4.28

Diagram indicating magnitude of error when using Eqr. (4.7) and (4.8).


235

STEADY-STATE VOLTAGE DROP BY USE OF CHARTS

Voltage Drop in Transformers. Figures 4.29 and 4.30 may be used to determine the approximate voltage drop in single-phase and three-phase 60-cycle liquid-filled self-cooled transformers. The charts are applicable for single-phase transformers by entering the chart a t three times the single-phase kva rating. Figure 4.29 covers transformers in the following ranges:

Single-phase : 250-500 kva, 8.6-15-kv insulation classes 833-1250 kva, 2.5-25-kv insulation classes Three-phase : 225-750 kva, 8.6-15-h insulation classes 1000-10,000 kva, 2.5-25-kv insulation classes An example of the use of the chart is given below. Example. Find the voltage drop in a 2000-kva three-phase KO cycle

236


transformer rated 4160-480 volts. The load is 1500 kva at, 0.85 power fact,or. Solution: Enter the chart on the horizontal scale a t 2000 kva. Extend a vertical line t o its interpeetion with the 0.85-power faetoP t u n e . Extend a line from this point horizontally to the left t o its intersection with the vertical scale of per cent voltage drop for rated load. Multiply this value by the ratio of actual load to rated load. Per cent drop at rated load

=

3.67

l5Oo X 3.67 = 2.75 2000 Actual voltage drop = 2.75 per cent X 480 = 13.2 volts Per cent drop at 1500 kva

=

~

Figure 4.30 applies to the 34.5-kv insulation class transformers in ratings from 1500 to 10,000 kva. These curves can be used t o determine the voltage drop for transformers in the 46- and 69-kv insulation classes by using appropriate multipliers a t all power factors except unity. To correct for 46 kv, multiply the per cent vokage drop obtained from the chart by 1.065, and for 69 kv multiply by 1.15. Example. Find the per cent voltage drop in a 5000-kva 69,00013,800-volt three-phase 60-cycle liquid-filled transformer carrying 3500 kw a t 0.8 power factor. Solution: Enter chart Fig. 4.30 a t 5000 kva and read per cent voltage drop where this transformer size intersects the 0.8-power factor curve. Per cent voltage 6 Q

5

05 w

=

4.25 for 5000 kva

NOTE: CURVES ARE BASED ON 6 PERCENT IMPEDANCE FOR 34.5 KV CLASS

I I I I I I I

I 4 -

&

54 a

0

U

0

u 3 0 4

5 8 , Iz Y

Y

U Y I

a

0

TRANSFORMER RATING-THREE PHDSE KVA

FIG. 4.30 Tronrformer voltage-drop curves for three-phase transformers, 34%-kv voltage class.

VOLTAGbSTANDARD RATINGS, VARIATIONS, CALCULATION OF DROPS

237

- 4380 kva 0.8 Multiplier for 69-kv insulation class = 1.15

Transformer load

= 3500

Actual per cent voltage drop

=

4380 4.25 X 1.15 X - = 4.28 per cent 5000

Voltage Drop in Cable. Voltage-drop curves, Figs. 4.31 to 4.34, may be applied with reasonable accuracy to all types of paper-insulated, rubber-insulated, and varnished-cambric-irisulated cable insulated for 600 or for 5000 volts. Two charts were prepared for each of these two voltage classes of cable to cover the different t.ypes of installat,ions for cable sizes No. 14 to X o . 4/0 Amg and 250 to 750 MCM. Voltage drop for loads between 0.7 power factor lagging and unity is shown for t,his range of cable sizes for three-conductor and three single-conductor cables in magnetic conduit. The resistance and reactance used in preparing these charts are taken from Chap. I. They are calculated values based on 75 C copper temperature and scattered tests. In determining reactances, it was assumed that for three conductors in conduit the cables d l lie a t random in the hottom of the conduit. If the cables are twisted together so that they operate in contact with each other, they should be regarded as a three-conductor cable. The chart,s are prepared for three-phase voltages. For single-phase circuits consisting of a two-conductor or two single cables in a conduit, the. voltage drop measured line-to-line will be 16 per cent higher than indicated in the charts. Use of Voltage-drop Charts for Cable. First, select the chart applying to t,he problem with regard to voltage and type of installation. Enter the chart a t the abscissa with the power factor of the load. Extend a line vertically from this point to the correct size cable. On the ordinate read the volts drop per 100 amp per 100 ft or per 10,000 amp-ft. Multiply this value by the multiple of 10,000 amp-ft, for the problem under consideration to get line-to-line voltage drop in a t,hree-phase system. For a single-phase system multiply the three-phase drop by 1.IG. Example. Assume that a 500-ft three-conductor rubber-insulated size KO. l/O-hwg cable in magnetic conduit is the feeder for a three-phase 440-volt 60-cycle 150-amp 0.8-power factor inductive load. Find the voltage drop. Solution: Enter chart, Fig. 4.31, at 0.8 power factor and move upward to the KO.l/O-Awg cable curve. From the point of intersection move to the left and read thc voltage drop as 2.08 volts per 10,000 amp-ft.

Ampere-feet in cable Actual voltage drop

= =

500 X I50

=

75,000

E 0 X 2.08 = 15.6 volts 10,000

238

W

LL w n

I 4

8 0

0 LL

w

0

u)

3

0

>

z &

z n W

c 3

4

50 >


VOLTAGGSTANDARD RATINGS, VARIATIONS, CALCULATION OF DROPS

FIG. 4.32 conduit.

239

Voltage-drop curves for three single-conductor 600-volt a b l e r in magnettc

240

FIG. 4.33


Voltage-drop curves for three-conductor 5000-volt cable in magnetic conduit

or interlocked-ormor cable.

242


OF DROPS

Voltoge Drop in Busway. Figures 4.35 and 4.36 may be used to determine the approximate voltage drop in a busway. Figure 4.35 applies to a busway that is designed specifically for low-voltage drop. Figure 4.3F applies t o a typical feeder busway of the type used with plug-in switches. Figure 4.35 gives the line-to-line voltage drop in volts for GOO-, 800-, 1000-, and 1350-amp low-voltage-drop busway. These curves apply only for balanced loading of the busway at an operating temperature of 70 C. The voltage drops for other than rated load may be obtained by multiplying the voltage drop for rated load by the ratio of actual load to rated load, Similarly, the voltage drop for lengths other than 100 ft may he

M

40 60 BO LOAD POWER FACTOR

W O

20

40 60 80 LOAD POWER FACTOR

W A D POWER FACTOR FIG. 4.35 Voltage-drop curves for low-voltage-drop burwoy ot rated load. operating temperature assumed.

100

70

c


243

obtained hy multiplying the voltage drop for 100 f t by the ratio of actual length to 100 ft. These corrections are expressed in the following formula: Actual line-to-:ine voltage drop =

voltage drop for 100 feet at rated load X

actual load rated load

actual length 100 ft

Example. Find the voltage drop on a 200-ft run of 800-amp husway carrying a 600-amp load a t a 90 per cent power factor. Solution: Enter Fig. 4.35 for au 800-amp husway at 90 per cent power factor on the horizontal scale. Follow a vertical line to its intersection

4 . 5 X 3 = 13.5 V O L T S

FIG. 4.36

Voltoge-drop curves for typical plug-in bvrwcly carrying rated load.


244

0 0 the curve and proceed horizontally to the left. The intersection of this line with the vertical scale is the voltage drop per 100 ft for an 800-amp busmay, 2.4 volts.

Line-to-line voltage drop = 2.4 X

600 200 X - = 3.6 volts 800 100 ~

Single-phase voltage drops may he obtained by multiplying the threephase voltage drop times 1.16. Figure 4.36 gives the line-to-line voltage drop in volts for a plug-in type busmay. An example is given with the curves to illustrate their use. Example of System Voltage-drop Calculation. The power system shown in Fig. 4.37 is used t o illustrate the use of the foregoing charts and formulas. Using the most critical feeders from the standpoint of voltage 33.5 KV TRANSMISSION L I N E 60 CYCLES OVERHEAD LINE

10,000 KVA PERCENT 2 = I.OPERCENTti6.0 PERCENT 34.400 -4160 VOLTS BUS A LOAD

9000 KVA 0.8 PF LAGGING 4160 VOLTS 3- CONDUCTOR 250 MCM

1500 KVA, 4160-480 PERCENT t = I . O PERCENT T t j 5 . 5 PERCENT 1300 aus KV A 0.8 a LOAD PF LAGGING 480 VOLTS

*-480

I I

VOLTS I 3 - k S o O MCM V C L I N CONDUIT

A

OHMS ?=0.0072tj0.009200 FEE1

W

250 KVA 440 VOLTS 0.7 PF LAGGING

FIG. 4.37 Calcdatio".

System one-line diagrom used

01 a

baris for examples of system voltage-drop

V O L T A G k S T A N D A R D RATINGS, VARIATIONS, CALCULATION OF DROPS

245

drop, four solutions involving varying degrees of accuracy were made to determine the operating voltage at the 4160- and 480-volt utilization buses and at the load end of a 480-volt feeder. In each solution except 4, it is assumed that the indicated load kva, power factors, and efficiency remain constant for voltage variatious due to regulation. I n other words, the load current varies with applied voltage to keep the kva constant. Table 5.12 lists the operating voltages obtained by the four methods of solution used. TABLE 4.12

1

Operating Voltages as Calculated by Four Methods

Solution

* Sce

I I

1 Equations' used

1 2

14.7) 14.31 end 14.41

3 4

Charts Charts

I i

Bur A

I I

Bur 8

Secondary feeder load

3900 3910/-2.2" -~

425 426/--4.40

418 419/~4~50

3925 3932

429 432

422 425

Eqs. (4.3), (4.4), and (4.7)

Solution 2 was made by using the exact formulas Eqs. (4.3) and (4.4).

It shows that the phase angle of each successive voltage level is shifted to lag slightly the no-load voltage. It should he recognized that the use of this exact formula does not necessarily mean that the answer is exact, because it is necessary to use a cut-and-try process in the solution. As with any cut-and-try process, a point is reached where the added accuracy to be obtained does not justify another trial, and therefore the answer is not absolutely exact. In solution 3, voltage-drop charts were used to determine voltage drop. The error involved in this method results from the greater margin of error in reading charts and in the arithmetical additiou of voltage drops slightly out of phase. Solution 4 involved the use of charts but neglected t,he cut-and-try procedure necessarily employed in the other solutions. The cut-and-try procedure was used in the other solutious because the load kva x a s assumed to be constant as the voltage changed and therefore t,he load currents changed. In this solution the current x a s assumed to remain constant as the load voltage varied. Solutioii 1 is given helow as an example. Solution 1: Calculatiou by approximate Eq. (4.7).


246

where v

I R X

e

line-to-line voltage drop line current, amp = circuit resistance, ohms = circuit reactance, ohms = load power-tactor angle = =

Bus A Voltage. From Fig. 4.37, Overhead line resistance = 1.97 ohms Overhead line reactance = 1.52 ohms Converting transformer per rent resistanre and reactance t o ohms by the formula Ohms =

%ohms X (kv)' X 10 kva

and using the principle that transformer impedance varies approximately as the square of the per cent voltage tap used, X (0.975)' K,. = 1.0 X (34.4)* 10,000

X 10

X (0.975)' X x, = 6.0 X (34.4)'10,000

Total ohms resistanre = 1.97 Total ohms reactance = 1.52

10

=

1,12 ohms

=

6.73 ohms

+ 1.12 = 3.09

+ 6.73 = 8.25

Assuming 4lFO volts at bus A and considering constant load, Bus A amperes =

9000 kva

=

Overhead line amperes =

1250 X 4.16 = 155 34.4 x ,975

Suhstitutingin the voltage-drop formula with cos 0

v

4X 4X

1250

4X 4.160 = 0.8 andsin

0 = 0.6,

+

155(3.09 X 0.8 8.25 X 0.6) 155(2.48 4.95) = X 155 X 7.43 = 1990 volts = =

4

Bus A volts = (source voltage

+

- voltage drop)

4.160 = (33,500 - 1990) 34,4 o,975

X (power transformer ratio)


Recalculating the voltage drop assuming 3910 volts on Bus A , Overhead line amperes Y

=

= 4160 -X

3910

155 = 165

4X 165 X 7.43 = 2120 volts

Bus A voltage = (33,500 - 2120)

4.160 34.4 x 0.9%

This value is assumed t o he close enough for practical purposes. Bus B Voltage. From Fig. 4.37, 5-kv cable resistance 5-kv cable reactance

= =

0.1094 ohm 0.0712 ohm

Transformer resistance

=

1.0% on its own base

Transformer reactance

5.5% on its own base X 10 1500 = 0.634 ohm =

5.5.X (4.16)' ~~

Total ohms resistance Total ohms reactance

= =

0.109 0.071

+ 0.115 = 0.224 ohm + 0.634 = 0.705 ohm

Assuming 450 volts on bus B , 1300 kva = 1670 0.450 480 5-kv cable amperes = 1670 X -= 193 4160 v = 4 I ( R cos 0 X sin 8 )

Bus B amperes =

&X

+

cos 0 = 0.8, sir1 0

v

0.6

4X

+

193(0.224 X 0.8 0.705 X 0.6) 43 X 193(0.179 0.423) = fi X 193 X 0.602 = 201 volts =

=

Bus B voltage

=

+

(bus A voltage - v) (transformer ratio) 480 = (3900 - 201) 4160 =

247


248

I{rcali~ulatrngI> a\suming 4"(i volts on bus H with same load,

450 5-kv cahlr amperes = 193 X 4%

v =

4 3

X 204 X O.GO2

Hns B voltage

=

204

=

212 volts 480 (3900 - 212) 4160 =

Secondary load voltage, assuming q20 volts a t load, 250 0.420 X Cable resistance = 0.0072 ('able reactance = 0.0090 Load amperes

v =

ros B

=

0.7,

sill B =

c = 4 3 = = =

=

=

4

344

4x r ( x cos B + X sin 8 )

0.714

x

344(0.007"

x

0.7

+ O . O O ~ Ox 0.714)

4X 344(0.00504 + 0.00643)

4 3 X 344 X 0.01147 G.9 volts

Load voltage = 425 - G.9 = 418.1 volts Since the most i.ritii.al feeders n-ith respect, t o voltage drop have been selerkd, the ralrulated load voltages a t hus A , bus B , arid at the secondary-load trrminals provide sufkieirt information t o analyze the system from the standpoint of voltage drop. Xct,ually, the 480-418 voltage spread at the serondary-load terminals iiidicates that the system is on the horder line and should he stiffened, possibly'hy using a larger 5-kv feeder cable. Howevw, this is beyond the scope of this problem, which i s mcrcly iriteiidcd to out,liiic t h e method of det,ermitriiig voltage drop. CALCULATION OF VOLTAGE DROPS DUE TO MOTOR STARTING INTRODUCTION

I t is rharactrristic of most a-c motors that the riirrent, which they draw startiirg is mu(.h higher t,han t,heir rrormal running ( w r e n t . Syni~hronousand sqnirrel-rape iudi~rtionmotors started 011 full voltage may draw a c u r ~ w i tas high as sevt!ii or eight t,imes their fnll-load running rurrcnt. This sriddeir increase in the (.usrent, drawn from the power system may r c s i i l t iii csressive drop i n volt,age unless it is considered in oii


249

the design of the system. Folloii-ing are methods for ralculatiug the voltage drop which results from startiug of three-phase induction aud synrhronous motors. M O T O R - S T A R T I N G METHODS

The motor-startiug kra, imposed on the power-supply system, and t,he available motor torque are greatly aKected by the method of starting used. l'ahlc 4.13 gives a conlparisoii of several common methods. Full-voltage Starting. This method usually provides the most torque hut muses the greatest load t o be applied to the system. The load applied equals (at motor rated voltage) the full-voltage starting kva of hhe mot,or. Frill xwlt,age is the least espeosive method of startiug. The full-voltage starting kva of syurhroiious and squirrel-cage indurtion motors ruuges from 230 to 800 per cent of their full-load h a input. The latter is approsimately cqual to t,hc horsepower rating of induction and 0.8-pomr-factor syirrhrorious motors and is approximately 80 per eelit of the horsepower rating of 1.0-poiver-factor syrichronous motors. If the starting curreut in ampercs is kno\vu, the startiug kva (of threephase motors) may he ralrulated from the formula Kva = 1.73 X amperes X

line-to-line volts

looo

Reactor Starting. With t,his method, a reactor is connect,ed in series with the motor aud is shorted out when the motor approaches full speed. 4 reactor starter redures the line current in proportion t o the tap used. For example, with a 50 per cent tap, the current is cut in half. The torque is reduced hy the square of the tap used. Hence, the torque is reduced more rapidly than the line current. Reactor st,arting is commouly used for large motor-generator sets. Resistor Starting. Resistor starting is similar t o reactor starting except that a resistor is used in series with the motor, instead of a reactor. The torque available for a given reductioti in startiug current is the same as with a reactor. The hie-voltage drop may be somewhat less because of the better power factor of a resistance-st,arting load. Resistor starting seldom offers a cost advantage, except wheu several steps are required, t o meet limitations established for the maximum kva applied at any oue step. Power companies sometimes establish such limitations. Use of several steps may permit a generat,or voltage regulator to restore voltage between steps. It also tends to make light flicker less noticeahle, even if most of the drop is in the distribution system and cannot be reduced by regulators.

250


Autotransformer Starting. If an autotransformer starter is used, the line current is reduced approximately as the square of the tap setting. For example, if an autotransformer at a 50 per cent tap is used, a motor starting load of 100 per cent of the rating of a generator will be redured to about 25 per cent. Table 4.13 shows 30 per cent because it allows for autotransformer magnetizing current. Autotransformer starting may cost more than reactor starting, but may be needed to provide adequate torque. The tap selected should always be high enough to accelerate the motor to a speed a t which the current will not be excessive after transfer t o the running connection. If the load torque is high a t the time of transfer to the line, a high transient inrush for a few cycles may occur at this time even if the speed is high. This is seldom sustained long enough to cause troublesome voltage dip, but may cause tripping of instantaneous overcurrent protection for the motor circuit.

TABLE 4.13

Comparison of Motor-starting Methods Line voltage = motor-rated voltage

Type of

starter* Motor voltage

line voltage

........................ ........................ ........................ ........................

Full.roltage stmrter. Autotransformer: 80 Per Cent t o p . . 65 per cent tap.. 50 per cent tap.. Resistor storm, single step [adjusted for motor voltage to be 80 per cent of line voltogel Reoctor; 50 per cent tap.. 45 per <*"I top.. 37.5 per Cent t o p . . Part-winding starter [low-speed m o l ~ r sonlyl: 75 per cent winding. 50 per cent winding.

....................... ........................ ...................... ..................... .....................

.I

t.rting

tolqY*

f"ll-"oltoge tarting torque

1 .O

1 .O

1 .O

0.80 0.65 0 .so

0.64 0.42 0.25

0.68 0.46 0.30

0.80

0.64

0.80

0.50 0.45 0.375

0.25 0.20

0.14

0.50 0.45 0.375

1 .O 1 .O

0.75 0.50

0.75 0.50

* The settings given %rethe more common for each type. Some motors can be provided with taps for part-winding starting. In such cases, power is first applied to a portion of the winding and later the entire winding is connected to the line. This is sometimes done in several steps, using increasing proportions of the winding. When only part of the winding is energized, the current and torque are Part-winding Starting.


251

less than for full-voltage starting. They are both changed approximately in proportion t o the amount of winding connected. That is, for a typical low-speed motor, at the half-winding connection, the current and torque are approximately equal t o one-half their full winding values. This method is comparable in cost with autotransformer starting, and also provides a smoother transfer to the running connection. However, this method is seldom advantageous for motors above 514 rpm (fourteen poles), because it provides relatively less torque for such motors. Starting of Wound-rotor Motors. Wound-rotor motors are invariably started on full voltage, but control is provided which inserts a high resistance in the secondary winding on starting and short circuits this resistance in one or more steps as the motor comes up to speed. This serves to limit the starting current drawn by the motor-usually to about 150 per cent of full-load current. Furthermore, this current will have a high power factor. Consequently, the voltage drop caused by starting this type of motor is comparatively small. On the other hand, wound-rotor motors and their control have a relatively high cost. TYPE OF VOLTAGE DISTURBANCE PRODUCED BY M O T O R STARTING

Generator Voltage. Figure 4.38 shows the behavior of the voltage of a generator when an induction motor is started. Starting a synchronous 100

90

80

$ 70

2 d" >

5 50 Y

u 40 IL

f 30 20

ov

10

0

2

3

4

5

6

TIME- SECONDS

MOTOR-STARTING XVd*IDDPfR

A B N

CENT OF

DENEMTOR RATING

-

NO INITIAL LOeiD ON GENERATOR

-

5 0 PER CENT INITIAL LOAD ON GENERATOR

- NO REGULATOR

FIG. 4.38

Typical generator voltage behavior.

151

V O L T A G S S T A N D A R D RATINGS, VARIATIONS, CALCULATION OF DROPS

motor is essentially similar, up to the time of pull in. I n the case illustrated, a full-voltage starter is used, and the full-voltage starting kva is ahout 100 per rent of t,he generator rating. I t is assumed that the generator is provided with an automatic voltage regulator. Curves .A and R show the performance, with the regulator operating, for init,ial loads on the generator of zero and 50 per cent, respectively. The minimum voltage is about 75 per cent and is not affected much by the iriitial load. This is typical with most initial loads which consist of a combination of lighting loads and partially loaded iuduction motors. The voltage regulator restores the voltage ton-ard normal in about 2 see. At, this time the motor is usually st,ill at low speed and drawing a high current. The initial load on the generator has an important effect on the value t o which the voltage is restored by regulator action. This is illustrated by curve B , for whirh the voltage is restored by the regulator to only about 85 per cent of normal. This restored voltage is the voltage available for breaking away and accelerating the motor. When the motor comes up to speed, its current becomes much less, so that t,he regulator then restores the generator voltage to 100 per rent. The reason the regulator usually cannot restore the voltage to 100 per cent when a large motor is started on a heavily loaded generator is that the exciter maximum (ceiling) voltage limits the available generator excitation. Sometimes it is only necessary to calculate the minimum voltage. In other cases it is also necessary to calculate the restored voltage available for break away and accelerations. Methods of estimating each of these voltages are included. Minimum voltage is needed to determine whether undervoltage devices and contactors connected to the system mill drop out, or running motors stall, during the disturbance. The minimum voltage is also a determining factor in light flicker. The restored voltage is necessary to estimate the torque available for starting the motor. Usually it is sufficient to determine the minimum voltage and the restored voltage based upon the current drawn by the motor at standstill, i.e., upon the locked-rotor current. It is sometimes necessary, however, to determine the restored voltage throughout the acceleration of the motor. Although the current drawn by a motor decreases as it comes u p to speed, resulting in an increasing generator voltage, the load torque may also increase with speed so that a higher voltage is necessary to ensure acceleration. In the case of a synchronous motor i t may be necessary to check the restored voltage at the speed at which field excitation is applied (95 per cent of synchronous speed or higher) to make sure that the motor will pull into step. The pull-in torque of a synchronous motor varies approxi-


253

mately as the square of the voltage at the motor terminals just before application of field. Distribution-system Voltage. Frequently there are transformers, lines, or cables between the motor starter and the generator or generators supplying the power for starting. The drop in the transformers, lines, or cables will be additional to the generator drop. Often practically all the drop is in this distribution equipment. The drop in this equipment is not reduced by the action of voltage regulators. Consequently, when practically all the drop is in transformers, lines, and cables, the voltage falls immediately and docs not rerover till the motor approaches full speed. ESTIMATING GENERATOR VOLTAGE DROP

Minimum Voltage. The curves of Fig. 4.39 may he used for estimating the minimum voltage occurring at the terminals of a generator supplying power to a synchronous or squirrel-cage induction motor which is being

254


started. The initial load on the generator, if any, is assumed to be of the constant-current type. The three sets of curves shown are for three ranges of generator speed. The generator reactances assumed to apply for each speed group are also given in Fig. 4.39. The curves show the minimum voltage, in per cent of the initial generator voltage, plotted against the "motor-starting kva" in per cent of generator rated kva. The "motor-starting kva" is the kva which would be drawn by the motor being started if the generator voltage were maintained at rated value. Since there is a drop in generator voltage, the actual kva drawn by the motor will generally be less than the value defined above, but the effect of this is taken into account by the curves. The several curves in each speed group-except those marked N a n d Eapply for various values of a factor K. This factor is the exciter response in volts per second divided by the exciter voltage for rat,ed generator voltage at rated load and multiplied hy the generator open circuit field time constant in seconds. Approximate values of K are given in Fig. 4.40. The values of Ii in Fig. 4.40 are based on the use of a self-excited excit,er controlled by a direct-acting rheostatic voltage reguhtor (such as the

GENEIIbTOR e I T E O K"& ~

W

T DIRECT-CONNECTED " EXCITER

.....~ WIT" ~ BELTED ..

EXCITER NUMBERS ON CURYES ARE R P N

NUMBERS I" BRACKETS &RE EXClTER R P H

*

FOR "IMIAITION OF EXClTER RESPONSE WlT" GENERATOR IN1TIAL L o l o lNlTlbL LOAD (PER C E N T , UULTlPL" I( B" (00 ,70 75 I55 50 I"5 25 I25 HULTlPLlERS TO *ILLOW

0

FIG. 4.40 Typical valuer of performance factor K for (I-c

,oo generators.


255

A-C GENERATOR

EXCITER FIELD

GENERATOR VOLTAGE REGULATOR

FIG. 4.41

1

u

''lL

Excitotion system for a-c generator.

General Electric Company Type GDA) as shown in Fig. 4.41. In this system the response of the exciter depends not only upon its design hut also on the setting of the exciter field rheostat. The latter is determined by regulator requirements. The rurves of Fig. 4.40 are based on a setting of the exciter field rheostat which makes available a maximum generator field current of 120per cent of its rated value. The K fartors given by the curves are typical only, and in an individual case K may vary considerably from the value shown. The curves of Figs. 4.39 and 4.40 allow an estimate to be made of the generator miuimum voltage directly from the generator kva rating, the generator speed, the exciter speed, and the motor starting kva. If guarantees of performance are required, a study based on romplete data should be made considering the characteristics and adjustments of generator, exciter, regulator, exciter rheostat, initial load, and the motor being started. Restored Voltage. The curves of Fig. 4.42 may be used for estimating the restored voltage of a generator, that is, the voltage attained after the regulator has acted to apply maximum excitation current to the generator (or has restored the voltage to its initial value) following the starting of a squirrel-cage induction or synchronous motor. The curves show the restored voltage in per cent of rated generator voltage plotted against the kva which would be drawn by the motor being started if rated generator voltage were maintained. The several curves apply for various values of initial load which is assumed to be a constantcurrent load of 0.8 lagging power factor. The excitation system is assumed to be such that a maximum excitation current of 120 per cent of rated generator field current can he obtained. If guarantees of performance are required, a study based on complete data should be made considering the characteristics and adjustments of generator, exciter regulator, exciter rheostat, initial load, and the motor being started.

156


MOTOR S T I R T l N G

I("&

OF

DROPS

I N P E R C E N T O F G E N E R l T O R K Y A A T R A T E O O E N E R ~ T O AVOLThQE

( B A S E D O N Y l x l Y u Y EXCITATION-IZOPER C E N T O F R A T E D G E N E R A T O R F l E L O C U R R E N T I NOTE: RESTORED VOLTAGE E W I L S VALUE READ FROM CURVE OR THE INITIaL YOLTlGL (REGULATOR SETTING1 WHICHEVER I S LOWER

FIG. 4.42

Restored generator vollage.

Advontages of Voltage Regulators. Figures 4.38, 4.39, and 4.42 show dashed curves, marked N , which indicate the results t o be expected if no regulator is used. It is apparent that regulators are very beneficial. They practically always justify their cost whenever the starting of large motors is involved. For example, consider a 480-volt 125-kva 1200-rpm generator. From Fig. 4.40 this may have a performanre factor K of about 1.7 with a regulator. From Fig. 2.39, 110 per cent motor-starting load or 138 kva will cause a 28 per cent voltage dip. This load would correspond t o starting a 25-hp motor at full voltage. T o obtain the same motor-starting performance without a regulator would require a 438-kva generator, because the curve N shows that about 32 per cent motor-starting load will cause a 28 per cent voltage drop if no regulator is used. (138 kva is about 32 per cent of 438 kva.) The 438-kva generator would cost over twice as much as a 125-kva machine. The best and least expensive arrangement mould be t o provide a regulator adding less than 15 per cent to the cost of the 125-kva generator. This mould permit successful starting of the 25-hp motor even against full-load torque and would improve normal generator performance. I n Fig. 4.39 are curves, marked E , which show the performance available when using an electronic exciter or some other very high-response excitation system. It shows there is a definite limit to the improvement

VOLTAGE-STANDARD RATINGS, VARIATIONS. CALCUUTION OF DROPS

157

which can be obtained by greatly inrreasing response; that is, the generator voltage will dip a t least a certain amount before the excitation system can do anything about it. Effect of Initial Voltage. Often the voltage rating of the generator supplying a motor is higher than that of the motor. A 440-volt motor might he supplied by R 480-volt generator and a 2200-volt motor by a 2400-volt generator. In such cases, the motor-start,ing kva should be adjusted t o take this into account,. The kva drawn hy a motor increases as the square of the line voltage. If t,hr startiiig inrush of a 410-volt motor is 1000 kva a t 440 volts, it will be 1190 kva at 480 volts because (480/440)* = 1.19. This is the value which should be used to determine the generator minimum voltage (from Fig. 4.39) regardless of the actual initial voltage. For example, assume that, with an initial voltage of 480 volts, the starting of the 440-volt motor (drawing 1190 kva at 480 volts) causes the voltage t o drop t o 75 per rent of the initial value, or 3G0 volts. If the voltage regulator is set t o hold a voltage of 440 volts, starting of the same motor will produre approximately the same voltage drop in per cent of the initial voltage, i.e., the voltage will drop t o approximately 75 per cent of 440 volts, or 330 volts. This shows that, from the standpoint of the minimum voltage, the regulator should be set t o maintain rated voltage on the generator even though the motor voltage is lower. As far as the restored voltage is concerned (Fig. 4.42), this is not affected by the initial voltage except that the voltage mill not recover t o a value higher than the initial voltage since this represents the setting of the voltage regulator. For example, if the initial voltage (setting of voltage regulator) is 90 per cent of rated generator voltage, the recovery voltage in per cent of rated generator voltage will be as shown by the curves of Fig. 4.42, except that all curves will become horizontal lines at 90 per cent voltage. Effect of Initial load. The voltage curves of Figs. 4.39 and 4.42 were prepared on the basis that the initial load on the generator draws constant current duririg the voltage disturbance. This sort of load characteristic is representative of many systems and results from the use of induction motors, all of which are not fully loaded. An induction motor at no load will draw a current approximately proportional t o the applied voltage, because the current is principally magnetizing current. A fully loaded induction motor will tend t o have constant kva input since its speed and power factor do not change much with variations in line voltage. Consequently, a fully loaded induction motor will draw more current if the voltage is lower, t o maintain the power constant, A system load consisting of both heavily loaded and

258

VOLTAGbSTANDARD RATINGS. VARIATIONS, UL6UUTION OF DROPS

lightly loaded motors will therefore tend to draw nearly constant current since a lowering of the voltage causes a reduction in the current to some motors and an increase in the current to others. A constant-current type of load will have very little influence on the minimum voltage during motor starting. It will, however, have an important effect on the value of the restored voltage of generators, as previously described. Lighting loads usually have little effect upon voltage disturbances due to motor starting. This is true because lighting loads usually constitute a small proportion of the total load on a generator, and also because of their high power factor. If the system load consists primarily of lightly loaded induction motors, the per cent minimum voltage and recovered voltage will both tend to be higher than indicated by the curves. If the initial load consists entirely of heavily loaded induction motors, the voltage disturbance from motor starting will be more severe than indicated by these figures. Initially connected synchronous motors are beneficial in reducing the disturbance due to motor starting. They are most beneficial when lightly loaded. Therefore, it is helpful to start synchronous motors first in a plant so that they will be on the line to help in the starting of large induction motors later. Synchronous motors will not be helpful, however, if the voltage disturbance is so great as to cause them to pull out of step. Although the curves in this section are based on initial loads of the constant-current type, they may be used for cases involving other types. This is done by adjusting the motor-starting kva by an amount corresponding to the change in current to the initial load, caused by the drop in voltage. The increase or decrease in motor-starting kva is such as to change the motor-starting current, a t the minimum voltage, by the same amount as the change in the lagging wattless component of the initial load. That is, the effect of the initial load is primarily due to a change in the wattless component, and this can be simulated by a change in the motor-starting kva. Since the change in current and the minimum voltage are dependent upon each other, a trial-and-error procedure is involved. The first trial is often sufficient,if the change in current is determined a t the voltage corresponding to the case of a constant-current initial load. For example, consider a generator whose voltage would dip to 75 per cent if a 100 per cent motor-starting load were applied when a 50 per cent constant-current initial load is being carried. If, instead, the initial load consisted of fully loaded induction motors a t 0.8 power factor, the dip would be more severe, because a t 75 per cent voltage the lagging wattless current to the running motors would be increased from 30 per cent of the


259

generator rating t o about 40 per cent. This increase could be approximately simulated by an increase of the motor-starting kva from 100 per cent to 113 per cent. This is true because a motor-starting load which would draw 13 per cent of generator rated kva a t full voltage will draw 10 per cent current a t 75 per cent voltage. Figure 4.43 shows the amount by which motor-starting kva should be increased to allow for the effect of an initial load consisting of fully loaded induction motors. Effect of Starting Power Factor. The power factor of most motorstarting loads lies between 10 and 40 per cent. Variations within this range do not materially influence voltage drop of generators. Wound-rotor motors have a starting power factor of about 80 per cent lagging. At this power factor the resulting voltage drop (initial voltage minus the minimum voltage) will not generally exceed 75 per cent of the drop caused by the same kva a t low power factor. Resistor starters

__

PF -R C F N T VOLTAGE

DROP

18 17 U

> Y

I !

g

10 09 08

20

13

15

12

'

07

2

03

LL

06

a w

04

1

25

14

z0 2

F

30

16 15

10 5

05

02

3

01

=

o

0

0'2

04

06

08

RATIO OF

10

12

14

16

18

20

I N I T I A L LOAD KVA MOTOR STARTING KVA

INCREASE MOTOR STARTING KVA BY MULTIPLIER SHOWN BEFORE USING CURVES OF FIG.4.39AND FIG.4.42 ( ! N I T I A L LOAD MAY THEN BE CONSIDERED AS CONSTANT CURRENT T Y P E )

FIG. 4.43

Approximate effect of initial lood consisting of fully loaded induction motors.

260

VOLTAGE-STANDARD

RATINGS, VARIATIONS,

CALCULATION OF DROPS

seldom cause the starting power factor to he high enough to reduce voltage drop greatly, except for the first steps when several are used. Effect of Drop in Generator Speed. Since the power factor of motorstarting kva is low, the amount of kw load applied to a generator is seldom large. Furthermore, the voltage drop, by reducing the electrical output, also reduces the new load applied. For example, a motor-starting load of 100 per cent of generator-rated kva at 0.3 power factor will involve a suddenly applied km load less than 30 per cent of rated kva, or 37.5 per cent of rated kw for an 0.8 power-fartor generator. The speed drop is not likely to be excessive if good governing means are employed. For most motor-starting problems, it may safely be neglected. As speed dips, a corresponding dip appears in the voltage, which is in addition t o the voltage drops considered in this section. For cases where speed dip may be sufficiently great to be important, this should be considered, but calculation of speed drop is beyond the scope of this book. ESTIMATING DISTRIBUTION-SYSTEM VOLTAGE DROP

The voltage drops in lines, cables, and transformers are often as important as generator voltage drop. In fact, they are frequently more important. For example, if the total kva of connected generators in the power system is more than 100 times the horsepower rating of the motor being started, then the generator voltage dip will be less than 1 per cent, and it will be quickly eliminated by regulators. In such a case, however, the motor will probably be supplied through a transformer bank. If the transformer-bank kva rating is only slightly larger than the motor rating, the voltage drop may be quite severe. Voltage Drop of Transformers. The curves of Fig. 4.44 may he used for estimating the voltage drop through typical transformers when starting a synchronous or squirrel-cage induction motor connected to the secondary of the transformer. The secondary voltage on starting of the motor, in per cent of the initial secondary voltage, is plotted against the motor starting kva. The latter is expressed in per cent of the transformer-hank kva rating and is the kva which wouldhedrawnhythemotor being started if rated transformer secondary voltage were maintained. The curves of Fig. 4.41 neglect the effect of primary-voltage drops caused by motor starting. Methods of taking these into account will he explained later. Note that the secondary voltage is plotted in per cent of its initial value. This initial secondary voltage is determined by the initial primary voltage, the t a p setting, and the initial load. It may he determined by measurement or by suitable calculations. It is usually slightly less than the rated secondary voltage.

VOLTACbSTANDARD RATINGS, VARIATIONS, CALCULATION OF DROPS

261

MOTOR STARTING KVA

1% OF B & N I K V A aiT RATED TRANSFORMER SECoNOAR” VOLTAGE1

FIG. 4 44 Transformer secondary voltage

The curves of Fig. 4.44 were prepared on the basis that the initial load, if any, draws constant current during the voltage disturbance. This is typical of a system consist,ing of both lightly loaded and heavily loaded inductiou motors. If the initial load consist,s largely of fully loaded induction motors, the curves of Fig. 4.44 may still be used provided that the motor-starting kva is first multiplied by the fartor shown in Fig. 4.43. The curves of Fig. 4.44 apply for motor-starting power factors in the usual range of 10 t,o 40 per cent. For wound-rotor motors which have a starting power factor of about 80 per cent, the drop in voltage will be about 70 per cent of that shown. Voltage Drop of Cables and Overhead Lines. The curves of Figs. -1.45 and 4.4G may be used for estimating the voltage drop through cables and overhead lilies n-hcn start,iiig synchronous and squirrel-cage induction motors supplied through these circuits. I n using these figures, it is first necessary t o determine the length of the circuit in feet, the initial voltage at the load end of the circuit, and the motor-starting kva a t the iuitial voltage. These quantities are combined to obtain the loading factor .If as follows: motor-starting kva at the initial voltage x (% ci:ri ne):,t M = (initial voltage)2

)

262

VOLTAGkSTANDARD RATINGS, VARIATIONS, CALCULATION OF DROPS

For example, if the motor-starting load were 1000 kva, the circuit 1000 ft long, and the initial voltage 2400 volts, the loading factor M would be

1000 x 1000 = o,1,3 (2400)' Figure 4.45 shows that for this case the voltage drop at the load end of a typical three-conductor cable is 1.5 per cent. This illustration gives data for three circuits: a three-conductor cable, a single-conductor cable, and an overhead line. It will be noted that the voltage drop in an overhead line is greater than that for a cable. If two circuits are in parallel, the drop is equal to that for a single circuit of one-half the actual length of each circuit. The voltage drop in a line or cable depends upon the conductor size and spacing. Consequently, for different cases than those illustrated in Fig. 4.45, the voltage drop may be somewhat different. This is illustrated by Fig. 4.46 showing the voltage drop for a range of circuit configurations. The points corresponding to the circled cases in Fig. 4.45 are circled in Fig. 4.46. Figure 4.46 applies for the condition hf = 1.0. It may be noted, however, that the curves of Fig. 4.45 are nearly straight lines. Hence, the voltage drop for other values of M may be estimated by multiplying the values of Fig. 4.46 by M . This provides a simple method of estimating the voltage drop for motor-starting loads. The power factor of the motor-starting load is assumed to be 0.3 power factor. For conductor sizes above No. 0 Awg, variations over the usual range from 0.2 to 0.4 power factor will not have an important effect on voltage drop. Figures 4.45 and 4.46 are based on a frequency of 60 cycles per sec. Lines and cables for systems operating a t lower frequencies mill have less voltage drop. The voltage drop will be reduced approximately in proportion to the frequency for all couductor sizes above KO. 0 Awg. For smaller sizes, the reduction will he less. Voltage Drop of Reactors. The voltage drop in a current-limiting reactor on starting a squirrel-cage induction or synchronous motor may be estimated from the transformer curves of Fig. 4.44. Current-limiting reactors are usually described as having a certain per cent reactance on a specified system-kva and syst,em-voltage base. The motor-starting kva of Fig. 4.44 should be that drawti at the specified system voltage expressed in per cent of the specified system kva. If the per cent reactauce of a reactor does not lie between 5 and 8 per cent, multiply the motor-starting kva by the ratio X / 5 , where X is the actual per cent reactance of the reactor, and read the voltage corresponding to this equivalent motor-starting kva on the 5 per cent reactance curve.

V O L T A G E - S T A N D A R D RATINGS, VARIATIONS, CALCULATION OF DROPS

263

0

100

" Y

90

10

80

20

70

30

0

0 0.1 02 0.3 0.4 05 0.6 0.7 0.8 0.9 1.0

LOADING FACTOR, M =

~

~

1.1

Q

1.2 1.3 1.4

( L E N G T H IN FT.)

~

(A) 3- CONDUCTOR C A B L E - (NO. 4 / 0 - A W G - I 5 K V )

(6) I - CONDUCTOR C A B L E - ( N O . 4 / O - A W G - 6

IN. SPACING)

(C) O V E R H E A D L I N E - N 0 . 4 / 0 - A W G - 1 5 K V CIRCLED P O I N T S APPEAR O N FIG. N O 4 46

FIG. 4.45

Variation of voltage drop with looding factor M for typical liner and cables.

c:

CONDUCTOR DIAMETER (INCHES) FOR MOTOR-STARTING LOADS OF 0 3 POWER FACTOR LO~DING FACTOR M:

(MOTOR-STARTING KVAI (LENGTH IN FEET1 (INITIAL VOLTAGE)^

*FOR FL4T SPACING, EQUIVALENT TRIANGULbR SPACING'; ADJACENT PHASES

FIG. 4.46

I0

I 2 6 TIMES SPACING BETWEEN

Voltage drop in lines and cables with loading factor M of unity.

264

V O L T A G b S T A N D A R D RATINGS, VARIATIONS, CALCULATION

OF DROPS

Effect of Series Capacitors. Sometimes it is advantageous t o include series capacitors in the distrihut,ion system t o neutralize the reactance of lines, cahles, or t,rausformers. Series capacitors redure voltage drop. The amount of redurt,ion depends upon the raparitor rating. For further informat,ion on series capacitors, refer to Chap. 8. Voltage Drop of Power Systems. Motors are frequently supplied from power systems cotisistirig of complicated uetworks of lines and cables for which a calculation of the voltage drop ~vouldhe difficult. The voltage drop may be est,imated, however, if t,he short-circuit current is known at the point of power delivery. The short-circuit rurrent is usually expressed in kva. When motor-starting kva is drawl from a system, the voltage drop in per cent of the initial voltage is approximately equal to 100 times the motor-startiiig kva divided by the sum of this kva and the short-circuit kva. The motor-starting kva used should be that drawn by the motor if the initial system \&age were maintained. For example, if a 1000-hp motor has a startirig kva of 5000 if initial system voltage were maintained and the system short-cirruit kva is 50,000, the voltage drop will be approximately

5000/(5000

+ 50,000) X 100

=

9 per cent of the initial voltage

In many systems the short-circuit kva varies over a wide range, depending upon the number of parallel h e s that are in service, system interconnections, etc. In such cases the highest short-circuit kva is the one usually determined since it must he the one used in selection of equipment which is t o carry or iritcrrupt the short-circuit current. For calculating voltage drop, oil the other hand, the minimum short-circuit kva should be used since the corresponding operating condition will give the highest voltage drop. The short-circuit kva of power systems varies over a wide range, as shown in Table 4.14. A corresponding variation occurs in the voltage drop produced by a certaiu motor-starting kva. TABLE 4.14

Power-system Short-circuit Kva Usual Range of

System Voltage 2,400 4,160 6,900

Short-circuit Kvo 15.000-1 50,000

25.000-250.000

50.000-500.000

13.800

100.000-1,000,000

23,000 34,500 69,000

150.000-1,500,000 150,000-I,500,000

I 15.000

250.000-2.500.000

I50.000-1,500,000

The method of calrulating voltage drop given above is not applicable at system locations where the short-circuit kva would be appreciably

VOLTAGE-STANDARD RATINGS, VARIATIONS, CALCULATION DF DROPS

265

affected by reactance of generators. I t should be used only when the impedances of transmission lines, transformers, reactors, and cables largely determine the short-circuit current. COMBINED VOLTAGE DROP

Series Circuits. Often a motor is supplied through cables, transformers, overhead lines, and generators, all in series. In such cases, the total voltage drop may be roughly estimated as the sum of the voltage drops given by the foregoing illustrations for each of the different parts of the system. However, the simple addition of voltage drops is not quite accurate because addition of impedance in series tends to diminish the current supplied to the motor. For more accurate work, the following procedure is suggested: 1. Determine the voltage drop in the circuit element nearest the motor, neglecting the other elements. For example, for a motor supplied from a generator, transformer, and cable in series, determine the drop in the cable first. 2. Multiply the motor-starting kva by the ratio of the load-end voltage to the initial voltage of the cable just determined. 3. Using this new value of motor-starting kva determine the voltage drop in the next circuit element. In the example selected, this is the transformer drop. 4. Now multiply the motor-starting kva by the product of the ratio of the load-end voltage to the initial voltage of the cable and the ratio of the secondary voltage to the initial secondary voltage of the transformer. 5. Using this new value of motor-starting kva determine the voltage drop in the next circuit element. In the example selected this is the generator voltage drop. 6. Continue the process until all elements in series have been considered. 7. Multiply the initial voltage a t the motor by the product of the final to initial voltage ratios of all the circuit elements. This result is the final voltage a t the load. An example a t the end of this chapter illustrates the procedure described. Parallel Circuits. If several sources are in parallel, the voltage drop is less than if the motor-starting load is supplied through any one of them. To determine the combined voltage drop, it is suggested that groups of similar generators may be treated as a single generator having the same total kva rating and the same performance factor as the individual machines. Transformer banks may also be grouped if they are supplied from the same primary bus and have the same per cent reactance and the same tap settings. To find the combined voltage drop for several parallel sources of different characteristics, it is suggested that the motor-starting load first

VOLTAGLSTANDARD RATINGS, VARIATIONS, CALCULATION

266

OF DROPS

be divided equally and the corresponding voltage drops determined. Then a new trial division of load can be made so as to increase the proportion of load carried by the sources with the least voltage drop. Usually oiily one or two trials are required to obtain a sufficiently accurate result. For example, consider the case of a motor which has a startiug kva of 1000 and is furnished with power by a 500-kva generator and a 300-kva transformer bank. The first trial division of load will he 500 kva each. Let us assume that this results in a minimum voltage of 75 per rent on the generator and 90 per cent on the transformer secondary. This means that the generator will actually accept less thari half the load. The drop in the generator is 2.5 times as great as in the transformer. Then assume that the transformer accepts 2.5 times as much load as the generator. This results in 285 kva being accepted by the generator and the remainder, 715, being imposed on the transformer (715 is 2.5 X 285). The voltage drop in the transformer for 715 motor-starting kva will be found to be practically the same as for 285 motor-starting kva applied to the generator. The drop obtained is the combined voltage drop. For the case illustrated, this voltage drop is about 14 per cent. A final check of the amount of voltage drop through each source is advisable, because the drop in a generator does not always vary directly with the amount of motor-starting load applied to it. This is especially true of the restored voltage obtained through the action of voltage regulators. FORMULAS FOR CALCULATING VOLTAGE DROP

The various curves and other data that have been presented allow estimates of the voltage drop due to motor starting to be made quirkly with minimum iuformation on the motor and circuit elements involved. For cases not adequately covered by these data, the formulas given below may he used. Static Circuit Elements Only. First assuming that all the voltage drop occurs in static circuit elements such as transmission lines, cables, transformers, and reactors, the voltage at the motor starter mill he equal to

+

Z.W

d(ttMRd2 where Z,

+ ( X , + Xd* X initial voltage at motor starter

(4.9)

impedance of motor being started (ratio of applied voltage to current drawn) R, = z, cos a, X , = Z, sin eM =


cos B.,

267

power factor of current drawn by motor being started total resistance of circuit, between motor and point in system where voltage remains constant, i.e., is not affected by start.ing of motor X s = hotal reartance of circuit between motor and point in system where voltage remains constant The impedance, resistances, and reactances in the above formula should all he expressed in ohms or all in per rent (or per-unit) on any convenient kva and voltage base. The mot,or impedance Z.,, expressed in ohms is

Rs

=

=

4

Voltage rating of motor in volts X starting current in amperes a t rated motor voltage

(4.10)

If a reduced voltage type starter is used, the starting current is that drawn from the line with rated motor voltage on the line side of the starter. Similarly, cos Ox is the poiver factor of the current drawn from t,he line. The voltage at the starter must, be multiplied by the motor voltage-line voltage ratio of the starter (see Table 4.13) to obtain the voltage at the motor t,erminals. The resistance and reactance of a transformer hank ran he expressed in ohms by multiplying its per cent resistance and per cent reactance, respectively, by (Secondary voltage rating in kv)2 X 10 Kva rating of bank

(4.11)

Circuit elements separated from the motor by a transformer should have their actual resistance and reactance values in ohms multiplied hy the square of the no-load voltage transformation ratio, that is, by (4.12) before adding to the ohmic resistances and reactances of the motor and other circuit elements on the serondary of the transformer. If two or more transformers are in series between the circuit element and the motor, the actual resistance and reactance in ohms should be multiplied by the square of the product of the various no-load voltage transformation ratios. For transformers equipped with taps 011 either primary or secondary winding, the voltage ratings used in the above formulas should correspond to the t a p setting. Using the per-unit system, it is generally convenient to select as base kva the kva drawn by the motor at rated motor voltage, which is X starting current in amperes X rated motor volts

1000

(4.13)

268

VOLTAGLSTANDARD RATINGS, VARIATIONS, CALCULATION

OF DROPS

and select rated motor voltage as the base voltage. Iii this case Z , = 1. The per cent resistanre and reactance of a transformer, with the motor connected t o its secondary, should be multiplied by

f Motor-starting kva at rated motor voltage \ G t i n g of transformer

\

)

Isecondary voltage ratiug\' of transformer rated motor voltage

\

1 (4.14)

A second transformer in series would have its per cent resistance and reactance multiplied b j the above expression and also bj' the square of the no-load volhage transformation ratio (secondary voltage divided by primary voltage) of t,he first transformer. The resistaiice and reactaiice of circuit elements that are expressed iii ohms should be multiplied hy

Motor starting kva at rated motor voltage ." (Rated "r volts ' x 1000

)

(4.15)

except where the circuit element is separated from the motor by a transformer, in vhich case the multiplier is Motor-starting kva at rated mot,or voltage Prjmary voltage r a h g of transformer rated motor volts -x Secondary voltage rating of transformer 1000

(~~__

) x 1000 (4.16)

If t v o or more transformers are in series bet,ween the circuit element and the motor, the transformer no-load voltage ratio which appears in the above espression should be replaced by the product of the no-load vokage transformation ratios of the various traiisformers. Where voltage taps are provided on a t,ransformer, the voltage ratirigs used in the above formulas should correspoiid t o the t a p sett,ing. The resistance and reactance of circuit elements connected in series can be added directly. For circuit elemeots connected in parallel, equivalent wlnes of resistance and react,ance can be det,ermined hy the method given in Chap. 1. If current to other loads is flowing in one or more of the circuit elements between the motor and the const,ant voltage point mhen the motor is started, the above formula for voltage a t the motor mil1 still apply, assuming that these other loads are of the constant-current type, i.e., the current drawn does not change ivhen the voltage drops. Such load currents must, of course, be considered io determining the initial voltage at the motor starter. A method for taking into account loads whose current varies u i t h voltage will be given later. Often it is desirable t o know the effect of motor starting on the voltage

V O L T A G b S T A N D A R D RATINGS, VARIATIONS. C A L C U U T I O N OF DROPS

269

a t various points in the system as well as a t the motor. The voltage a t the motor starter divided hy rated motor voltage and multiplied hy the current drawn at rated motor voltage gives the actual current drawn from the line. This current can be comhined with any load current flowing through the various circuit elements and the voltage a t any point calculated hy the methods given earlier in this chapter. For the case where motor starting current only flows in the circuit elements between the motor and a point in the system, the voltage a t this point will he equal to

d(Rw d(RX

+ + Rs)*+ (x.w (X, + + R i ) 2+

where R I Xi

x1)2 X XS)2

initial voltage a t motor starter

(4.17)

resistance of circuit betweeo motor starter and specified point reactance of circuit between motor starter and specified point R a , X,, Rs, and X s are as previously defined If any load drawing current through the circuit elements in series with the motor is not of the constant-current type, the voltage a t the motor starter can still be calculated hy the formula given provided that the initial voltage a t the motor starter is calculated using the current drawn by the various loads aft,er the motor is started. Since these currents will depend upon the voltage drop occurring when the motor is started, a trial-and-error solution is necessary. Thns the voltage a t the various loads eaii first be estimated from calculations based on ali loads drawing a constant current. The current drawn by each load a t the estimated voltage is used to calculate a new value of initial voltage a t the motor starter from which the voltage a t the motor starter and a t the various loads can be recalculated. If the load voltages do not agree closely with those estimated, nem estimates can be made and the process repeated. In many cases the voltage drop can he caleulated with little error, considering only Lhe reactance of the circuit elements in series with the motor and using the formula =

=

Voltage a t motor starter

where X s

=

zx z.w + x s x

initiai voltage a t motor starter (4.18)

total reactance of circuit betmeen motor and point in system where voltage remains constant Z, = impedance of motor heing started When the reactance-to-resistance ratio of the eircuit elements (X,/Rs) is 2 or greater, this formula gives a voltage drop which is generally within 10 per eent of the correct value. Transformers rated 100 kva or larger usually have a reactince-to-resistance ratio greater than 2. =

270

VOLTAGF-STANDARD

RATINGS, VARIATIONS. CALCULATION OF DROPS

Effect of Generators. h’ext consider the case ivhere generator voltage drop as well as the voltage drop through static circuit elemeiits must be considered. If there is no initial current flowing through the circuit elemeiits mhen the motor is started, the voltage a t the gerierator terminais may be determined from the curves of Figs. 4.39 aiid 4.42 using a value of “motor-startiiig kva at rated geiierator voltage” equal t o

Starting kva drawn by motor if voltage at motor starter mere maintained at the initial value X

Z I,

d(fi,, + n,)z +

rated geiierator voltage

(x,“+ xsjz

)

x(.initialgeiieratorvoltage ..

(4.19)

where Z,,,, R , , and X , are as previously defined Ra = resistance of circuit betweeii motor starter and geiierator terminals X, = reactaiice of circuit betmeeii motor starter aiid geiierator termina Is The pomer factor of the current drawn from the generator will equal (4.20)

Haviiig dctermiiied the voltage a t the generator termiiials, the voltage at the motor starter cari be calculated as it xill equal

zw

d(&, + K s ) ? + (XI, +

XS)*

voltage a t generator _ terminals _ X initial ~ motor voltage initiai generator voltage

(4.21)

If currents t o other loads (of constant-current type) are floniiig through the circuit elements mhen the motor is started, the voltage drop may be determined by trial aiid error. The formula gireri above, Eq. (4.19j, for motor-starting kva a t rated generator voltage may be used for the first estimate and the correspoiiding value of generator voltage determiiied. From this the voltage at the motor starter may I i c calculated. It is equal t o /initial voltage at motor\ starter which mould apz >< (4.22) pear if generator voltage drop had already occurred Having the voltage a t the motor starter, the kva drawn by thc motor caii be calculated. The equivalent motor-starting kra a t rated generator voltage wili equal the actual kva drawn by the motor multiplied by

V O L T A G L S T A N D A R O RATINGS. VARIATIONS, CALCULATION OF DROPS

271

(Rated generator voltage)z Actual voltage at motor starter x actual voltage at, gcrterator

(4.23)

~~

~

~

If there is a transformer between the generator arid the motor, the vollagc a t the mot,or starter should be multiplied by the no-load volt,agc trausformation ratio (primary voltage ratiiig divided by secondary voltage rating) of the transformer before suhstitutiiig it in the above formula. With t,wo or more transformers in series, use as a multiplier the product of their no-load voltage transformation ratios. If the calculated mot,orst,arting kva a t rated generator vokage differs appreriably from the first estimate, a serond estimate based on the calculated value can be made and the calculatioiis repeated until a close rherk is obtained. Motor-starting Power Foctor. Use of the preceding formulas requires a knoivledge of the motor-start,ing power fartor ((WS 8.,,). The starting power factor of squirrel-cage induction and synchronous motors varies over a rather wide range, depending upon the rating and desigii characteristics. Approximat,e values of starting power factor for typiral squirrel-rage induction motors are given in Fig. 4.47. Low-speed (450 rpm aiid below) synchronous motors for reriprovatirrg compressor drive usually have a start,itig p o m r fartor bet,ween 0.20 aiid 0.40. Synchronous motors for rrntrifugal pump drive, on the other hand, have starting power fartors generally between 0.15 and 0.35. Where motor-start,ing power factor must be kuo\vn more acrurately, a value should be ohtailled from the motor manufacturer. With reduced voltage starting, the p o m r factor of the rurreut drawl from the line may be somewhat different from the motor-starting power factor. An autotransformer starter has oiily a small effect on the porver fact.or, but the magnetizing current of the autotransformer makes the power factor of the current drawn from the line slightly less t,han the motor-starting p o w r factor. With a reactor st,arter, the power factor

-"

0.70

50.60 0

-=

0.50

A

0.40

0.M

=w 0.20

B

O.I0 0.001 5

FIG. 4.47

I

K)

I

I5

I

20

I l l

I

I

30 40 50 75 100 HORSEPOWER RATING

1

1

150 M o

I

300

I

500

I

I

700 1000

Approximate 3tor:ing power factor of typical squirrel-cage induction motors.

272

VOLTAGE-STANDARD

RATINGS. VARIATIONS. CALCULATION OF DROPS

of thc. riirri.iit dmwi from t h r liiie \ri11 eqiial the motor-startiiig power f w t o r miiltiplird Iiy thi: volt,age ratio (motor volt,age divided hy liiie voltage) of t h r startcr. .i rc:sist,»r starter, oii the other haiid, results i i i a power fartor for t h r riirreiit drawii from the liiie equal t o

To illiistrate, assume t h a t a motor Iiaviiig a startiiig poirer factor of 0.30 is providrd with a resistor starter dcsigtied t o reduce the voltage applied t o the mot,or t o íi3 pcr cciit of ratcd motor voltagr. Thc p o w r factor of t h r (wrrriit drawii hy tliis motor-start,cr combiiiatioii \vil1 iie ~~

-\/I-

~(0.íi5)2X [1--(0.30)*]

=

0.785

REDUCED-FREQUENCY STARTING

Ociasioiially i i i ordcr to start a largc motor, t,he system frequeiicy is rcduccd to a Ioiv valiic i i i ordcr to iiirrcasc t,hc ratio of tlie motor torquc to thc motix-startiiig ciirreiit. At rcduced frcqiiciicy the applied volt,age is l o w r , liiit i i i thc iisual applicatioii of tlic schcmc, thc applied voltage is rrdiii,rd oiily to t h r samr rxtriit as the frrqiieiivy; that is, t,hc geiicrator exvitatioii is maiiitaiiied at tlie same valiie as heforr. Motor torqiie aiid wrrriit varg irith rediiiiiig frequeiivy i i i t,he samp iray as t,hey do with iiirrrasiiig spwd. sitiw i i i either rase t,he rotor frequeiicy is redured. C:«iiseqiieiitly, at 10 pcr i c i i t frcqiiciicy, the torqiie delivered and the wrreiii d r a w i \vil1 iic approsimatcly thc same as at 90 prr r r n t speed. IItari: tlir torqiie is griierally highcr aiid the ciirreiit loi\-er thaii at standstill. h t ttirse lon- freqiiriiries the effertive liiir resistarire is grcatly iiirrrast:d so that RII this iortliic is iiot rcalizcd. Severtheless, t,he scheme will eITrctively iii(.rcase t,lic toripie a\-nilat>lefor startiiig aiid aweleratiiig tlie motor. H o w v e r , thcre are scvcral disadvaiitages which usually makr it impractical: i . T o ohtaiii miich improvemriit thí! frequciicy must be redured t o a \-cry Iow valiie, iisiially M o \ \ - 50 per w i i t frequeiii.y, ivhich is difficult for some typcs of geiirrator ilrives. 2. i i i i iiidcpciideiit drive for the exeiter must he proridcd as direetcoiiiiectcd (ir I>rltrd exriters uill iiot provide suficieiit excit,atioii at 1ow geiicrator speeds. 3 . Loivcriiig tho system freqiieiiry may adversely affwt other equipmeiit coiiiiected to tlie systrm. Coriscquciitly, t h r svhrme is usually applicahle only for a generator supplyiiig a siiiglr motor ivheir excitatioii is supplicd by ai1 excitcr driven by a sepaiate steam tiirhiiic or aii eqiially iritiepeiidciit excitation source. 111 siicli cases, t h e schemc may be quite advaiitagrous.

V O L T A G ~ S T A N D A R DRATINGS, VARIATIONS, CALCULATION OF DROPS EXAMPLE OF CALCULATION

OF

VOLTAGE DROP DUE T O M O T O R STARTING '

Data (see Fig. 4.48) Generators: Two identical turbine-driven generators, 3600 rpm Total output rating = 10,000 kva Voltage rating = 6900 volts Voltage-regulator setting = 6700 volts Overhead line: 3-ft equivalent delta spacing Length = 5000 f t Conductor size = KO, 4/0 Awg Transformer hank: Output rating = 2000 kva (three-phase) Transformer voltage rating = 6600-2400 volts Motor starter: Autotransformer type Tap = 65 per cent Motor: Synchronous motor Output rating = 1000 hp Full load input = 1000 kva, 0.8 power factor Voltage rating = 2200 volts Full-voltage starting kva = 500 per cent Full-voltage starting torque = 65 per cent Initial conditions: Initial voltages At generator bus

=

6700 volts (regulator setting)

I rrT"

FIG. 4.48 Circuit diagrcm of power supply to motor.

LINE TRANYORMER BANK

MOTOR

273


274

At transformer primary = 6700 volts At transformer secondary = 2440 volts Initial loads At generator bus = 5000 kva (50 per cent of generator rating) of constant-current type No initial load on overhead line or on transformer Requirements : Minimum allowable voltage a t generator bus = 90 per cent of initial voltage Motor starting torque must be at least 25 per cent Voltage calculations: Starting kva drawn with rated motor voltage a t autotransformer starter = full-voltage starting kva X multiplier from Table 4.13 = 5 X 1000 X 0.46 = 2300 kva Kva applied to transformer a t rated secondary voltage =

starting kva a t 2200 volts X

=

2300 X

2400

(2200)

=

rated secondary voltage 2200

2735 kva = 137 per cent of bank rating

Transformer secondary voltage (neglecting primary voltage drop) is obtained from Fig. 4.44. For banks rated 15 kv and below and a starting kva of 137 per cent, it is 93 per cent of the initial secondary voltage. Kva applied to transmission line a t initial voltage secondary voltage of transformer = starting kva a t initial voltage X initial secondary voltage =

2300 X

( )' . 4

X 0.93 = 2620 kva

~~,

Loading factor = kva applied a t initial voltage length in feet - = 2620 X 5000 - 0.292 (6700)* (initial volts)* ~

From Fig. 4.46, for M = 1, 4/0 line, 3-ft spacing, voltage drop is 11.5 per cent. Since M = 0.292, drop in line is 0.292 X 11.5 = 3.36 per cent. Voltage a t end of line (neglecting generator voltage drop) is 100 - 3.36 = 96.64 per cent of initial voltage (6700 volts). Kva applied to generator a t rated generator voltage voltage a t end of line = starting kva a t rated generator voltage X initial line voltage


( itransformer n i t i a l secondary secondary voltage voltage X 0.9664 X 0.93

=

=

)

=

2300 X

(

275

i

mo

6900 X 22002400)

2690 kva 26.9 per cent of generator rating

From Fig. 4.40, performance factor K for a 5000-kva 3600-rpm generator a t 50 per cent initial load is 1.9. From Fig. 4.39, minimum generator voltage, for 26.9 per cent starting kva and K = 1.9, is 92.5 per cent of the initial voltage (6700 volts), or 6200 volts. From Fig. 4.42, restored generator voltage for 26.9 per cent motorstarting kva and 50 per cent initial load is equal to the initial voltage or 6700 volts. The minimum voltage a t the motor starter is equal to the initial voltage a t the motor starter multiplied by secondary voltage of transformer Minimum generator volts Initial generator volts initial secondary voltage voltage a t end of transmission line initial voltage a t end of line = 2440 X 0.925 X 0.93 X 0.966 = 2030 volts

) (

(

The restored voltage a t the motor starter is equal t o the initial voltage at the motor starter multiplied hy secondary voltage of transformer Restored generator volts initial secondary voltage Initial generator volts voltage a t end of transmission line initial voltage a t end of line = 2440 X 1.00 X 0.93 X 0.961 = 2200 volts

) (

(

Since the restored voltage is equal to rated motor voltage, the starting torque on the 65 per cent autotransformer tap = 65 X (0.65)' = 27.5 per cent The minimum voltage a t the generator bus (92.5 per cent of initial value) and the motor starting torque (27.5 per cent) both meet the requirements. Next the formulas for calculating voltage drop will be used to solve this problem. I t will be assumed that Motor-starting power factor Transformer resistance Transformer reactance Transmission-line resistance

= = = = =

30 per cent 0.7 per cent 5 per cent 0.0573 ohm per 1000 f t 0.287 ohm total

276

VOLTAGE-STANDARD RATINGS, VARIATIONS, CALCULATION O F DROPS

Transmission-line reactance

=

=

0.121 ohm per 1000 f t 0.605 ohm total

The per-unit system will be used n.ith base kva equal to the motorstarting kva a t rated motor voltage (2300 kva) and base voltage equal to rated motor voltage (2200 volts). On this basis, the motor constants are

z, = 1 cos On/ R.u

B M = 72.5'

=

0.3

=

Z . w cos BM = 0.3

X M = Z M sin Bar

=

0.954

The resistance and reactance of the transformer vill equal the per cent values multiplied by secondary voltage rating of transformer rated motor voltage

Motor-st,arting kva a t rated motor voltage Kva rating of transformer

Transformer resistance = 0.7 X 0.0137 = 0.0096 Transformer reactance = 5 X 0.0137 = 0.0685 The resist,ance and reactance of the transmission line will equal the ohmic values multiplied by Motor-starting kva a t rated motor voltage rated motor Primary voltage rating of transformer volts x 1000 1000 Secondary voltage rating of transformer 2300 -

=

0.06275

(4.16)

1000 Line resistance

= =

0.287 X 0.06275 0.605 X 0.06275

= =

0.0180 0.0380

The total resistance and reactance between the motor starter and the generator will be Rs = 0.0096 0.0180 = 0.0276 X s = 0.0685 0.0380 = 0.1065

+

+

The equivalent motor-starting kva at rated generator voltage = starting kva drawn by motor if voltage at motor starter were maintained at the initial value


VOLTAGE.-STANDARD

X

d ( R , ,+ &)2 =

2300 X

z

277

(rated generator voltage initial generator voltage 1 X d ( 0 . 3 0.0276)* (0.954 0.1065)'

Y

+ (X,, + Xs)* ~~

(-)2440

+

2200

+

x (o)*

+

= 2700 kva

(4.19)

This is substantially the same as previously determined; so the generator voltage drop will h r essentially the same, that is, the minimum voltage will be 6200 volts a i d the restored voltage, 6700 volts. The voltage at the motor starter will equal the voltage at the geuerator multiplied by initial motor voltage zw x ( .initial . . generator voltage d(~,, + n,)z + ( x ,+~xS)2

zo0.328 + 0.276)' + (0.954 + 0.1065)* X '6700 1

-

~

d(0.3

=

Thus t,he minimum voltage at the motor starter will he 6200 X 0:328 and t,he restored voltage

i d

=

2030 volts

=

2200 volt,s

he

F700 X 0.328

(4.21)

Chafiter 5

by R. H. Kaufmann and Maynord N. Halberg

Sys tern Overvoltages-Causes and Protective Measures Electric insulation in energized systems is continuously under stress. To make the most economical use of insulation, operating overvoltages should he curbed in so far as is reasonably possible. The application of additional insulation to accept higher overvoltage levels entails several rather obvious disadvantages: (1) increased cost, ( 2 ) increased size and weight, (3) increased resistance to the flow of heat from the currentcarrying conductors. In the case of a-c systems, the electric potential is varying substantially as a sine wave. The crest potential will be 41 per cent greater than the rms value. Under ideal conditions the line-to-ground voltage stress mill he less than the line-to-line operating voltage. In the case of direct current or single-phase alternating current, this ideal line-to-ground voltage would he E L L / 2 ,or 50 per cent of the line-to-line value. In the case of three-phase a-c systems, this ideal line-to-ground voltage would be E L L / f i , or 58 per cent of the line-to-line value. Throughout this section, overvoltages will he expressed as multiples of the ideal balanced voltage stress in three-phase systems. Electric systems are subject to disturbances of many types which unavoidably produce overvoltages. However, the application engineer has at his command many system design principles which will greatly curb the magnitude of overvoltages. It is important to note that a-c systems are subject to many types of overvoltages not to be found in d-c systems; hence a-c systems deserve more careful consideration of the overvoltage problem. Electric insulation exhibits the effect of fatigue. Insulation will fail upon repeated or prolonged application of a given voltage stress which is 278

SYSTEM OVERVOLTAGES-CAUSES AND PROTECTIVE MEASURES

279

far below the single-impulse \vit,hstaiid abilit,y. One may conclude that a reduct,ion in either the magnitude or duration of overvoltage stress will in general result in longer useful life. OVERVOLTAGE SOURCES

There are many varied sources of overvoltages of sufficient magnitude to be damaging t o the insulation of a-c industrial power distributioii systems. 111t,liis chapter the mechanism by which the more prominent overvolt,ages are created v i l l be described and preventative measures suggested. ‘Treatment of t,he following will be included: I . Static 2 . Physical contact nith a higher voltage system 3. I1esouani.e effects ill series inductive-capacitive circuits 4. Repetitive int,ermittent short circuits 5 . SIT-itrhing surges (i. Forced-current zero-current interruptiou 7. Autotransformer connections 8. Lightiiitig Of these, most are the result of effevtsdirectly within the electric system itself. I n contrast, lightning (a vicious source of overvokage) is communicated to the electrical system from nature’s powerhouse in the heavens above. STATIC

Wind-blown sand or dust can become highly charged and impart relatively high voltage to exposed overhead electric conductors. Moving belts rutiiiing on iioiimet,allic pulleys can also develop high voltages by st,at,icmeans which may in turn be communicated t o electric system conductors if electric enclosing frames arc improperly grounded. The rate a t wtrirh electric i,harge is communicated t o electric system conductors by stat,ir means is extremely low. Even a rather high-resistance ground i~iiincrtionon the electric system n d l discharge these stat,ic currents t o ground as fast as they are rereived with negligible overvoltages. I n addition to grounding the elect,ric service system, it is important that electric machiue frames arid all metallic enclosures which contain electric circuit conductors be effectively grounded (see Chap. 7). PHYSICAL CONTACT WITH A HIGHER VOLTAGE SYSTEM

If the conductors of a high-voltage electrical circuit come in contact with those of a lower voltage circuit, then the same potential will exist on

280

SYSTEM OVERVOLTAGES-CAUSES A N D PROTECTIVE MEASIJRES

both circuits at the point of contact. If Lhe low-voltage circuit does not have its neutral grounded, its potential will be increased t o t h a t of the high-voltage system or flashover mil1 occur. If Lhe low-voltage system is anchored close t o ground potential as hy Lhe use of a solidly grounded neutral, high values of current may flow from the high-voltage system, b u t a much lower voltage will appear than with an isolated neutral system. Accidental cootacts hetmeen primary and secondary voltages on industrial systems are guarded against by the use of metal enelosures and metal barriers which separate conductor systems of different operating potentials. In some cases overhead circuits have both primary and secondary on the same pole, but substantial clearances reduce Lhe danger of accidental contact t o a minimum. Occasional cross-ups have occurred between primary and secondary on overhead circuits, and a few cases are known where failure has occurred between primary and secondary inside a transformcr. UNINTENTIONAL CONNECTION

7 PHYSICbL CONNECTIONS

P,

I I

'..

/

2 I 3LI 1

N O R M b L POSITION O F 4 8 0 V VOLTAGE TRIbNGLE

C xq I

I

Eb= 2 4 0 0 V

\:ol,

\,\

L,--' I

e0

b RESULTING VECTOR VOLTbGE DIbGRbM

FIG. 5.1 Overvollage on 480-volt ungrounded ryrtem rerulting from contcxt with a higher roltoge ryrtem.

Figure 5.1 illustrates this type of fault connection. It can be responsible for dangerous overvoltages on ungrounded low-voltage systems. The most effective protection against that type of overvoltage is grounding of Lhe lowvoltage system mith the grounding impedance made low enough t o accept Lhe maxirnum line-to-ground fault current of the high-voltage system without biasing the neutral of the low-voltage system by a dangerous amount.


281

RESONANT EFFECTS I N SERIES INDUCTIVE-CAPACITIVE CIRCUITS (LIMITED TO A-C SYSTEMS)

Ungrounded-neutral a-o systems are most commonly subject t o overvoltages originatiiig from this cause. It is import,aot t o recognize that ungrouiided-iieutral systems are actually capavitively roupled t o ground rather than truly divorred from ground. They are ungrounded in the sense that no int,er(.oiinection with ground has purposely been made, but every element of the electric system incorporates some capacit,aiice t o ground which constitutes an inherent caparitive impedance interconnection tietween the elertrir system conductors and ground. Every ungrounded elertric system contains the essential elements presented i n the upper diagram of Fig. 5.2. The electrical behavior of any one phase conductor relative t o ground rail be determined by a much simpler equivalent rirruit, as indicated in the lower sketrh of Fig. 5 . 2 . A

.

'S

A PHASE

I

GENERATOR OR TRANSFORMER

3- PHASE ESSENTIAL ELEMENTS

xs

A PHASE

"A"*"

Eg

-E'.c %

EQUIVALENT CIRCUIT REFERRED TO A PHASE CONDUCTOR

FIG. 5.2

Elemental composition of an ungrounded system

In terms of this simpler equivalent circuit it will be possible t o understand readily the effect of connecting different types of impedance hetween line and ground as shown in Fig. 5.3. I t becomes evident that the connection of any value of either resistance or raparitanrc tietween one line and ground produres no dairgerous overvoltages. The potential on the phase to which the impedance is connected progressively diminishes from normal value t o zero. The potential t o ground on the remaining two phase conductors will be increased t o full line-to-line value at the time the first

282


6

5

I

I

.

" 4

Y

J

z

LL

4 3 "Y 7

Ec ' 2

. i

FIG. 5.3 Overvoltages on ungrounded systems result

I

0

0

2 3 4 R4TlO OF ZF TO X C O , ~

S

6

from a high-inductive-reactonce connection between line and ground.

phase conductor has been reduced to zero potential. This represents an overvoltage of only 73 per cent, which is not dangerously high and will normally produce no ill effect unless continued for a long time. The connection of an inductive reactance between line and ground, on the other band, can be responsible for the production of serious overvoltages to ground. It is the ratio of the inductive reactance of the lineto-ground circuit to the total capacitive reactance of the system to ground which controls the degree of overvoltage. The highest overvoltage will occur when these two reactances are equal, and a t this point they may be as much as ten times normal. It is significant to note, however, that over a two-to-one range of reactance, overvoltages of three times normal or more would be produced. The unintentional connection of an inductive reactance between a phase conductor and ground can occur in a number of ways, some of which are illustrated in Fig. 5.4. The operating magnetic coil of a motorstarter contactor may be inadvertently connected between phase and ground by a ground short circuit in the control wire to the push-hutton station or the slip of a maintenance man's screwdriver. Any time that the inductive-reactance value, which becomes connected from phase to ground, falls in the danger region indicated on Fig. 5.3, dangerous overvoltages to ground will be produced which are communirated over the entire metallic conductor system of that operating voltage. Overvoltages originating from this canse can be completely suppressed by a relatively light-resistance ground on the electric system neutral. A grounding resistor of about the same ohmic value as the total charging

C

B

"--

C4SE I

AN INDUCTIVE WINDING CONNECTED BETWEEN 01 GROUND

FIG. 5.4

-

f

I

CbSE 3

Examples of unintentional high-reactance connections between line and ground.

ONE BROKEN OVERHEbO LINEGROUNDED ON T H E L O A D SIDE OF T H E B R E A K CONNECTS T H E REACTANCE OF TRANSFORMERS 12 AND 13 I N PARbLLEL B E T W E E N L I N E b N D GROUND [NOTE II

BROKEN L I N E GROUNDED

b GROUNO FAULT bT A F U S E PROTECTED T R I N S F O R H E R C I N BLOW ONE FUSE LEbVING THE REACTbNCE OF TRANSFORMERS 12 4 N D T 3 I N P b R A L L E L B E T W E E N L I N E AND GROUND 1 NOTE I I

CbSE 2

_

Y U N G R O U N D E D T R I N S F O R M E R CONNECTIONS WOULD PRODUCE T H E SAME EFFECT

NOTE I

:CIDENTbLLY PHASE h N D

II

284


capacitive reactance t o ground is sufficient t o eliminate overvoltages almost completely. It will be evident t,hat there is good reason t o adopt electric system neutral grounding with a much lower value of grounding resistance for ot,hcr reasons (see Chap. 6). Figure 5.3 has been computed on the basis t,hat the inductive reactance is linear. If this reartance incorporates a n iron core which during the mode of operation being considered should encounter magnetic saturation, the performance will be somewhat different. Under such conditions the effective reactance of t,he inductive circuit can become much lower than the unsaturated reactance, and the voltage will tend t o oscillate automatically betveen vokage limits which cause the effective inductive reactance to match the capacitive-reactance value. This character of operation has been named ferroresonance. The maximum voltage so developed may not be so high as would be produced by a linear reactor but, may still be in excess of two or three times normal. Substantial overvoltages may result by ferroresonance when the unsaturated reactance is many t,imes the capacitive reactance to ground. The application of grounded-Y potential transformers on ungrounded systems with a Y or broken-delta secondary connection can be responsible for damaging overvoltages as a result of resonant or ferroresonant action since the magnetizing reartance of the pot,ential transformers becomes connected from phase conGROUNDED WYE- BROKEN DELTA POTENTIAL TRANSFORMEIS FOR GROUND ductors t o ground. A comINDICATOR OR ZERO SEQUENCE VOLTAGE plete descriptionof thisphenomenon need not be taken UNGROUNDED NEUTRAL SYSTEM u p here as it has been adequately treated in an AIEE technical paper (see reference3). Thesesystemvoltage oscillations will not occur if the electric system U neutral is grounded. Freedom from this particular type of voltage oscillation TO INSTRUMENTS can be obtained even with R ungrounded-neutral operation byusingpotential transTO INSURE FREEDOM FROM UNWANTED LINE-TOformers with a line-to-line GROUND VOLTAGE OSCILLATIONS :

r Tz. T3, W I T H THE L I N E - T O - L I N E RATED VOLTAGE 2 APPLY A SECONDARY LOADING RESISTOR WITH A RESISTANCE NOT GREATER THAN 4 0 PERCENT OF THE TRANSFORMER MAGNETIZING REACTANCE. NOTE- THE LOADING RESISTANCE CAN BE APPLIED TO EACH SECONDARY BUT WILL THEN CDNSUME POWER AND LIBERATE HEAT CONTlNUOUSLY I SELECT P T s TI.

FIG. 5.5 Grounded-Y brokendelta potential transformers for ground indicator or zero-sequence voltage detector.

SYSTEM OVERVOLTAGES-CAUSES

AND PROTECTIVE MEASURES

285

voltage ratiiig and the applirat,iori of shuiiting resistors on the secondary windings as is outliried iii Fig. 5.5. Series-capacitor melders are occasionally applied, particularly in the case of large-sim machiiies because of their ability t o reduce the kva demaiid aiid improre the operating poiver fartor t o substantially unity. However, the series-raparitor welder preseiits a definite voltage hazard to aii uiigrouiided-iieutral a-c supply system. Duriiig welder operation the voltage arross hoth t,he series raparitor and the weldiiig transformer primary v i l 1 he severa1 t,imes the rated line-to-liiie voltage. The physical electrir roiiiieitioiis aiid the associated vector voltage relationships are iiidicated iii Fig. 5.íi. 48OV 3-PH 6 0 C Y

o

s i i o m m ~ uTO~GROUND -

?--.

PHVSlCbL CONNECTDNS

NORMAL POSITION OF IP UOLTME TRIANGLE;

-,,, /' b

RESULTING VECTOR VOLTbGE OIAGRPH

FIG. 5.6

Overvoltager on ungrounded syrtemr os a rerult ot o ground contact on a ieriei capocitor welding mochine.

Should a fault t o grouiid occur at the juiiction hetveeii the series capacitor aiid the weldiiig transformer (poiiit, P ) , the lorat,ioii of ground poteiitial will teiid t o become t,hat of this juiictioii poiiit iiistead of the center of the a-r system voltage triangle. The t,otal system eapacitiye impedaiice t,o grouiid would geiierally be expeeted to be high, relatire to that of the welder series rapacitor, aiid thus offers practically no opposition t o this shift in the loration of ground potential. Iii the case illustrated iii Fig. 5.G, it will be evideiit that the poteiitial of the A-phase roiiductor may be elerated to ahout 2000 volts to grouiid, which is about seveii times iiormal. As iii the other cases, this overvoltage is commuiiicated to a11 equipmeiit metallically iiiterroniierted a t this commoii operatiiig voltage. AI1 these resonant inductive-caparitive overvoltage hazards can be elimiiiated by electric system neutra1 groundiiig.

186


INTERMITTENT GROUND FAULTS

Substantial overvoltages can he developed in ungrounded a-c industrial systems by sputtering or intermittent ground-faulting connertions. The intermittent character of the short-circuit path may be the result of vibration which causes an electrical conductor to make contact intermittently with ground, the result of scattering particles of molten conductor metal which intermittently establishes a conducting path to ground, or as a result of successive breakdown and seal off of the separating space between conductor and ground. In the last case involving a fixed separation between conductor and ground, a progressively increasing breakdown voltage across this gap is an essential element in the build-up of severe overvoltages. Intermittent ground-fault conditions on lom-voltage ungroundedneutral systems have been observed to create overvoltages of five or six times normal quite commonly. An unusual case involved a 480-volt ungrounded system. Line-to-ground potentials in excess of 1200 volts were measured on a test voltmeter. The source of trouble mas finally traced to an intermittent ground fault in a motor-starting autotransformer. About two hours elapsed while the source was being located, during 13 hich time between 40 and 50 motors broke down. Electric systems which are grounded through reactanre of too high an ohmic value ( X a more than ten times XI) are also subject to overvoltage by this same mechanism acting in a little different form. An understanding of the manner in which a discontinuous electric connection can he responsible for the generation of overvoltages can he most easily acquired by examining the case of a sputtering or intermittent line-to-ground fault on an ungrounded-neutral system. I n Fig. 5.7 at A is shown the vector voltage pattern of a three-phase a-c system as it would operate under normal balanced conditions. The voltage vectors E., Eb, and E, rotate about the neutral at synchronous speed. The electric neutral is a point of central symmetry and remains constant at ground potential if the individual phase voltages are pure fundamental-frequency sine waves. Should the A-phase conductor become grounded, the system voltage triangle mould become displaced as illustrated in B . At the phase position illustrated in B , the A-phase voltage is at its maximum value at which instant the charging current to ground (90' ahead of the voltage) is passing through zero. In case the short circuit contains a small gap or an arc, the arc current would become extinguished at this point. Note that the trapped charge on the line-to-ground capacitance will tend to maintain the voltage triangle in the same displaced position. I n other words, the potential of the neutral (relative to ground) would tend to


287

remain at a d-c potential equal to the crest value of the a-c voltage wave. All this merely says that there will be little tendency for any voltage to reappear across the gap in the short circuit immediately following the current zero which occurs at B . During the next half cycle, however, the a-c generat,ed voltages will reverse their polarities (vectors rotate 180°),which would cause the threephase vector voltage pattern to assume the position shown in the upper part of C . Kote that during this one-half cycle time interval, the potential of the A phase has progressively inrreased from zero value to about twice the normal line-to-neutral crest voltage relative to ground potential. This value of line-to-ground potential of the A phase may he sufficient to break down the gap in the ground-fault circuit arid reestablish the corinection between the A phase and ground. If so, the A-phase potential will tend to be suddenly yanked to ground potential. Iuevitably there will be some system reactance in the A-phase conductor to the ground shortcircuit point which would result in an oscillation of the A-phase-conductor potential between plus 2 and minus 2 at a frequency probably 20 to 100 times normal. If the short circuit consisted of a solid metallic connection, this oscillation would decay to zero, leaving the A-phase conductor at ground potential. Xote that associated with this high-frequency transiEi"'

E;

Y--

+ $

CYCLE

--tC

$

CYCLE

4

€0

NORMAL

A

FIG. 5.7

B

C

D

Overvoltages on ungrounded systems due to repetitive momentary contact between one line and ground.

288

SYSTEM 0VERVOLTAGES.-CAUSES


tory oscillation will he a corresponding trarisitory charging current t o ground. This transitory charging current t o ground, or restrike current, will again reach zero value when the system voltage swing is at the maximum excursion i n the negative direction, as showri in the lower part of C . Thus, an opportunity is afforded for the gap in the ground short circuit t o rerlear. If reclearing does occur, a charge is again trapped on the system rapacitance t o ground which would tend t o maintain a constant d-c potential t o ground on the system neutral. In the course of the next following half cycle, the voltage vector system will again rotate 180°, causing the potential of the A-phase conductor to ground t o he elevated from minus 2 to minus 4 as indicated by the transition from the lower part of C to the lower part of D. This increased voltage across the short-circuit gap may again result in restrike, in which case the voltage triangle would tend t o be thrown in the positive direction in the form of a high-frequency oscillation between poteutial limits of minus 4 and plus 4, which in the presence of a solid metallic connection would gradually decay t o zero. I n this explanation of the mechanism, it will be noted that all conditions have been most favorable to the creation of the highest possible restrike voltages in the shortest possible time. The restrike has been assumed t o occur at the time the maximum recovery voltage was reached but not before. Likewise it has been assumed that a reclear occurs at the first current zero after restrike. Under these conditions a line-to-ground potential of five times normal has been developed in less than two cycles. I n practical cases, t,he restrike may occur before the maximum recovery voltage has been reached, and several cycles of the transitory oscillation may take place before the short cirruit reclears. While in theory it might he possible progressively t o increase the line-to-ground voltage by successive restrike without limit if the dielectric strength progressively increases, voltage measurements on actual systems indicate that voltage levels of five t o six times normal are rarely exceeded. There is reason t o believe that damaging overvoltages of repetitive restrike origin are far more common on ungrounded-neutral systems than mould a t first he suspected. The case which was mentioned in an earlier paragraph is unusual in t,hat the obnoxious restriking conditions persisted for a long interval of time while t.he source was being located. A farmore common occurrenre is one in which several pieces of electric equipment on the system suffer electrical breakdown apparently simultaneously and one or more of the fault conditions were known or believed to involve ground. These multiple failures are commonly associated with ungrounded-neutral system operation. It is also known that a solid metallic ground connectioo on one phase may exist for subshntial intervals of time without producing multiple breakdowns in equipment,

SYSTEM GV'ERVOLTAGES-CAUSES


289

although it does produce 73 per cent overvoltage on two of the phase conductors. It therefore seems reasonable to assume that the multiple failures result from the appearance of overvoltages considerably in excess of 173 per cent normal. Distribution-system ox,ervoltages of repetitive-restrike or intermittentground origin can be entirely eliminated by effective system neutral grounding (see Chap. 6). Resistance grounding with a resistance ground fault of any value upward of the line-to-ground charging current mill be effective. For various other reasons it mill he evident that higher values of available ground-fault current will he desirable. If reactance grounding is contemplated (it rarely finds application in industrial systems), it is important to keep the reactance of the grounding circuit sufficiently low so that the ratio of X o is no more than ten times X , . If this grounding reactanre value is exceeded, opportunity is given for another type of repetitive restrike action which can result in overvoltages t o ground. The ground-fault neutralizer (Petersen coil) represents one special case of high-reactance grounding which is free of overvoltages by repetitive restrike action. This is due t o the fact that the reactance value is carefully selected so that the oscillating circuit formed hetmeen it and the system-to-ground capacitance will oscillate a t normal line frequency. Following a ground-fault cnrrent shutoff point as at B in Fig. 5.7, the potential of the electric system neutral with respect t o ground would oscillate between plus and minus 1 at fundamental frequency as controlled by the tuned grounding reactor and system capacitance t o ground. Thus as the potential of the n-phase conductor with respect t o the neutral due to the generat,ed voltage in the supply system alternates from minus 1 t o plus 1, the free oscillation of the zero-sequence circuit remains in step with it, with the net result that the potential of the A-phase conductor tends t o remain at ground potential. Voltage of normal frequency gradually reappears as the free oscillation in the zero-sequence circuit decays. I n general, some 15 or 20 cycles will elapse before the potential of the previously shorted phase increases t o three-quarters of normal value. Thus, the freedom from restrike is due t o the long-delayed reappearance of voltage across the line-to-ground circuit. SWITCHING SURGES

Circuit switching operations constitute abrupt changes in circuit parameters and can be responsible for the creation of overvoltages although generally of short duration and not in excess of two to three times normal. It will be important t o recognize that normal a-c switching interrupters offer very little opposition t o the flow of circuit current during the course of current flow but do act t o build up dielectric strength rapidly during a

290

SYSTEM OVERVOLTAGES- CAUSES AND PROTECTIVE MEASURES

normal current zero and prevent reestablishing current flow during the following half cycle. As a result of this action it is unnecessary that the stored magnetic energy in the inductance of the circuit be disposed of during interruption. Interruption takes place at a normal current zero, at which time the stored magnetic energy is zero. A quaiitative understanding of the mechanism whereby such overvoltages are generated will be useful. Of first consideration is the amount of voltage change which would tend to appear across the switching contacts if they were switched open. For example, in Fig. 5.8, a line-to-line short-circuit condition between phases A and B is illustrated. With the circuit breaker still closed, the potential of a' and b' must be common and will lie midway between potentials e, and ea, as indicated in the vector diagram. With the vector relationships shown in the figure, the current in the faulted circuit will be going through zero, which affords an opportunity for the circuit breaker to make an interruption if the contacts have parted. If current flow is interrupted at this current zero, the potential of a' tends to return to e. while that of point b' tends to return to eb. OVERVOLTAGE IN CLEARING A LINE- TO- LINE CIRCUIT FAULT

SHORT CIRCUIT

;-,

- ___-

-\,

, , , "

'

"

VOLTAGE RELATIONSHIP WITH SHORTCIRCUITON AT THE T I M E OF A CURRENT ZERO IN THE SHORT CIRCUIT CIRCUIT e.' = eb' (VOLTAGE mob AT MAX VALUE1

-

IF CURRENT INTERRUPTION OCCURS AT THIS CURRENT ZER? THE POTENTIAL OF POINTS 0 AND b WILL TEND m SNAP BACK TO ea AND Ob RESPECTIVELY BUT DUE TO PRESENCE OF L AND C I T WlLL TAYE THE FORM OF h TRINSITORI OSCILLATION W l C H WlLL OVERSHOOT END

POINT eb'' 113 PERCENT OR

MAX e.'DR

73 PERCENT OVERMLTAGE

FIG. 5.8

Overvoltages due to interruption of

(I

line-to-line short circuit at current zero.

SYSTEM O V E R V O L T A G E S - C A U S E S A N D PROTECTIVE MEASURES

291

There will inevitably be inductive capacitive constants which cause this return to take the form of an oscillation of relatively high frequency; this causes the potential of points a' and b' to overshoot their final value by about a n equal amount. In this illustrative example, the potential of point b' would transitorily swing t o a value of 1.73 times normal crest voltage in the positive direction while that of the point a' would make a corresponding swing to 1.73 times normal crest value in the negative direction. Circuit breakers which introduce substantial resistance drop during current flow tend to reduce the magnitude of switching transient voltages. As a result of the higher power factor of the short circuit, the point at which a current zero is reached will approach more rlosely to the point at which a voltage zero would also he reached, which thus lessens the magnitude of voltage that tends to appear across the contacts immediately following current zero. Another form of switching transient which develops overvoltage primarily on the utilization machine on contact closing is illustrated in Fig. 5.9. Here illustrated is a n open-cycle autoPOSSIBLE SWITCHING OVERVOLTAGES ON CLOSING L I N E BREAKER WITH OPEN CYCLE AUTOTRANSFORMER START

ASSUME0 VOLTAGE RELATlONSHlP JUST PRIOR TO CLOSING L I N E BREAKER IAUTOTRANSFORMER STARTI CBPOLENOI ISTHE FIRST TO CLOSE

4e, c,

MOTOR TERMINAL B W I L L TEND TO ABRUPTLY JUMP TO e. BUT OUE TO

MAX TRANSLTORI VOLTAGES-MOTOR TERMINALS TO GROUND TERMINAL B - 2 5 0 PERCENT 1150 PERCENT OVERYOLTAGEI TERMINALS A 8 C - 325 PERCENT'1225 PERCENT OVERVOLTAGE1

FIG. 5.9 Possible switching overvoltage when motor running breaker closes lopen-cycle autotransformer start).

292


transformer starting arrangement. It has been assumed that 65 per cent voltage has been applied on the starting connection and the machine rotor brought up to near synchronous speed. The motor was then disconnected from the starting tap preparatory to reconnection across full-line voltage. During this interval it is possible that the internal generated voltage within the motor has dropped to 50 per cent of rating and has slipped back in angle so as to he 180' out of phase with respect to the supply system. At this point, the potential difference appearing across each of the three line-switching contacts is one and a half times normal line-to-neutral voltage as indicated by the vector relationships. Suppose that the line-switching unit is now closed and that pole 1 is the first to make contact. The potential of motor terminal R would tend to abruptly assume the potential e,, but the inevitable transitory overshoot would carry it on up to 250 per cent normal with respect to ground. The potential of the motor terminals A and C would tend to be carried along and suffer a transitory excursion to 325 per cent voltage with respect to ground unless contacts 2 and 3 close at almost the same instant as contact 1. Closed-cycle starting arrangements such as reactor starting or Korndoerfer autotransformer starting minimize the overvoltages which may be developed in this manner. One of the most severe sources of switching overvoltages is associated with the separation of two system sections which have become unsynchronized and are switched apart when the generated voltages in the two sections are nearly 180' out of phase. The elements of this case are illustrated in Fig. 5.10, which shows a synchronous motor t h a t has pulled out of step and the internal generated voltage of which is 180' out of phase with respect to the system. The main supply system on the left is considered to be operating with grounded neutral and contains a much smaller reactance than the motor circuit shown on the right. All three poles of the switching interrupter have been maintained in a closed position up to the time indicated by the vector diagram. It has been assumed that, in the course of pull-out operation, the demagnetizing reactive current which has been flaming in the motor stator windings has caused the internal generated voltage ahead of transient reactance in the motor to he depressed to 50 per cent of normal value. With the vector system in the position shown, the current in the A phase is going through zero, which affords an opportunity for interruption if the contacts have parted. If the current in the A-phase conductor is interrupted at this point, the potential of the motor A-phase terminal (point a2) will tend to jump to the right to its new steady-state position E,. The inevitable transitory overshoot will cause its potential to swing about an equal distance the other side of point E,, as shown hy the dotted line. At the

SYSTEM OVERVOLTAGES-CAUSES A N D PROTECTIVE MEASURES

293

maximum of this transitory excursion, the potential of point a2 mould reach about 3% times normal crest t o ground in the positive direction. I n coutrast t o the examples just cited, the more usual switching operation which is involved in separating a normally operating rotating machine or composite system of rotating machines involves very little switching surge voltage. The systems on both sides of the switching interrupters contain internal sourres of generated voltage which are of almost the same magrrihde and very close t o the same phase position. Very little change in potential tends to occur on either side of the switching device at the time interruption takes place. Arc-furnare circuits can be sources of rather severe overvoltage if switched off while an arc is in progress within the furnace. As the priS

e

.

d

.

5

\i, c2

e,.... i

.. . .... .. A L L C B POLES S T I L L CLOSE0 ASSUME C B TRIPPED AND POLE I (PHASE A 1 IS THE FIRST TO INTERRUPT AT T H I S CURRENT ZERO 0 1W I L L T E N 0 TO JUMP TO e0,AND

Q p T O T H E NEW EA WITH .TRANSITORY EXCURSIONS SHOWN BY DOTTED LINES

FIG, 5.10

Possible overvoltager when interrupting o synchronous motor during out-of-

step conditions.

294



mary circuit-breaker contacts part, current at the breaker contacts can be forced to zero while current still continues to flow in the furnace arc. Thus the circuit breaker accomplishes a n interruption of line current with current flow still Continuing in the secoudary circuit. As the furnace internal current diminishes, the potential across the furnace arc increases in accordance with the normal inverse volt-ampere characteristic of an arc. The arc voltage progressively increases as the current dimiuishes and can result in a substantial voltage drop as the arc snaps out. While this voltage may not be high as referred to the arc-furnace anode, it still may be many times normal operating voltage and will be reflected to the primary side of the transformer by the turn ratio. The voltage developed at the transformer high-tension terminals may be dangerously high and sufficient to produce flashover. Special consideration is given to arcfurnace transformers, and preventative measures take the form of shuntcapacitor applications at the transformer terminals on older uuits or internal Thyrite* shunting resistors across sections of the winding 011 iiew units.

FORCED-CURRENT-ZERO INTERRUPTION

The discussion of switching overvoltages so far has considered interruption only a t a normal current zero. The term forced current zero or interruption of of current zero is used to describe an interrupting mechanism (be it a fuse, switch, section of small wire conductor, etc.) that has the property of developing a large countervoltage in opposition to rurrent flow which can force current to zero value at a time quite different from the normal inherent current zero of the rircuit. Should any element in an electric circuit have the ability to develop a high potential drop during current flow, the potent,ial so developed would appear on connected circuit conductors. The overvoltages so developed would persist until the stored energy in the inductive elements of the circuits has been dissipated (a current zero has been forced). A high rurrent short circuit created through a length of small wire conductors can be responsible for developing dangerous overvoltages in this manner. As current builds up in such a circuit, stored magnetic energy is heing accumulated in all inductive elements of the circuit. When the fusing point of the conductor is reached, the conductor copper tends t o separate into a loiig string of tiny globules of molten copper with a small arc between adjacent globules. The total voltage drop across the entire section of conductor may be several times the normal operating voltage of the circuit. During this interval of overvoltage, the magnitude of current is being diminished;

* Registwed

tradr-mark of Grncral Elrrtrir Cornpang-.


295

however the overvoltage will persist until the magnitude of current has been returned t o zero value. Because of the overvoltage problems, the vacuum contact switch finds little applicatioii. The vacuum switch tends t o shut off current completely the instaiit that the contacts part. Unless suitable overvoltage suppressors are associated with such an interrupter, high voltages will be developed if applied in inductive circuits. The overvoltages so produced may he sufficient t o sparkover the outside of the vacuum switch unless some other portion of the circuit breaks dowu a t a lower voltage. Current-limiting fuses constitute an example of a forced current interrupter. They possess the property of being able t o reduce the rurrent t o zero value ahead of a normal current zero. Overvoltages are developed during the operation of such an interrupter. As supplied by reputable manufacturers, the design of the internal elements contains special features mhirh rontrol the magnitude of such overvoltages, and full-srale tests are applied to prove the resulting performance t o ensure that overvoltages so developed d l be within the safe withstand value of the electric insulation of the voltage class t o which it is t o be applied. Because of the overvoltage problem, current-limiting fuse interrupters of a particular voltage rating should not be applied t.o electric systems of lower operating voltage. I n other words, a 7500-volt rated currentlimiting fuse should not he applied on a 2400-volt operating system because overvoltages developed iu its operation will be dangerous t o a 2400-volt insulation level. AUTOTRANSFORMER CONNECTIONS

Autotransformers for interconnecting two electric systems of different insulation level should be avoided in industrial systems unless both are solidly neutral grounded. The common metallic interconnection between t,he two systems which is formed by the autotransformer windings tends t o subject the lower voltage system t,o nearly the same transitory voltages as would be expected on the higher voltage system. There are some exceptions, and a specific example mill serve t o illustrate the nature. Should a system be planned which is to operate initially a t 2400 volts and later be converted t o 4160 volts with all equipment therein contaiuing insulation levels commensurate with 4100-volt operating potential, i t would be sitisfactory t o employ a suitable autotransformer for interconnecting this 2400-volt system with another 4160-volt syst,em. An unusual var'ation of autotransformer action which has been responsible for system overvoltages in a number of instances is represenled by a transformer with extended windings operating on an ungrounded-neutral system such as illustrated in Fig. 5.11. Applications of this sort are most

296

SYSTEM OVERVOLTAGES-CAUSES A N 0 PROTECTIVE MEASURES

often found in test areas or developmental areas which contain multipurpose transformers with a multiplicity of taps to permit a wide variety of output voltages to be obtained. If operated with system line voltage impressed across a fraction of the total winding, the vector voltage at the end of the winding extension will be as illustrated in Fig. 5.11 because the volts per turn developed in the winding extension will be exactly the same as the volts per turn in the excited winding. Should the end of the winding extension be inadvertently connected t o ground or develop a short circuit to ground, the point of ground potential would tend to move away from the center of the voltage triangle to the potential of the extreme end of the winding extension resisted only by the high system-to-ground capacitance coupling. It will be evident that, as a result of this action, the presence of any extended winding would cause the potential of one phase conductor to be elevated to more than 173 per cent of normal operating potential. The degree of overvoltage may be much more severe if greater amounts of winding extension are present. It is important to realize that these overvoltages would be carried to all apparatus connected to the same metallic system. Thus, a ground short circuit on a winding extension of a transformer in a small test area at one corner of a building might impose overvoltages on all equipment fed from the same load-center substation which might include half the productive machinery in that building. As has been true so many times before, grounding of the electric supply system neutral will cure this type of potential overvoltage also. A system grounding equipment which makes available a ground-fault current which is equal t o or greater than the short-circuit current resulting from short circuit of the extended winding portion of the offending transformer will keep the system line-to-ground potentials within safe bounds. It is quite generally true that transformers of this a

480V W Q H 60 CY

I

"

I

PHYSICAL CONNECTIONS

\ \ \

RESULTING (IOLTAGE

vEcmR

I

i

I

'"' b DIAGRAM

FIG. 1 1 1

Overvoltage on ungrounded systems due to a ground connection on the winding former.

of an autotrans-


1P7

character to be found in test areas are of relatively small physical size and do not impose restrictive requirements on the necessary system grounding equipment. As a matter of fact, on all low-voltage-system equipment (GOO volts and less) it is the standard practice to ground the neutral solidly. The application of three-phase transformers or three-phase banks of single-phase transformers, mhich do not incorporate a closed-delta winding in their make-up, should in general be avoided or quite rarefully examined to ensure that the resulting operation will be free of damaging overvoltages. This would be equally true of Y-connected autotransformers (see reference 4). Berause of the nonlinear shape of transformer magnetizing curves, the required transformer magnetizing current to produce a fundamental frequency sine wave of voltage will contain rather prominent amounts of harmonic currents. In a Y-connerted transformer system energized from a three-phase supply in the absence of a deltaconnected winding, the transformers are unable to obtain a sourre of third-harmonic current or multiples thereof because these are of zero sequence. As the result of the inability to obtain a third-harmonic exciting current, there will appear a third-harmonic voltage whirh may be as much as 50 per cent of the normal operating potential. Should the neutral of such a transformer system become grounded intentionally or accidentally and the supply system be ungrounded or high-resistance grounded, this third-harmonic voltage will be imparted to and appear on the system phase conductors and represent a sustained source of overvoltage. Even though the transformer system neutral is ungrounded, some fraction of the third-harmonic voltage will appear on the phase conductors, depending on the ratio of capacitance to ground within the transformer structure to the distributed capacitance to ground of the rest of the system. Core-type three-phase transformers present a fairly low zero-sequence magnetizing reactance which would hold the zero-sequence voltage to much lower levels than shell-type three-phase transformers or banks of three single-phase transformers and are thus much less susceptible to overvoltage difficulties. If operated with grounded neutral on an ungrounded-neutral system, a careful check should be made to ensure freedom from neutral instability, as treated in reference 3. While grounding the electric system neutral may not solve all the troubles of the Y-Y transformer connections, it will eliminate appearance of overvoltage on the phase conductors of a system to which such a bank of transformers might be connected. Overvoltage Example. A great many specific cases of system overvoltages have been analyzed, identified, and catalogued. All types are well represented. Space will not allow a lengthy treatment of these

298

A


DISTRIBUTION BUS (UNGROUNDED SYSTEM)

1

B C FUSE CUTOUTS

-

I

Q Q

PT2

PHYSICAL CIRCUIT CONNECTIONS

(A1

POWER SYSTEM EOUIVALENT CIRCUIT

PROTECT1V E EOUIPMENT CIRCUIT

A-PHASE FUSE OPEN

EOUIVALENT CIRCUIT FORMED BY OPENINGOF T H E A- PHASE FUSE ( FJI

FIG. 5.1 2

Circuit conditions responsible for an orenoltoge experience on an ungrounded power system.


299

specific overvoltage cases. However, it will be interesting to review one case. The one here described has been selected because it discloses how obscure may be the basic overvoltage cause. Note that the series resonant circuit created by the opening of one fuse might very easily fail to be identified, leaving the overvoltage source to remain a mystery. A metal-products plant in the North Central section of the country had made application of a set of rotating-machine protective capacitors and arresters a t the main bus of a medium-voltage distribution system through a set of fuse cutouts. To monitor the fuses, two potential transformers and voltmeters had been applied on the load side of the fuses, as illustrat,ed in Fig. 5.12A. As a result of opening of the fuse unit in the A phase it was observed that voltmeter V , went off scale, potential transformer 1 overheated and melted out the compound, the gap shunting resistor 011 the A-phase arrester was destroyed, and phase-to-ground overvoltages appeared on the phase conductors of the service system. Not until the resulting circuit is redrawn as in Fig. 5.12B is it apparent that the overvoltages result from series resonance (probably of ferroresonance character). System-neutral grounding is to be adopted to ensure freedom from overvoltages on the distribution system conductors. (Additional corrective measures are needed to ensure freedom from overvoltage trouble in the local protective equipment circuit-potential transformer and capacitor shunting arrester.) PROTECTION OF POWER SYSTEMS AGAINST THE OVERVOLTAGES CAUSED BY LIGHTNING The highest overvoltages to which industrial power systems are subjected are those caused by lightning. Limiting these overvoltages by suitable protective measures is essential if costly equipment failures and service interruptions are to be avoided. NATURE

OF THE OVERVOLTAGES

A lightning stroke to earth represents the sparkover of a highly charged condenser, a cloud forming one plate, the earth the other, and the air between the dielectric. The initial charge has been estimated to be as high as 1 billion volts, and stroke currents as high as 200,000 amp have been measured. Although lightning may strike directly a t the terminals of outdoor electrical equipment, this can generally be avoided by proper shielding. Thus, the overvoltages usually reach the equipment (both indoor and

300


outdoor) through exposed overhead lines which often bring power t o the plant or, in some cases, distribute power withiu the plant. Direct Strokes and Induced Surges. Lightning may produce an overvoltage on a transmission line either by a direct stroke to the line or by electrostatic induction from a stroke t o earth iri the vicinity of the line. The probable maximum voltage appearing ori a liiie by a direct stroke is 15 million volts and for an induced surge, 500,000 volts. These voltages appear between conductor and ground. Wave Shopes. Although the voltage surges produced hy lightning have high magnitudes, their duration is very short. I t is measured in microseconds (millionths of a second). Typically, the voltage rises very rapidly (in 1 t o 10 psec) t o the maximum or “crest,” value and theu decays more slowly, reaching 50 per cent of the crest value in 20 t o 150 psec. As illustrated in Fig. 5.13, the shape of a voltage or current, surge produced hy lightning (and those produced artificially for test purposes) is customarily expressed by two numherç. The first, is the time from the “virtual zero” of t,he wave front t o the time the wave reaches crest value, while the second numher is the time from the virtual zero t o the time the voltage or current has decreased t o 50 per cent of the crest value. The

-WAVE-FRONT

-

t

WAVE - TA1 L -CREST

VALUE

f

. I I

ZERO TIME O F CURRENT WAVE ZERO TIME O F VOLTAGE WAVE

I

b

&tut

3

I I

4

T i a N MICROSEMXIDS

-

I

1.

_ I

WAVE- SHAPE OF VOLTAGE WAVE ti X 12 CURRENT WAVE t 3 X t e

FIG. 5.13

Termr ured to dercribe voltage cind current waves.


301

virtual zero of a wave front is the intersection with the zero axis of a straight line drawn through the points on the front of the wave which are 30 per cent and 90 per cent of the crest value for a voltage wave and 10 per cent and 90 per cent of crest value for a current wave. Both times are usually expressed in microseconds. To illustrate, a 95-kv lf.5 X 40-psec wave is one that has a crest value of 95 kv, rises to crest value in 134 pser from the time of virtual zero, and decays to 50 per cent of crest value (47.5 kv) in 40 psec from the time of virtual zero. Traveling Waves. The voltage surge produced on a transmission line by lightning does not appear simultaneously at all points on the line; instead, it appears at successively later intervals of time as the distance from the point of the st,roke increases. Furthermore, the magnitude and shape (voltage vs. time) of the surge remain approximately the same at all points of a uniform line, but are simply displaced in time phase. In effect then t,he surge which appeared as a voltage-time wave on the line where the stroke occurred becomes two identical voltage-distance waves on the line which travel at uniform velocity in oppvsite directions from the point of origin. Keglecting all resistances, it can be shown that 1. The voltage waves travel along the conductor without change in magnitude or shape with a velocity equal to l / d T C fps, where L is the inductance in henrys per foot of line and C i s the capacitance in farads per foot of line. 2. A current wave accompanies the voltage wave and is of exactly the same shape, that is, a t any instant at any point on the line, the current flowing in the conductor is directly proportional to the voltage from conductor to ground. 3. The ronstant of proportionality between the current and voltage is called t,he surge impedance Z and is equal to 4 r C ohms, where I, i s the inductance in henrys for any unit length of the line and C is the capacitance in farads for the same unit length. The current in amperes is equal to the voltage in volts divided by the surge impedance in ohms. The inductanre and caparitance of an overhead line are such that the velocity of a current or volt,age wave (called velocity of propagation) is equal to the velocity of light in free space, which is 984 ft per psec. In most ralrulations the round number 1000 is used. The propagation velority in a cable varies with its construction, but a typical value is 600 f t per psec. The surge impedanre of an overhead line varies with the size of the ronductor and its height aboveground, but is usually between 400 and 500 ohms. A typical value for a cable is 30 ohms. Reflection of Traveling Waves. A change occurs in a traveling wave when it reaches the junction between two conductors of different surge

302


impedance, for example, an overhead line and rahle. The original wave, called the inrident wave, gives rise to two waves at the t,ransition point, namely, a “refracted” wave whirh rontinues on through the second conductor and a “reflected” wave which starts traveling hack over the first conductor. If, at any instant, E is the voltage of the incident wave at the junction, then E X (Z, - Z,)/(Z, ZJ is the voltage of the reflected wave, where Z, is the surge impedance of the first rouductor (over whirh the surge arrived) arid Z , is the surge impedaure of the second ronductor. The voltage of the refracted wave at the junrtiorr is the sum of the voltages of the incident and reflected waves, that is, it equals E X (222)/(Z2 Zi). Reflected and refracted current waves accompany the corresponding voltage waves, the constant, of proportionality being t,he surge impedanre ZIor Z2 of the conductor the wave is traveling oil. A reversal of dirert,ion of a voltage wave, without change i n polarity, reverses t,he direction of flow of current. As indirated by t,he equations, if Z 2 is greater than Z,, a voltage wave reflects positively at, the junctioo and the voltage a t the junrtion (equal to the voltage of the refracted wave) is greater than the vokage of the incident wave. In the limiting rase if 2%is infinite (the line is open), the voltage at t,he junction is double the voltage of the inrident wave. On the other hand, if Z,is less than Z , , the wave reflerts negatively and the refracted wave is less than the incident wave. For the limiting rase of Z2 equal t o zero (the line is shorted t o ground), the volt,age a t the junrtion is, of course, equal t o zero. The current t o ground will equal twire the current of the incident wave. Although neglecting all resistances represents an idealized condition, the simplified relations this makes possihle are useful in many practical situations.

+

+

INSULATION CHARACTERISTICS

It is characteristic of most insulations that t,he maximum voltage which they can successfully ivithstatid varies inversely with the duration of the voltage. Since power systems are subject t o various types of overvoltage, some of long and some of short duration, power distribut,ion equivment is usually required t o withstand at least tivo different types of dielect,ric tests. The first are the so called “lorn-frequency” (00-cyrle) tests, usually of 1-min duration, that cstahlish the ahility of the insulation t o withstand moderate overvoltage of relatively long durat,ion. The others are the “impulse” tests which prove that, the insulation will not break down on vokage surges of high magnitude but short duration. Since the overvoltages produced by lightning are surges of high magnit,ude and


303

short duration, it is the impulse tests that are important as far as protection against these overvoltages is concerned. Basic Impulse Insulation levels. The impulse test which is most commonly used consists of the application of a 155 X 4O-psec full-wave voltage surge of a specified crest value to the insulation of the equipment involved. The crest value of the wave is called the basic impulse insulation leuel (abbreviated BIL) of the equipment. T o simplify the design and appliration of elertrical equipment, the Joint Committee on Coordination of Insulation of the American Institute of Electrical Engineers (AIEE), the Edison Electric Institute (EEI), and the Xational Electrical Manufacturers Association (KEMA) have established a series of Standard Basic Impulse Insulation Levels. These are listed in Table 5.1. It was the intent that the impulse level assigned t o any equipment should he taken from the standard series. This has generally been done, but in some cases the value adopted for a given insulation class is that shown in Table 5.1 for a different reference class. TABLE 5.1

Standard Basic ImDulse Insulation Levels Boric

Reference

impulse in."lation led. kv

Reference

150 200 250 350

138 161 196 230

~

1.2 2.5 5 8.7

15

30 45 60 75 95* 110

23 34.5 46 69 92 115

650 750 900 1050 1300 1550

*The 95-kv BIL was estahlished for rertain types of equiprnrnt in t h e 15-kv class.

The standard BII, of most pover distribution equipment whose insulation class is 23 kv or higher is the value assigned to the corresponding reference class, as shown in Table 5.1. This is true for oil-immersed transformers, oil-immersed induction- and step-voltage regulators, oilimmersed reactors, instrument transformers, apparatus bushings, air switches, and bus supports. However under special conditions, equipment having lower impulse ratings may be furnished. For example, on high-voltage systems (115 kv and above) that are very well grounded, transformers having a RIL one step below the standard value have been successfully applied. These are referred t o as reduced-insulation transformers, while those having a BIL in accordance with Table 5.1 are called fully insulated.

304


The standard RII, of distribution a n d power transformers, reactors and voltage regulators (all oil-immersed), and instrument transformers whose insulation class is 15 kv and below are given in Table 5.2. TABLE 5.2

Standard Impulse Tests

Oil-immersed distribution transformers and "Oltage regulotorl; in.trument trondormers'

Oil-immersed power transformers and current-limiting reactors

lnlUlotior

.I.%, kv

56

Chapped-wore test

I .2 2.5 5.0 8.66 15

30 45 60 75 95

ii 69 88

110

-

1

,mat,

Chopped-rare test

kv

crest, kv

45 60 75

1 %i I

3-11,,

1 .o 1.25 1.5 1.6 1.8

I

!

"II-r.Yc tell

Min time to Rashover,

Crest, kv

X 40

Min time to Ro-rhovar. p*oc

I .5 1.5 1.6

I

* Thr YSIUP

ivm for the 15-kv insulation rlass apply to instrument transiormers oi the 151.-kv ulatiou PLSS. For the 1511-kv class thc full-wavr test is 110 kv and the rlropp~rl-wnrrt p s t is 130 kv with 2.0 ~ S C Cto flashover.

S o industry standard impulse levels have been established for drytype transformers, hut present practire is to use the following combinations of insulation class arid UIL, both in kv: I .2

811 10

2.5

20

I".Yl.ti."

Class

5

25

8.66

35 50

~~

I5

The impulse levels of power circuit breakers, switchgear assemblies, and metal-enrlosed huses for the various voltage ratings, in kv, are as fOllO\VS

: Voltage Roting

811

Voltage Roting

2.4 4.16 7.2 13.8 14.4 23 34.5

45 60

46 69 92 115 138 161 230

* 95 ior rnctal-rlnd gear with

75' $3 110 150 200

oillcss hreakcrs.

811 250 350 450 550 650 750 900



305

Impulse testing of rotating machines has not been adopted. I t is generally considered that their impulse level is the crest value of the 60-cycle dielectric test. The rms value of the latter is twice rated line-toline voltage plus 1000 volts. Chopped-wave Tests. In addition to the 145 X 40 full-wave test, oilimmersed transformers (reactors and voltage regulators) and instrument transformers are given a “chopped-wave” test. In this the applied voltage is built up at a predetermined rate and then reduced substantially to zero by sparkover of an air gap. The crest voltage reached and the minimum time to sparkover of the air gap for the chopped-wave tests are given in Table 5.2 for equipment having an insulation class of 15 kv or below. For the higher voltage insulation classes the crest value of the chopped wave is approximately 1.5 per cent higher than the BIL and the minimum time to sparkover of the air gap is 3 psec. PROTECTIVE EQUIPMENT

The protection of electrical equipment against the overvoltages caused by lightning depends primarily upon the proper application of lightning arresters. How lightning Arresters Operate. Lightning produces overvoltages between the line conductors of a power system and ground. A lightning arrester limits the overvoltage by providing a conducting path of relatively low impedance betmeen the line and ground. The resulting current flow to ground, through the surge impedance of the line, limits the line-to-ground voltage. But this low-impedance path must not exist before the overvoltage appears, and it must be broken immediately after the voltage has returned t o normal. This is accomplished in a lightning arrester by (1) an enclosed gap, or several gaps in series, which will withstand the normal operating voltage but sparkover and become conducting a t some higher voltage; and (2) a device which in conjunction with the gaps interrupts the flow of currentfrom the power system, called “follow current,’’ after the lightning surge has passed. Two different principles are used to interrupt follow current, and arresters may he classified according to which of these they use. Expulsion-type and Valve-type Arresters. As the term implies, expulsion-type arresters interrupt the Row of follow current by expulsion action. The gap is arranged so that upon sparkover the arc must pass over the surface of gas-evolving material; for example, the gap enclosure may he a gas-evolving fiber tube or the gap may he filled with celofiber spheres. As the gas is emitted it rushes out through a suitably placed opming in the arrester case, blowing out the arc. Interruption takes

306


A N D PROTECTIVE MEASURES

place as the a-c current goes through zero. The action is similar t o the operation of an expulsion fuse. I n the valve-type arrester, on the other hand, interruption of follow current depends upon having in series with the gaps a column of material whose resistance varies inversely a s some power of the voltage applied. Hence, this “valve ” material exhibits a relatively low resistance when the overvoltage due t o lightning exists, but as soon as the voltage returns to normal its resistance increases t o a high value. This reduces the magnitude of the follow current to a value which can be interrupted by the series gaps. Xormally interruption takes place the first time the a-c current goes through zero. The construction features of one design of valve-type arrester are shown in Fig. 5.14. Expulsion-type arresters have assigned current interrupting ratings

FIG. 5.14 construction.

A valve-type lightning arrester with section removed to show features of


307

slid should not he applied to systems whose fault current exceeds such ratings. Furthermore since some of the gas-producing material is destroyed each time the arrester operates, there is a limit t,o the number of operations t o which they can he safely subjected. Valve-typc arresters have ueither of these limitations. The expulsion arrester also has a higher sparkover voltage, although following sparkover it exhibits a lower resistanre t,o the flow of lightniug discharge current than does the valve-type arrester. Finally the gaseous disrharge from an expulsiorrtype arrester makes it uusuitable for moiiriting wit,hin equipment enclosures or in close proximity to other elertrical apparatus. For these reasons the valve-type arrester is used almost exclusively for the prot,ection of equipment on industrial power systems. Voltage Rating of Arresters. The voltage rating of an arrester is defined as the highest a-c voltage (rms value) hetween its line and ground terminals a t which it is desigued t o perform its operating duty cycle. I n effect it represents the highest voltage at, which it is guaranteed to interrupt the follow current after sparkover on a voltage surge. It does not represent the voltage at which the arrester sparks over; in fact, industry standards specify that an arrester shall not sparkover at any 60-cycle voltage less than 150 per cent of its rating. As n.ill be shown (see Application Proredure) the proper voltage rating of a n arrester for any system depends not only on the syst,em voltage but also on how the system is grounded. Protective Characteristics of Arresters. The two characteristics of a lightning arrester which determine the degree of protection it can provide are (1) its impulse sparkover voltage and (2) its discharge voltage, i.e., the voltage which appears across its terminals during the passage of discharge current. The latter is sometimes referred t o as the I R voltage drop or simply I R drop. Two different sparkover voltages are usually published by the arrester manufacturers. One is the “critical sparkover voltage” with a l!i X 40psec wave, i.e., it is the crest value of the 1>6 X 40 wave which will cause sparkover on 50 per cent of the applications of this wave. Sparkover occurs on the tail of the wave. The other is the average voltage at which front of wave sparkover occurs with the voltage wave rising at the rate specified in the AIEE standards for arrester tests, namely, 100 kv per psec for each 12 kv of arrester rating. This sparkover voltage is generally higher-as much as 50 per cent higher for some arresters-than the crit,ical sparkover voltage for a I f 5 X 40-psec wave. Arrester discharge voltages usually published are the average crest values of the voltage appearing across the arrester terminals when discharging a 10 X 20-psec current wave having various crest values such as 1500, 3000, 5000, 10,000 and 20,000 amp.

300


From the average protective rharacteristirs of lightning arresters xyhivh are puhlished, the masimum values can be determined h y means of iirdustry rerogniaed toleraiires. As shown in Table 5.3, these give the amount hy whirh the masimum sparkover and disrharge voltages of a n arrester may be eupeited to exceed the average values. The various types of arresters listed in Table 5.3 are defined under the heading (:lassification of High-voltage Arresters which follovs. TABLE 5.3

Tolerances in Performance of Valve-type Lightning Arresters

Type of Arrester

1

ayeroge "(IiYe, per cent

I Sparkover voltage .......... ................. .............

Distribution. Line Stotion.

25

Discharge voltage

20

20 15

I5

10

Effect of Altitude. Since the sparkover voltage of a gap varies with the atmospheric pressure, the protective characteristics of arresters are afferted by the altitude a t which they are installed. This is true even if the arrester has a sealed gap since the seals employed are not expected t o maintain a pressure different from the surrounding atmosphere for any extended period. Standard arresters are considered suitable for altitudes up to GOO0 ft. Special arresters are available for altitudes of 6001 t o 12,000 ft and for altitudes of 12,001 t o 18,000 f t . Classification of High-voltage Arresters. Arresters in ratings of 1000 voks and higher are classified in accordance with their principal charact,eristirs and field of application as follows: 1. Distribution-type arresters 2. Line-type arresters 3. Station-type arresters Distribution-type arresters are available in voltage ratings of 1, 3, 6, 9, 12, 15, arid 18 kv. Though designed primarily for the protection of dist,ribut,ion transformers, they are also used to protect other equipment such as metering and switching devices, voltage regulators, distribution rapacitors, and cable. The arresters are small, lightweight units t h a t are readily mounted on poles or crossarms, have reasonably good protective rharacteristics, and are very low in cost. Line-type arresters are available in voltage ratings of 20, 25, 30,37, 40, 50, GO, and 73 kv. They are relatively small and lightweight, are moderate in cost, and have good protective characteristics. They are used for the protection of the smaller transformers and substations in the mediumvoltage range.

309


From the average protective rharacteristirs of lightning arresters xyhivh are puhlished, the masimum values can be determined h y means of iirdustry rerogniaed toleraiires. As shown in Table 5.3, these give the amount hy whirh the masimum sparkover and disrharge voltages of an arrester may be eupeited to exceed the average values. The various types of arresters listed in Table 5.3 are defined under the heading (:lassification of High-voltage Arresters which follovs. TABLE 5.3

Tolerances in Performance of Valve-type Lightning Arresters

Type of Arrester

1

ayeroge "(IiYe, per cent

I Sparkover voltage .......... ................. .............

Distribution. Line Stotion.

25

Discharge voltage

20

20 15

I5

10

Effect of Altitude. Since the sparkover voltage of a gap varies with the atmospheric pressure, the protective characteristics of arresters are afferted by the altitude a t which they are installed. This is true even if the arrester has a sealed gap since the seals employed are not expected to maintain a pressure different from the surrounding atmosphere for any extended period. Standard arresters are considered suitable for altitudes up to GOO0 ft. Special arresters are available for altitudes of 6001 to 12,000 ft and for altitudes of 12,001 t o 18,000 f t . Classification of High-voltage Arresters. Arresters in ratings of 1000 voks and higher are classified in accordance with their principal charact,eristirs and field of application as follows: 1. Distribution-type arresters 2. Line-type arresters 3. Station-type arresters Distribution-type arresters are available in voltage ratings of 1, 3, 6, 9, 12, 15, arid 18 kv. Though designed primarily for the protection of dist,ribut,iontransformers, they are also used to protect other equipment such as metering and switching devices, voltage regulators, distribution rapacitors, and cable. The arresters are small, lightweight units that are readily mounted on poles or crossarms, have reasonably good protective rharacteristics, and are very low in cost. Line-type arresters are available in voltage ratings of 20, 25, 30,37, 40, 50, GO, and 73 kv. They are relatively small and lightweight, are moderate in cost, and have good protective characteristics. They are used for the protection of the smaller transformers and substations in the mediumvoltage range.

310



t,he arresters must withstand is 100,000amp for thestation typeand 65,000 amp for the distribution and line types. TABLE 5.4

Industry Average Protective Characteristics of Valve-type Lightning Arresters A v e r a g e impulse rporkover voltage on AlEE test wove, kv

V0ltog

rating, kr

Average discharge oltage with 10,00O-~mp 10 x zo-psec CUrlent wave, kv

__ Distribution OrreSler.

~

~

3 6 9 I2 I5 18

I8 34 48 61 71 84

13 23 35 43

53

...

Line

30 37 40 50 60 73 97 I09 121 145 169 195 242

75 93 110 136 147 183 220 267

... ... ... ... ... ... ...

44 55 69 78

11 22 33 44 54

Line

Or,e.te,S

20 25

I5 30

O,,&e,*

72 89

92 Ill

I06

I35

131 136 178 214 261 345 388 430

I64

51s 602 691 860

I77 222 271 328

... ... ... ... ... ... ...

72 90 108 132 144 179 217 262 349 394 438 523 610 698 872

Arresters and Capacitors for Rotating-machine Protection. A variant of the station-type arrester designed particularly for rotatingmachine protection is offered by some manufacturers. One version (see Fig. 5.16) has characteristics similar t o that of standard station-type arresters but differs mechanically in that it has a porcelain top with the line-terminal connection brought out through the center. This allows placing the three arresters of a three-phase installation close t o each other, thus reducing space requirements to a minimum. The arresters are available in voltage ratings of 3 t o 27 kv with the 3-, 4 . 5 , 6-, 7.5-, 9-,


31 I

12-, aiid 15-kv rat,ings of particular interest for industrial npplicat,ions. The 4.5- and 7.5-kv voltage ratiirgs are not, available in t h e standard station-type arresters. They are iiicliidrd in this line to give bet,ter protection to 1.16- and 6.0-kv machines tliaii rim be provided by t h c standard 6- aiid 9-kv arrest,ers. Tlic latter ~vouldotherwise be required where the paver syst,cnis are riot, sufficiently well grounded to permit t,lie USC of 3- and ti-kv arresters on 4.16- and G.!bkv rnachiiics ( Arrester T’oit,age Ratings). The coristructioii fcatorcs niid additioiial voltage ratiiigs available make these arresters dcsirahle for iit,her app1ii.atioiis such as t.he protection of switchgear. Surge protective capacitors are also available for rotatiiig-mii~hiiie protection. They are used to reduce tlic stcepriess of the wdve front of lightning surges aiid arc available in ratirigs of OM50 volts with 1 .O pi per pole, 2.1, 1.16, 1.8, arid 6.9 kv with 0.5 pi pf’r pole, and 11.5 and 13.8 k v with 0.2.5 pf p t pole. ~ l’liese capacit,ors differ from thc staridad porver-fact,or impr(iviiig capacitors i i i that they are designed t o withstaiid higher test, voltages and have low interrid inductance. A typiciil unit is shovii in Fig. 5.17. Low-voltage Arresters. For thc prntectiou of etluilimixit on circuits whose line-to-ground voltage is iri the 110- to 125-volt range, a 175-volt

FIG. 5.16 Rotating-machine form of station-type lightning arrester rated 6 kv.

FIG. 5.17 Surge protective capacitor rated 6900 volts, 25 to 60 cycler, 0.5 ilf.

312

SYSTEM OVERVOLTAGES--CAUSES AND PROTECTIVE MEASURES

liglituing arrester is avsilal-lie. This is built in a two-pole lorm; so a single unit will provide protect,ioii to the common 1 15i230-volt siriglephase tliree-wire grounded-iieutral circuit,. A t y p i d iiistallation is showii in Fig. 5.18. For a two-wire circuit, grounded o r 1 oiie side, the two poles of the wrcster arc generally roniiecied in parallel between the uiigrouiided h i e arid gruund. For three-phase circuits such as those supplied from a208Yjl20-volt grounded-imitral system, t x o arresters arc required. For the protection of equipmerit on higher voltage circuits-up t o 600 volts-~-~twu forms of arresters areavailahle, both rat,cd 650 volts. One has a port:elaiii housing (see Fig. 5.19), is for oiitdoor service oiily, arid is availablein a singlepoleaiidatno-poleform. T h e other has t i niet,al enclosure (see Fig. 5.20), is suitable for either indoor or outdoor service, axid is availsblc in one., two-, arid t,lirce-polc forms. This unit also has better prok c l i v e characteristi(,s and so FIG, 5.18 lnrtallotion of two-pole 175-volt is t,hc oIic usually selected for lightning orrester on o 115,'230-volt single-phore protection of indust,rial plaiit three-wiro circuit. equipmeti t. Arresters for D-C Systems. .krrcst,ers designed for use on a-c power systoms are iiot getierally suitable for service on d-c employed t o interrupt follow r'urrrirt is not cffectiv diics not periodir:ttlly go through zero. Arresters, hinrevcr, arc availrtiilo for d-? scrvicc. The moderir forms arc simply capacitors having iiot less tiinti 4 pf of capnt,itaricc. Tiicy are coiiriected from line to ground arid limit, tlit. tw:st, - i d t i e of a volt,age surge by absorbing the current as a charge o i i ttic capwitor. fleiice llie effectiveiiess of the arrester in limit,irig the \-oltnge of ail iticwmitig surge depends upon the duration as i\-ell as the magriitude of the surge. lrmvever, it, also ser t o slope t,he froiit of tiit? ivavc a r i d tliiis reduce the turii-to-turn voltage sircss on the d-c rotating mnr.liities. 'I'hc arresters are available i i i three voltage classes, iiamely, Obi30 volts (illustrated iii Fig. 5 . 2 ! ) , 751-2000 v o h , and 2001:?&00\-o1ts. (I


FIG. 5.19


313

Single-pole lightning arrester

FIG. 5.20 Three-pole lightning or-

with porcelain housing roted 650 volts, for outdoor service.

rester in metol core roted 650 volt>. for indoor or outdoor service,

FIG. 5.21 cirwiti.

Capacitor-type lightning orrerter rated 0 to 750 volts, 4 rrf for w e on d-c lnruloting cop and sleeve removed ot one end to show terminol.

314


APPLICATION PROCEDURE

Every exposed overhead line distributing power within or supplying power t o an industrial plant represents a possihle sourre of destruitire overvoltages. Lightning arresters should be so applied that a voltage surge from any of these sources will be reduced to a ralue ~ v e l lheIoi\- the impulse strength of all apparatus involved. The application procedure consists of (1) selecting t,he voltage rating of the arresters t o he used, (2) choosing t,he types of arrest,ers needed, and (3) determining where the arresters should be located to ensure adequate yet economical protection. Selection of Arrester Voltage Ratings. The protective characteristics of an arrester are hetter and, in general, its cost, is lower, the lower its voltage rating. On the other hand, if the line-to-ground system voltage after sparkover of a n arrester should exceed its voltage rating, the arrester may not interrupt follow current and then iI-ill fail very quickly. This makes it important t o determine the maximum lilie-to-groutid system voltage at the point at which the arrester is applied. 111 so doing it is necessary t o consider all abnormal conditions which ran exist, particularly those conditions which are likely t o exist when the arrest,er sparks over. Under normal balanced operatirig conditions, the voltage from each line t o ground on a three-phase system is the syst,em line-to-line voltage divided by the square root of 3. This applies vhethcr the system neutral is grounded or ungrounded. There are, however, many abnormal rotiditions which can occur that result in higher t,hari normal line-to-ground voltages. Hut the one that is most likely t o exist a t the time of arrester sparkover is a line-to-ground fault. For example, if a lightning stroke causes flashover and hence a fault on one phase of a transmission line, the voltage indured on the sound phases is apt, t o cause sparkover of the arresters connerted t o these phases. These arresters must then interrupt, follow curreut, with a line-to-ground fault on the system. The voltage ratings of arresters are, therefore, generally selected 011 the hasis of the system voltage t o which they arc subjected under line-to-ground fault condi t,ions. The voltage from sound conductors t o grouud with a line-to-ground fault 011 a system depends upon how the system neutral is grounded. For the usual ungrounded or resistailre-grounded system, t,his vokage will be essentially equal to the system line-to-line voltage, and the lightning arresters used must be selected 011 this basis. Thcse are siimetimes referred to as ‘‘ 100 per cent arresters.” However, for solidly grounded or reactance-grounded systems the sooiid-rotidurtor-to-ground voltage with one line grounded may be as low as the system line-to-neutral volt-


315

age. It depends upon the relation between the zero- and positivesequence impedances of t,he syst,em. For example, if the ratio of zerosequence reactance X Ot o the positive-sequence reactance X I is positive and less than 3 and the rat,io of the zero-sequence resistance R , t o the positive-sequence reactance X I is less thau 1, the voltage from sound conductors to ground will not exceed 140 per cent of the system liue-toneutral voltage or about 80 per cent of t,he system line-to-line volt,age. Such a system is said to he “effei.t,ivelygrounded,” and t,he arresters used are referred to as “80 per rent arrest,ers.” Some syst,ems are grouudcd so that arresters of even lower voltage rating can he used as far as the orervoltage caused by line-to-ground f a u h is concerued. This, however, should he done only after a careful check of the possible overvoltages from all sources t,o make sure that v o h g e s in excess of t,he arrester rating are not likely to occur at the time of sparkover. Table 5.5 lists the voltage ratings of arresters usually selected for (1) ungrounded or resistauce-grouuded systems and (2) “effectively grounded” systems. Selections are show1 for all system voltages likely to he encountered in industrial plants. As shown in Table 5.5, 3-kv arresters are often used on 2.4/4.1C,Y-kv grounded-oeutral systems and 9-kv arresters on 7.2/12.47-kv grouridedneut,ral syst,ems, akhough in t,hcse cases the arrester rating is only 125 per cent of t,he nominal system line-to-neutral voltage. Before using these lmi-er rat,ed arresters, the maximum operating voltage and the rise iu soulid-conduct,or-to-ground rokages with a linn-t,o-grouud fault, should be determitied t o make sure that under these conditions the voltage applied to the arresters will not exceed their rating. I n geueral they should not be used on industrial pmver systems unless (1) the ratio of zero-sequenre reart,ance X o to the positive-sequence reactance X I is less thau 1.5 and (2) the ratio of the zero-sequence resistance Ro t,o t,he positive-sequence reactance X I is less thau 0.5. Even though a system meets the qualifications of an eflectively grounded system at the power source, it may not a t other points in the system because of the impedance of intervening lines. Furthermore, the system may be “effectively grounded” under uormal operating conditions, but certain faults or other emergencies may result in the opening of switches which leaves a portion of the system ungrounded but still energized either from generators or from mot,ors whirh can temporarily act as generators. Such possibilities should he considered before selecting the voltage rating of arresters to he applied on what appears t o be an effectively grounded system. Choice of Arrester Type. Where the arrester voltage ratiug required is 3 t o 15 kv, a choice must be made between the distribution-type and the station-type arrester. Similarly, if the rating required is hetween 20

316


TABLE 5.5

Voltage Ratings of Arresters Usually Selected for Three-phase Systems Voltage roting of arrester, kv

Nominal system voltage, kv

0.120/0.208Y 0.240 0.480 0.600 2.4 2.4/4. I6Y 4.16 4.8 6.9 12 7.2112.47Y 13.2 (or 13.81

23 34.5 46 69 115

I38

Sy*tom "e"Ir.1 ungrounded or 'eiirtonce groundeq

0.65 0.65 0.65 0.65

System neutral effectively grounded

0.175 0.65 0.65 0.65

3

t 4.5,;

4.5. or 6 4 . 9 or 6 6

3

or 6 4.5. or 6 4.5* or 6

6

7.5*or9

12

I5

9 t o r I2

15 15

12

25 37 50

20 30 40

73 121 145

60 97 121

* The 4.5- and 7.5-kv arresters are available only

;he station type. 1 less than that necessary to make the system "effectively grounded" (see accompanying text)

t The use of these arresters requires an X o / X ,

I

and 73 kv, either the line-type or the station-type arrester must he selected. The value of the equipment protected and the importance of uninterrupted service in an industrial plant generally warrants the use of stationtype arresters throughout their voltage range. However, for the smaller (liquid-filled) transformers and substations, say 1000 kva and less, distribution- or line-type arresters are frequently used. Similarly, for the protection of short lengths of cable joining overhead lines and apparatus, these lower cost arresters are generally chosen. They are also used to protect small breakers, disconnecting switches, and similar outdoor switching equipment. Finally, distribution-type arresters are often used in the protection of rotating machines, thereby supplementing the protec-


317

tioii provided h y statioii-type arresters (sec l'rotcrtioii of h-C Iliitatiiig hlarhiiirsj. Location of Arresters. The ideal location of lightiiiiig arrestcrs, from the staiidpoi~rt,:if the prutrrtioii whirh they provide, is directly at the terminals of t h e apparatus heiiig protwted. .kt this location, aiid with the arrester groutid leads i.oiinerted direi.tly to the tank, framc, or other metallii, strnctiire h i r h supports the iiisrilated parts, thc surge voltage applied to the itisrtlatioti will he limited to the sparkover vultage aiid the discharge voltage of the arresters. Iii some cases, howe\-er, it might I)P quite costly or aivk\\-ard t o muiiiit the arresters at tlie apparatus tcrmiiials. Furthermore, i i i somc iiistatlations, if the arresters are mo\-cd away from the trrmitials 11f the protected equipme~it.a single set of arresters caii he lorated \\-here they will intercept all lightiiiiig surges to two or more pieres of apparatus. H o ~ e v e rsuch separation hetwecti lightning arresters alid thc eqriipme~it,does mean some itiiwase i n the magiiitude of the voltage surge 11-hivh is applied t o t h r eiluipmmt. First, the equipmelit protevted will ofteir have a highrr surge impedtilice than that iif the h i e or mhle over \\-hich the lightiiiiig srirge arrives. This means that thc voltage wave will refle1.t positix-ely nt the equipmetit termiiials aiid the 1-oltage rearhed at this poiiit n-ill al\\-ays he lriglrcr than the sparkover v d t a g e of the arrester. T h e amoriiit of the itirvmse will depend upoii ( I ) the steepiirss of the froiit of the srirgc viiltagr, (2) the relative surge impedance of the eqnipmeiit aiid the circuit hetiweti the arrester and the protected equipmeiit. (3) the sparkowr \-iiltnge of the arrester, and (4) the length of the rirt,nit hrtivreli the arrester and the protwtrd eiluipmeiit. The greatest i i i i ~ r a s riii voltage wciiss if the cirruit is iipeir at the protected eiluipmetit (iiititiite surge i m p d a t i w j . 111 this rase tlie voltagr will IK dinible the arrester sparko\-er voltagr if the sepitration distairre is such that parko over ownrs before tlic voltage wave reflected from the eiliiipmriit arrives hack at the arrestrr, U'ith less separatiiiii the voltage will iiot iiirreasr a s miidi. This is showi h y the iwrves of Fig. 3.22. Citrve ;I applies if the overhead liiie. over \\-hirIi the surge arrives. estends past the arrester to the priitri,tid eqiiipmetrt, i\-hile curre B applies i f a i,ahle of typical chnravteristiis forms the cirruit het\\-eeti the arrester aiid t h r proterted equipment. The voltage whirh appears arross an arrester after spnrkovrr, i.e., its disrharge voltage. is also magnified by separation atid priidrwrs ii Iiighrr voltage at the protevtcd equipmeiit. Fnrthermiirr. if thew is ail? appreriable lciigth of lead hetween t h e h i e rolidrii~tiir atid the arrester or het\\-een the arrester atid griiulid, the voltage drop wross surh a lead adds to the discharge voltage of the arrester aiid is also itiiwased by separation betxi-eeii the arrester and the protected eqiiipmrtit. Finally, if a11 arrester located away from the protected equipment has a11isolated co~iiicctioiito

318


ground, the additional voltage drop resulting from discharge current flowing through the ground resistance also adds to the line-to-ground voltage a t the arrester and a magnified addition appears a t the protected equiqment. Certain installation practices help to reduce the difference between arrester discharge voltage and the corresponding voltage a t the pro-

FIG. 5.22 Effect of reparotion between a lightning arrester and the protected equipment on the rotio of the maximum voltage a t the equipment to the sparkover voltage of the arrester (doer not include any effect of the voltage at the arrester following its sparkover).



319

tected equipment. For example, where an arrester is connected between an overhead line and ground, the leirgth of the line and ground leads can both he reduced to a minimum by use of the V connection. The arrest,er is placed a t ground level, and the line coriductor is brought down to the arrester and then back up, forming a V. The angle hetween the two sides of the V should not be less than 30" to minimize their mutual inductance. The effect of high ground resistance a t the arrester ran be minimized by interconnection of the arrester ground terminal with the tank or enclosure of the protected equipment, the station steelwork, and the ground mat. Finally, where the circuit hetween an arrester and the protected equipment consists of cable having a contirruous metallic sheath, the arrester ground terminal should he connected directly to the cable sheath and the sheath connected to the equipment tank or enrlosure. In this may arrester lead lengths can he kept to a minimum and the effect of ground resistance eliminated. More specific recommendations covering the application of arrest,ers for the protection of various types of equipment,, including suggested maximum separation distances, are given in the remainder of this chapter. PROTECTION OF TRANSFORMERS

Transformers generally constitute one of the must important elements of any industrial power system. Furthermore they are frequently connected directly to exposed overheadlines and so are suhject to destructive overvoltages unless properly protected by lightning arresters. A liquid-filled (oil or askarel) transformer having arresters mounted a t its terminals is well protected against the overvoltages produred hy lightning, with the possible exception of those result,ing from severe direct strokes to the transformer terminals or to the conneitcd lines close t,o the transformer. Furthermore, the possibility of such direct strokes can he essentially eliminated by proper shielding. Often, howerer, in order t o protect (with the same set of arresters) switching and other equipment located between the transformer and the exposed lines, or to protect two or more t,ransformers connected to the same line, it may appear desirable to mount the arresters some distance away from the transformer terminals. The maximum permissible separation distanres depend, among other things, upon the magnitudes and rates of rise of the voltage surges which can he expected to reach the arresters. Until more statistical data on these surges are available, no determination of permissible separation distances can be considered final. Hou,ever, making That appears to he reasonable assumptions, a Working Group of the AIEE Suhcommittee on Lightning Protective Devices (of the AIEE Committee on Protective Devices) proposed the maximum separation distances shown in Table 5.6.* The installation conditions on which these distances are

* Ser AIEE Misccllaneous Paper 51-285.

320

SYSTEM OVERVOLTAGESS-CAUSES AND PROTECTIVE MEASURES

based arc that ( 1 ) the transformer is fully insulated (liquid-filled), (2) statioii-type arresters are used, (3) arrester lead lengths are zero (V r~~nncctioir or eqnivalent), (4) ground resistance is negligible, and (5) the transformer is a t the elid of a single overhead line (the worst condition) with the arresters located on the line directly in the path of incoming sr1rgcs. TABLE 5.6

Separation Distance Permissible between Station-type Arresters and Transformer Bushings Separation diitonce, ft

Tiomformer i"lYl.ti0" CI.I.,

kv

Botic impvke inrvlotion

IWel, kv

System neutral

System neutrd

ungrounded or esistance grounded 1100% arrosten)

effectively grounded

180% arresters)

-___ 25 34.5 46 69 92 115 138

I50 200 250 350 450 550 650

25 25 25 ~. 25

30 35 30

70 70 70 70 75 85 95

For transformcrs of lower volt,age ratings (15-kv class and below) which are not covered i t t Table 5 . 6 , permissible separation distances have not becti estahlished. Severtheless it appears that for these ratings any apprecialile scparatiott should be avoided, that is, the arresters should he mounted 011 t,he transformer itself or closely adjacent to it. In ratings of 15 kv and helow, transformers are often connected t o exposed overhead lines through a length of cable. I n this case fully insulated liquid-filled power transformers ronnerted t o the overhead line through a cable having a continuous metallic sheath will be adequately protcvtrd by statioii-t,vpe arrestcrs located at the junction of the cable and the overhead l i t e. Thc arrester ground terminals must he connected directly t o thc catilc sheath, and at the transformer the cable sheath must lie rotinerted t o the transformer tank. If the transformers are of the distribution rathcr thaii the po\ver rlass or if distrihution-type rather than station-type arresters are provided at the junction of the cable and overhead line, it may he necessary to add a set of arresters at the t,ransforme,r terminals to eiisurc adequate protection. Dry-type transformers, \\-hose impulse level is ahout half that of the


321

liquid-filled type, are not generally recommended where connection t o exposed overhead lines is required. If used they should definitely be protected by station-type arresters located at the transformer terminals regardless of whether the connection t o the exposed overhead h i e is direct or through a cable. If a liquid-filled transformer is connected t o an exposed overhead line only through another transformer which is adequately protected by lightning arresters, no additional protection is required. In the same situation a dry-type transformer should, preferably, have station-type arresters mounted a t its terminals since analysis iiidicates that the surges that come through the other transformer can have magnitudes greater than the recognized impulse level of the dry-type units.

PROTECTION

OF

METAL-CLAD SWITCHGEAR*

Metal-clad switchgear (used on 2.4 t o 13.8-kv circiits) is often connected t o an exposed overhead line either directly from roof bushings or through a moderate length of cable. In either case it is esseiitial that adequate lightning protection be provided. If the switchgear is connected directly t o the overhead line from roof bushings, lightning arresters should always be provided a t the gear. Although the arresters are sometimes mounted on the first structure away from the gear which supports the overhead h e , the resulting separation between the arresters and the protected equipment substantially reduces the effectiveness of the protection. Heiice locating the arresters at the gear is definitely recommended. They may be mounted on the roof of the switchgear enclosure adjacent t o the bushings or inside the enrlosure but on the line side of the breaker. Since the former arrangement generally requires an extra ground bus, the inside mounting is usually selected. The arresters should preferably be of the station type (rotating-machine form), but space limitations may sometimes make i t necessary t o use the distribution type. The voltage rating selected should he the lowest that is consistent with the system voltage and method of grounding. If the metal-clad switchgear is connected by cable t o the exposed overhead line, the first requirement is that arresters be provided a t the junction of the cable and the overhead line in order t o protect the cable. Then if the cable does not have a continnous metallic sheath, a second set of arresters should be provided at the switchgear. I n this case dis-

* Adapted from Dillow, Gittings. Halherg. Hoffman. Howard. and Hontrr, Lightning Protection of Mptalclad Saitchgear and Unit Substations Connected to Overhead Lines, Gen. Eke. lieu., March, 1949.

322

SYSTEM OVERVOtTAGE5,-CAUSES


tribution-type arrcsters are usually used a t the junction, but those at the switchgear should preferably be of the station type (see Fig. 5.23). If the cable eoiiiiectiiig niet,al-clad gear t o an exposed overhead line does have ii coritiiiiious metallie sheath, the set:orrd set, of arresters at the gear may or may not be rec.]uired. It depends upon (1) tho iiisulation level of the gear, ( Z j the type and ... . . . .. voltage rating of the arresters provided at the junction, aiid (3) the length of the cable. An analysis of this ease was made on t,he basis of the followiiig assumptions : 1. The arrestors at the jiinctioii maintain a voltage at, this point which does not exceed the sparlrover voltage of the arrester as given in Table 5.8. 2. The maximum voltage at the switchgear must be limited to 80 per cent, of its BIL. 3. The volt,age waves which appear on the overhead liue arid reach the cable junction have a const,ant rate of rise which does not exceed 1000 kv per psec. 4. The surge impedance of the overhead line is 500 ohms, and t h a t of the cable is 30 ohms. FIG. 5.23 Stofion-type lightning orresterr 5 . The velocity of propagat,ion (rotating-mochine form) mounted in metalof the surge iii the cable is GOO f t clad switchgear. per $see. The r e s u k of the analysis are shown in Table 5.7. I n all cases the grouiid terminal of the junction arrester should be coriiiected to t,lre cable sheath as me11 as t o ground, aiid at the switchgear the cable sheat,h should he eonri d to the ground bus (see Fig. 5.24A). This is essential if no arresters are provided in the gear aiid is desirable in any case. Where large single-coiiductor cables are used, it may not be desirable t o ground bot,b ends of the sheath because of excessive sheath curreiit. I n this case the lightning-arrester ground terminal should still be coririected directly t o the cable sheath arid the sheath grounded at the switchgear, hut the connection t o ground at the arrester should be made through aii isolatiiig gap, as shown in Fig. 5.24R.


TABLE 5.7

Protection of Metal-clad Switchgear Connected to Overhead lines through Continuous Metallic Sheath Cable Arresters in witchgeor (required or not requiredl

System voltage

Voltage roting m d

811 of switchgear. kv

4.16

Neutral ungrounded

effectively grounded, kv

or

resistance grounded, kv

2.4or4.16

13.8

4.8 4.8 or 6.P 11.5 13.8

’The use of

Voltoge rating of .2,T&*,S

With dirtribution-typo

iundion, k?

arresters 01

the iunction

4.5 6

....

Not required Not required

Required

(55 h)f

t

Not required Not required Not required Not required (30 ftlt Required*

4.5

4.8

6

6.9

7.5

(75 tilt

t

P

6.P 11.5 13.8

12 15

-

With station-Wpe arresterr rrt the junction

t

Not required

3.

2.4 4.16 4.16

160 811)

I95 8111

323

Required Required Required

arresters on a 4.16-kv system requires an X d X , ratio IPSS than that necessary t o make the system “effectively grounded” (see Selection of Arrester Voltage Rating). t The 4.5- and 7.5-kv arresters are available only in the station type. t Arresters required in snitchgesr if length of cable exceeds this value. TABLE 5.8

Y

Sparkover Voltage of Arresters Used in Analysis of Protection Rewired for Metal-clad Switchgear Sparkover voltogs. k r

V0lt.go rating of arraters, kv

-I

Distribution-type .r,der*

Stofion-type .lr,e*t*r*

I

3

22

15

6

42.5 60 74 81.5

25 37 52 64

P 12

15

324

SYSTEM OVERVOLTAGLS-CAUSES AND PROTECTIVE MEASURES

type arresters and a set of protective capacitors (as used for rotatingmacliitie protection) at the junction of the rable and overhead line. The ground terrniual of both the arresters aiid the caparitors should be connected to the rable sheath as \yell as t o ground (directly or through a n isolating gap), aud the (.able sheath should he eoiniected t o the ground bus a t the switchgear.

EXPOSED OVERHEAD LINE

11,

a l l - P ~ ~ I?

CABLESHEATH

THIS ARRESTER MAY NOT BE REQUIRED

-

4

q&2 ’

-T

. 1

-<4p

~

4 -

PROTECTION OF SUBSTATIONS

Outdoor substation equipment should be protected against direct strokes of lightning by proper shielding. This may take the form of steel



325

masts (see Fig. 5.25) or extensions of tlie steel structure arrarrged so as t o divert, t o themselves all lightning strokes which might otherivise strike a bus, disconnecting switch, bushing terminal, or other exposed currentcarrying part,. The mast,s, or eqiiivaleiit, are designed so as t o form a “protective zone” ivithin which all vulnerable parts will lie. JVith a single mast the protective zone is usually corrsidered to be a cone hax-ing its apex at the top of the mast and whose sides make an angle with the vert,ical of 30 t o 15”. With two or more masts the protective zotie of each is iricreased somewhat in the area betweeir t,hem. 1‘his may be considered as an iiicrease in the angle (made with tlie vertical) of the side of each protective cone which lies bet,rveeii two masts. With the usual spacings between masts, this angle may iiicrease to 60’. It is also desirable t,o shield the inc,omiiig lines, by ovcrhead ground wires, for a distarice of a t least 2000 ft out from the statiou. This r e d i m s the possibility of direct strokes t,o tlie lines i l l t,he riciiiit,y of the statioir and thus limits both the Inagnitude arid rate of rise of the voltage surges which reach tlie station. The overhead ground wire should lit: grounded. with as low a ground resistarrce as it is practicable t o obtain, at each p o k , and i t should he connected to the ground bus a t the substat,ion. Loir ground resistance is particularly important for the ground rotinertion at the first few Doles adjacent to t,he substation.

FIG. 5.25 Substation with lightning masti for direct stroke protection and station-type lightning arresters far protection agoinrt surges entering the station over the incoming liner.

326


In addition to proper shielding against direct strokes, substation equipment should be protected against voltage surges entering over the incoming lines by the proper application of lightning arresters. The type, voltage rating, and location of the arresters should he selected (by the methods that have been described) so as to protect all the equipment in the substation. Typically, a set of arresters is required on each exposed overhead line as it enters the station to provide protection to disconnerting switches, buses, etc. Whether or not these arresters will also protect the transformer depends upon the system voltage, method of grounding, and circuit distance between the arresters and the transformer (see Protection of Transformers). It may prove necessary to install an additional set of arresters at the transformer. Although the feeders from an industrial-plant substation are usually underground cahle circuits with no lightning exposure, occasionally overhead feeders are used. These represent additional sourres of voltage surges from whirh the suhstation equipment should he protected by the proper application of lightning arresters. PROTECTION

OF

AERIAL CABLE

The best protection that can he provided for aerial cahle against direct lightning strokes consists of grounding the messenger and sheath a t every pole and securing as low a ground resistance as possible. This is to allow a lightning stroke to the messenger to drain off by current flow to earth without causing the voltage of the messenger and sheath to rise excessively above the voltage of the cable conductors. If an aerial cable joins a n open-wire line, lightning arresters should be installed at the junction to protect the cable insulation against lightning surges which arrive over the open line. The ground terminal of these arresters should be connected directly to the cable messenger and sheath as well as to ground. Since the voltage and current surges produced in the messenger of aerial cable by a lightniug stroke to the messenger result in voltage and current surges in the cable conductors, it is generally recommended that aerial cahle be considered the same as open-wire lines as far as the protection of terminal equipment is concerned. PROTECTION OF A-C ROTATING MACHINES

Rotating machines present a special problem in lightning protertion. First the insulation of the stator windings of a-c rotating machines has a relatively low impulse strength. The highest test voltage it must with-


327

stand is simply the crest of the fi0-cycle high-potential test whose rms value is twice rated (line-to-line) voltage plus 1009 volts. This means that special effort must be made to limit the magnitude of the surge voltage which reaches the terminals of the machines. Secondly, the steep front of the voltage surge produced by lightning may damage the turn insulation even though the magnitude of the surge is limited t,o a value which can be safely withstood by the major (ronductor-to-ground) insulation. Such damage is avoided by reducing the steepness of the voltage wave which reaches the machine. Finally, as a result of the above limitations, lightning protertive equipment must be considered even though the machine is connected to the exposed overhead line through a transformer whose line side is adequately protected by a lightning arrester. A voltage surge of a magnitude and a steepness of front u,hich will damage machine insulation can be t,ransmitted through a transformer by electrostatic and electromagnetic coupling. The scheme of protection recommended differs somevhat for (1) machines connected direct,ly to exposed overhead lines and ( 2 ) machines connected to exposed overhead lines through transformers. Protection of Machines Connected Directly to Exposed Overhead lines. First to protect the turn insulation, the maximum rate of change

of voltage (steepness of wave front) applied to the machine must be reduced to a value which will limit the resultant turn-to-turn voltage to a safe value. This is accomplished by (1)connecting a protective capacitor between each line and ground in the path of the incoming surge (preferably a t the terminals of the machine) and ( 2 ) connecting a distributiontype arrester from line to ground a t a distance of 1500 to 2000 f t out on each directly connected exposed line. Then to ensure reliable protection of the major insulation, a station-type arrester should be connected in parallel with the protective capacitor. When located at the terminals of the rotating marhine, the ground terminals of both the arresters and capacitors should be connected directly to the machine frame, which of course should be connected to the plant ground bus. When the protective devices cannot be located directly at the terminals of the machine, it is preferable to bring the incoming lines to the terminals of these devices and then on to the machine, as shown in Fig. 5.26A4,rather than use separate leads from the machine terminals to the protective devices, as shown in Fig. 5.2CR. For additional improvement in the protection provided, the exposed lines should be shielded by overhead ground wires for a distance of approximately 2000 ft out from the plant, This reduces the possibility of direct strokes terminating on the circuits close to t,he station. It also ensures that a voltage surge originating on the line, beyond the arrester which is installed 1500 to 2000 f t from the station, will have its amplitude limited

328

SYSTEM 0VERVOLTAGES.-CAUSES


to the sparkover voltage and I R drop of this arrester regardless of the arrrst,er grouiid resistarice. The overhead ground mires should be directly ronnerted to the arrester grounds; they should he mel1 grounded a t each pole strurture; and they should be eoniiected to the plarit ground bus. l'he complete protective scheme is shown in Fig. 5.27. Where overhead groiind-wire shielding of lines does riot give effective protertion against direct strokes (hecause of inadequate line insulation

r3 J

n

L

3'I

INFERIOR METHOD

PREFERRED METHOD

(01

(AI

FIG. 5. 5 Allernaiive methodr of making connertions to rototing-machine protective equipment where thir equipment cannot be located directly a t the machine terminal%

t----

MACHINE

I

-

' 6 f+t

DISTRIWTION-TYPE ARRESTER

Ls

ARRESTER

c GROUND CDNNECTION

t

1500-2000 FT

I

-I

FIG. 5.27 Arrmgement of lightning protective equipment for nected directly 10 on exposed overheod lhe.

rototing mochine com


329

in relation t o the resistance of the pole structure grounds), an alternative method is availahle. This roiisists of installing a sperial ronrentrated inductanre of about 200 mirroheiirys, and of suitable ampere rating for the line rurrent, i l l earh phase b e t ~ e e i the i exposed lines and t,he parallelconnected statioir-type arresters and protective capacitors. 111 addition a distribution-type arrester should he installed on the line side of each ronreiitrated itidnctanre. Protection of Machines Connected to Exposed Overhead Lines through Transformers. In this case no arresters are reqnired out on the

line, hut instead a station-type arrester should be installed on the line side of the t,ransformer (see Fig. 5.28). In addition, to give the most reliable protection to both the major and the turn insulation of the rotating machine, a set of station-type arresters and protective caparitors should be installed between the transformer and the machine, preferably a t the machine terminals. The roniiertions t o this protective equipment should he as outlined for machines connected directly to exposed overhead lines. Protection of Machines Connected to Exposed Overhead lines through Reactors or Regulators. The protection provided in this case

should he the same as for marhines roiinected dirertly to exposed lines. I n addition a11 arrester should be applied dose t o the line terminals of the reactor or regulator, as sho\vn in Fig. 5.29, t o protert this equipment. Protection of Machines Connected Both Directly and through Transformers to Exposed Overhead lines. As shown in Fig. 5.30, the protection in this rase is provided by a combination of the arrangements described above; that. is, a station-type arrester should he provided on the line side of the t,ransformer, a distribution arrester 1500 t o ZOO0 ft out on the line t o which the machine is ronnerted dirertly, and a station arrester and protective capacitor at the machine terminals.

FIG. 5.28

Arrmgemenl of lightning protective equipment for nected to on e.,,ored overhead line through a transformer.

D

rotating machine con-


330

MACHINE

f7

REGULATOR O R

CONNECT ARRESTER AND CAPACITOR GROUND T E R M I N A L S T O MACHINE FRAME AND TO RELIABLE STATION GROUND


OVERHEAD GROUND WIRE

r-------7----ARRESTER *GROUND 'CONNECTION

-.ARRESTER CONNECTION

1500-2000F T

FIG. 5.29 Arrangement of lightning protective equipment for o rotating machine connected to an exposed overhead line through a voltage regulator or through a currentlimiting reactor. ARRESTER GROUNDCONNECTEDTO TRANSFORMER TANK OVERHEAD GROUND WIRE MACHINE DISTRIBUTION-TY PE ARRESTER

OONNECTYM TO MACHINE FRAME AND TO RELIABLE STATION GROUND

1500 -2000 FT

FIG. 5.30 Arrangement of lightning protective equipment for a rotating machine connected to exposed overhead lines both directly and through a transformer.

Effect of Cables and Switching Equipment between Rotating Machine and Exposed Overhead Lines. There will normally be one or more

lengths of cable as well as switching equipment between the rotating machine and the exposed overhead lines. For the lengths of cable normally encountered in industrial plants, this does not affect the application of rotating-machine protective equipment. However, the cable and switchgear should also be provided with adequate lightning protection (as previously described), and it may be found that some of the devices used can also serve in the rotating-marhine protective scheme. Protection of low-voltage Machines. As indicated by their GO-cycle high-potential tests, low-voltage machines (600 volts and helow) have relatively higher dielectric strength than the higher voltage machines. Where such machines are connected only to exposed overhead lines


331

t,hrough transformers which have adequat,e lightning protection on their primary, w additional lightning protective equipment is generally warranted. However, where Ion-voltage machines are supplied direct,ly from exposed overhead lines, lightning prot,ection shonld he provided. l y p i r a l installations of t,his t,ype include motors installed in oil fields and in quarries. Protection of Machines Having Single-turn Coils. Rotating machines above a certain size (lower limit varies wit,h speed and voltage rating) are generally built with single-turn coils in which the coil insulation also serves as the turn insulation. For 3GOO-rpm turbine generators, typical lower limit,s are 2500 kw a t 2400 v o k , 5000 k w at 4160 volts, and -10,000 k w a t 13,800 volts. For such machines, protective raparitors are essential only if they are required t o limit the rise of volhage at the neutral of the machine due t o positive reflection of the surge voltage wave a t t,his point, Hence t,hey are not required if the machines (having single-turn coils) are connected t o the exposed overhead lines through delta-Y or Y-delta transformers. Xeither are they essential if the machine neutral is grounded t,hrough a neutral resistor of 25 ohms or less or through a neutral reactor of 0.1 ohm or less (on a 60-cycle basis). Protection of Two or M o r e Machines on the Same Bus. The installation of station-type lightning arrest,ers and p r o t e h v e raparitors at t,he terminals of each rotating machinc is always an ideal arrangement as far as the protection provided is concerned. However, where two or more machines in a plant are connected t,o the same exposed overhead lines, there are obvious economies in plaring t,he protective equipment, on a rommon bus or a t some other point where it will be in the path of the lightning surges t o all machines. Where there are a number of machines involved, a set of protective equipment on each incoming line may he the most economical arrangement,. In such inst,allations, if the protective equipment is not over 500 ft from the rotating machines and is placed directly in the path of t,he incoming surges, there is relatively little loss in protection. A compromise arrangement places the protective capacitors a t the terminals of each machine with a single set of arresters at the common point. If the protective equipment is not located at the machine terminals, careful grounding of the arrester and capacitor sod a n interconnection between this ground and the marhine frame is quite important. If the circuit between the protective equipment and the machine frame consists of continuous metallic-sheath cable or the cable is run in metallic conduit, the arrester and capacitor ground terminals should also be connected t o the cable sheath (or conduit) and the latter should be joined t o the machine frame.

332


Ratings of Protective Equipment Recommended. Table 5.9 shows the ratings of the lightning arresters and protective capacitors recommended for protection of three-phase rotating machines of the popular voltage ratings. The ratings of the arresters required on the line side of any transformer between the rotating machine and an exposed overhead line are not shown. Such arresters must he selected to match the voltage and method of grounding of the line-side system. For single-phase machines the same recommendations apply except that only two single-pole units are required if neither line is grounded and only one (on the ungrounded line) if one line is grounded. However, for TABLE 5.9

Protective Equipment for Three-phase A-C Rotating Machines For instoilofion IS00 to 2000 ft out on directly connected exposed overhead lines

For indallation mt machine tsrminoli or on mochine bus

Protective capacitors Mochine

- -

st.tion-type

1

orrester,

Distribution-type orrc~tcrs

~

~

*olt.ge rating

Voltage rating

Iphose-to phase1

Voltage rating

~

Voltw e di"(

MiUO

forad per pole

~

Single p01e ""it. re.

Effective1 groundec system

quire.

i"gleUnpole units grovnde< iff&i"*l: or regroundoc "ired rnidonca system ground=<

Singlepole ""it. required

system

- 0-650 2,400 4.160 4.800 6,900 11.500 13.800

0-65 2.40 4.16 4.80 6.90 11.50 13.80

.o

3'

0.5 0.5 0.5 0.5 0.25 0.25

3* 3*

I

I-

~

3

3 3 or 6 3 or 6

650 3,000 4,500 6,000 7,500 12.000 15.000

~

650 3,000

650 3,000

3,000$ 4,500 6,000 9,000

I2.000

3 3

3 3 3

6,000 6,000 9,000 12,000 15,000

3,0001 6,000 6,000 9,000 12,000

3 3 3 3 3 3

3

--

-- - - - * A single three-pole

unit is commonly used. units (0.5 pf per phase) where both of the following conditions apply: (1) Maehinc IS directly connected to t h e exposrd overhead lines. is connected through a n autotransformer. or is eonnectcd throneh a Y-Y transformer with both Y's grounded. (2) Machine is ungrounded, is neutral grounded through a resistance greater than 50 ohms, or is neutral grounded through a reactance greater than 5 ohms (60-cycle basis). In all other cases three capacitor units (0.25 pf per phase) will suffice. $The use of 3000-volt arresters on a 4160-volt system requires s n X o / X Iratio less than that newssary to make the system "effectively grounded." (See Selection of Arrester Voltage Rating.)

t Use six capacitor

~~

~~0~~~~

~

~~~


333

CkG5O-volt marhines a t,hree-pole protertive caparit.or is commonly used in any rase. T\\-o-pole 2400-volt capacitors and two-pole GO-volt arresters are available for use on single-phase systems i n which neither line is grounded. As shown, the voltage rating of the protective capacitors rerommended matches the system phase-to-phase vokage for both effectively grounded and ungrounded systems. They arc generally designed so that t,hey rail be used on any marhine whose voltage rating does not exreed 110 per rent of the capacitor voltage rating. Where the voltage rat,ing of a marhine falls between the voltage ratings of the prot,ertive raparitors available, one of the next higher voltage rating r a i l alivays be used. PROTECTION OF D-C ROTATING MACHINES AND RECTIFIERS

D-C motors and generators connected to exposed overhead lines should be proterted hy suit,ahle d-r arresters, snrh as the capacitor type. They may he installed at the machine terminals, on the bus, or at the station on each outgoing feeder. Mercury-arr rectifiers and t,heir transformers (Fig. 5.31) may be protected hy a set of station-type or distribution-type arresters on t,he supply side of the transformer and, if the d-r feeders are exposed, suitable d-r arresters a t the d-c terminals of t,he rertifier, on the d-r hus, or on the exposed d-c feeders. In addition t o this protection, rectifier transformers are often supplied with built-in nonlinear resistors or “surge eliminators” installed on one of the secondary Y’a or zigzags, and also installed across AC SUI

1

RECTIFIER TRANSFORMER

ANODES MERCURY CATHODE

I

~

+

DC

T

I

I/

4

THYRITE SURGE ELIMINATORS (BUILT IN AS INHERENT PART ff GENERAL ELECTRIC RECTIFIER TRANSFORMERS I

I L L

FIG. 5.31

4 Typical scheme of lightning prokction for a mercury-arc rectifier.

334


the interphase transformer winding, to absorb the high peaked surges of small energy associated with mercury-arc phenomena. PROTECTION OF COMPLETE POWER SYSTEM

The protection of a complete power system is accomplished simply by providing adequate protection to each component of the system by the methods which have been described. In some cases it will be found that the protective equipment required for one piece of apparatus will also serve to protect other pieces. The effect of separation between arresters and protected equipment must, however, be considered. Typical arrangements of protective equipment for an industrial power system are illustrated in the one-line d i a g a m of Fig. 5.32. As shown,

FIG. 5.32 ryrtern.

Typical arrangement of lightning protective equipment on on industrial power

SYSTEM OVERVOLTAGES-CAUSES AND PROTECTIVE MEMURES

335

lightning arresters are provided on the high-voltage incoming line. It is assumed that the distance between these arresters and the transformer terminals is short enough so that no additional protection is required for the transformer. The generator and large motor are provided with both lightning arresters and capacitors a t their terminals, while the smaller motor has the capacitors only. Lightning arresters are also provided on the line side of the breakers to which the overhead feeders are connected. Arresters are also required a t the junction of the overhead line and cable, but it is assumed that the length of this cable is such that these arresters do not give adequate protection to the switchgear. Finally, arresters are installed a t a distance of 1500 to 2000 f t nut on each overhead feeder to complete the protection required for the rotating machines. In all cases the type and voltage rating of the arresters would be selected as outlined under Application Procedure. No attempt has been made to show the details of the connections to the arresters. These should be made in accordance vith the recommended practices that have been outlined. REFERENCES 1. Modern Conwpts of Lightning ProtPetion for Transmission and Distribution Circuits, Ocneral Electric Company Publication GET-I720A, 1948. 2. Lightning Protective Equipment for Rotating Machines, Gerieral Electric Company Puhlication GEA-l743H, 1953. 3. Shott. €1. S.. and H. H. Peterson, Critoris. for Xeutral Stability of Wye-Grounded Primary Ijroken Delta Secondary Transformw Circuits, Tmns. A I E E , vol. 60, November, 1941. 4. Blrmie, L. F., and A. Bwajian, "Transformer Engineering," 2d ed., John Wiley & Sons, Ine.. New York, 1951. 5. Schroeder. T. W.,The Cause and Control of Somr Typcs of Switching Surges, Tions. A I E E , vol. 6 2 , November, 1943. 6. AIEH Committee Rcport, Poner Systmn Overvoltages Produccd by Faults and Switching Operations, 1948. 7. AIEE Committee Rcport, Corrdation of System Overvoltagcs and Pystem Grounding Impedance, 1943. 8. Lewis, W. W.."The Protection of Transmission Systems Against Lightning," John UIIPy & Sons, Inc., New l-ork, 1950. 9. Brwley. L. V., "Traveling Waves on Transmission Systems," 2d ed., John WilQy & Sons, Inc., New York. 1951. 10. Joint Cornmittcc on Coordination ai Insulation of AIEE, E E I , and KEMA, Standard Basic lmpulsc Insulation l,evels, EEI Publication No. H-8, NEMA Publication No. 109, 1941. 11. National Clcetrieal Manufacturers Association, Standards of Lightning Arresters, Publiention Kos. LAl-1852 to I,A5-1952, 1952. 12. "American Standard for Lightning Arrestprs for Alternating Current Power Circuits," ASA Standard C62.1, 1944. 13. Hunter, E. M.,E. Pragst, and P. H. Light, Dctermination of Ground-fault Current and Voltages on Transmission Systrrns, Cen. Elec. Rev., August and November. 1939.

336


14. Wagner, C. F., G. D. McCann, and C. M. Lear, Shielding of Substations, T ~ a n s . A I E E , vol. 61. February, 1942.

15. Boehne, E. W., Voltage Oscillations in Armature Windings under Lightning Impulses, Trans. A I E E , vol. 49, 1930. 16. Rudge, W. J., R. W. Wiesernan. and W. M. Lewis, Protection of Rotating A-C Maehinps Against Traveling Wave Voltage Due to Lightning, Tians. AZEE, vol. 52, 1933. 17. Hunter, E. M., and N. E. Dillow, Surge Protection of Rotating Machines, Gen. Eke. Rev., May, 1950. 18. General Electric Application Committee, Lightning Protection of Metalclad Switchgear and Unit Substations Connected to Overhead Lines, Gen. Elec. Re"., March, 1949. 19. Towne, H. M., Lightning Protection of Substations. supplement to Dist7ibulion Magdzine, July, 1951. 20. Allen, E. J., Protecting Meter Equipment from Lightning, supplemmt to Distribulion Magazine, July. 1952. 21. Rudge, W. J., W. A. MeMorris, S. B. Howard, and T. J. Carpenter, The New Thyrite Mrtgne-Valve Ststion Arrester, AIEE Conference Paper, 1954.

Chapter 6

by

L.

J. Carpenter and L. G. Levoy, Jr.

System Grounding' About midafternooil one day i l l a West Coast manufarturi~igplant, normal operations herame suddenly disrupted. The first evidenre of trouble came in the form of a motor failure 011 the 480-volt system, then another, and still anot,her in close succession. h i i inspertion of switchboard voltmeters (measuring line-to-line volts) and ammeters indicated no unusual conditions. System equipment continued to fail. A test voltmeter W:LS rigged up having a full-srale ralihratioti of 1200 volts. Upon coiiiiecting it phase-to-ground, the pointer went o f srale. A phaseto-ground potential on a 480-volt system of more t,haii 1200 v o l h existed! At once the inroming service t,raiisformers were snsperted of iiiternal breakdown hetween high- and Ion-voltage windings, As the last of these transformers was isolated and individually tested, it hemme evident that they mere not at fault. System equipment rontiiiued t o fail, and the situation was desperate. A frantic group went into a huddle aiid derided that, t,he only way out, was t o trip the main inroming service breaker n-hich would deenergize the entire system. A t this point one of the workmeii noticed a small wisp of smoke coming from a motor-starting autotransformer and, upon approaching, could hear a buzzing noise inside. This circuit was switched rlear of the system, and the overvoltages disappeared. During the two-hour period that this arcing fault existed, hetween 40 and 50 motor windiiigs had failed. Finally it was found that the autotransformer enclosing case had been bashed in and was practically in contact with the coil. The spot where arcing had taken place was evident although not badly burned. An attempt mas made t o show the plant engineer what had been the trouble. A solid connection was made between the frame aiid the burned

* Crrdit for much of the original analytical work on this srhjcat is duc to IV. I i . Roire, who was formerly a mcmher of the Industrial EnginPcring Srction. Orrwral Electric Company. 337

338

SYSTEM GROUNDING

spot on the coil. Much to the bewilderment, of the operating men and according to the expectations of the plant engineer, no mnre than the 73 per cent increase in the voltage to ground on the other two phases occurred. The main ingredient of the overvoltage (discontinuous conduction) had been omitted. This is an actual case of severe prolonged experience of overvoltage of repetitive restrike origin on a 280-volt ungrounded system. This story is spectacular because of the magnitude of the disturbance and consequential damage. Similar occurrences of lesser extent are not uncommon, however, and there is evidence that they are more frequent than realized. It is a characteristic of ungrounded systems that they are subject to relatively severe transient overvoltages. This trouble can be avoided by proper grounding of the system, and other importaut benefits are also obtained. For a detailed explanation of the nature and causes of these overvoltages, refer to Chap. 5 . System grounding has been practiced since the beginning of electric power systems. This method of operation has not been universally uniform even within a given voltage class of systems or between various operating companies. On the other hand, rertain systems are nearly always grounded, for example, 120-volt lighting circuits. The problem of whether or not a system neutral should be grounded, and how it should be grounded, has sometimes not had the complete understanding and engineering analysis which it deserves. As a consequence, the grounding of many systems has heen hased on past experience or opinion, and therefore system grounding practice is found to vary widely on existing systems. On the other hand, most new systems conform to modern grounding practices. A comprehensive review of the problems involved in grounding the neutrals of industrial power systems clearly shows that it is generally advantageous to ground all power-system neutrals regardless of voltage or of process in the plant. The application in practical systems, however, must be tempered by the availability of standard a p p a r a h s for new systems and the equipment and practices in an existing plant. DEFINITIONS

The word “grounding” is commonly used in electric power system work to cover both “system grounding” and “equipment grounding.” To avoid confusion or possible misunderstanding, this chapter is devoted exclusively to the subject of system grounding. The following chapter (7)is devoted to equipment grounding. These terms are defined by the National Electrical Code as follows:

SYSTEM GROUNDING

339

System Ground: A system ground is a connection to ground from one of the current-carrying conductors of a distribut,ion system or of an interior wiring system. Equipment Ground: An equipment ground is a ronnertion to ground from one or more of the non-current-carrying metal parts of the wiring system or of apparatus connected to the syst,em. As used in this sense, the term equipment includes all such metal parts as met,al conduit,s, metal raceway, metal armor of cables, outlet boxes, cabinets, switch boxes, motor frames, and metal enclosures of motor controllers. The following definitions are taken from AIEE Standard S o . 32. Seutral Grounding Devices. System Neutral Ground: A system neutral ground is a connection to ground from the neutral point or points of a rircuit, transformer, rotat,ing machine, or system. The neutral point of a syst,em is that point whirh has the same potential as the point, of junction of a, group of equal nonreactive resistances if connected at their free ends t o the appropriate main terminals or lines of the system. (Except where specifically stated to be otherwise, the srope of this chapter includes and relates t o neutral grounding of three-phase a-r systems in industrial plants.) Grounded Sydern: A grouuded system is a system of conductors i n which at least one conductor or point (usually the middle wire or neutral point of transformer or generator windings) is intentionally grounded, either solidly or through a current-limiting device. KOTE:Grounded systems may be subject to various steady-stat,e and transient overvoltages depending upon the ratios of X o / X , and R o / X , as viewed from the fault location. Elo, X o , and X,are, respectively, the zero-sequence resistance, the zero-sequence reactanre, taken as positive if inductive and negative if capacitive, and positive-sequence subtransient reactance. Ungrounded: Ungrounded means without an intentioual connection to ground except through potential-indicating or measuring devices. Solidly Grounded (Directly Grounded) : Solidly grounded means grounded through an adequate ground connection in whirh no imcedance has been inserted intentionally. Resistance Grounded: Resistance grounded means grounded through impedance, the principal element of which is resistance. Reactance Grounded: Reactance grounded means grounded through impedance, the principal element of which is reactance. Resonant Grounded (Tuned Grounded) : Resonant grounded means reactance grounded through such values of reactance that, during a fault between one of the conductors and earth, the rated-frequency current

340

SYSTEM GROUNDING

flowing in the grounding reactanres and the rated-frequency capacitance iwrrent flowing between the unfaulted condurtors and earth shall he substantially equal. In the fault, these two components of the fault current will be substantially 180’ out of phase. Ground-fault Neutralizer: A ground-fault ueutraliaer is a grounding device which provides an inductive component of current in a ground fault that is substantially equal t o and therefore neutralizes the ratedfrequency capacitive component of the ground-fault current, thus rendering the system resonant grounded. Grounding Transformer: A grounding transformer is a transformer intended primarily t o provide a neutral point for grounding purposes.

CHARACTERISTICS

OF UNGROUNDED SYSTEMS

The term ungrounded system is used t o identify a system in which there is no intentional connection betneen the system condurtors aud ground. However, in any practicalsystem, therealmaysesists a rapacitive coupling between the system conductors aud ground. Consequently, the so-called “ungrounded system” is in reality a “capacitively grounded” system by virtue of the distributed rapacitanre from the system rondurtors t o ground. When the neutral of a system is not grounded, it is possible for destructive transient overvoltages, of several times normal, t o appear from h i e to ground during normal switrhing of a circuit having a line-to-ground fault. Tests have shown that overvoltages may be developed Sy repeated restriking of the arr during interruption of a line-to-ground fault, particularly in lom-voltage systems. Experieiice has proved that these orerUNFAULTED CIRCUIT UNGROUNDED POWER SOURE

-

4

)TRANSIENT

fI

1 4,

I MAY CAUSE

I

SECOND FAULT

I HERE

-

clb-

POTENTIAL

, ~

BREAKER FIG. 6.1

OVERVOLTAG ES

+-

t

-SINGLELINE- TO GROUND FAULT

INTERRUPTING FAULT

Transient overvoltages due to ground-fault interruption on ungrounded system

may cause other faults to occur on system.

SYSTEM GROUNDING

341

voltages may cause failure of itisulat,ioii at other lo(.atiotis on the system than t,hc point of fault,. Thus, a litie-t,o-ground fault 011 one circuit may result ill damage t o eiluipmeiit aud interruption of service on other rirruit,s. The same condition will result from the repeated restrike of the arc: in at1 arcing fault from line t,o ground. The condit,ion described is illustrated ill Fig. (i.1. In aii uiigrouiided-neutral system, a serond ground fault 011 another phase may occur heforc the first fault, is removed. The second fault may he on the same cirruit as t,he original fault or OII another. In any event,, the resulting line-to-line fault will avtuate relays or circuit breakers and

i SECChD ' I GROUNC FAULT

FIG. 6.2

Double line-to-ground faults on ungrounded system result in outages of two circuits and high-level fault currents which can cause severe damage to equipment.

FIG. 6.3

One ground foult on

an ungrounded ryrtem may cause ground foultr in other

connected apparatus.

-:;i:::VOLTAGE

FU LL LIN E -T O LINE VOLTAGE NEUTRAL NORMAL GROUND VOLTAGE

*

GROUND P O T E NT I A L

Y

(A) NO FAULT ON SYSTEM (N E U TR A L FL O A T S AT GROUND POTENTIAL)

FIG. 6.4

Effect on line-to-ground voltages of

grounded neutral system.

(01 SINGLE-LINE-TOGROUND FAULT ON SYSTEM (ONE LiNE AT GROUND POTENTIAL1 (I

single line-to-ground fault on an un-

SYSTEM GROUNDING

343

cent higher than normal. Figure G.4 illustrates the increase i n line-toground voltage due t o a ground fault. Usually the insulation het,ween each line and ground is adequate t,o withstand full line-to-line voltage. However, if this voltage is applied for loug periods, it may result in failure of insulation which may have deteriorated hecausc of age or severe service conditions. Line-to-ground faults on ungrounded-iieutral systems muse a very small ground-fault current t o flow through the raparkanre of cahles, transformers, and other electrical equipment ou t h e system. This current may have a magnitude from a few amperes t o 25 amp or more 011 larger

rLA J

I

1 T

T

1Y

1Y

- ~.

I

h

s

2Y

Y

s5

T

3 "

5

P

FIG. 6.5

Location of ground faults m a y be troublesome on ungrounded neutral systems.

344

SYSTEM GROUNDING

systems. This is not, in general, enough to actuate protective devices, bnt it may do considerable damage if allowed to flow for a long period. Ground detectors on an ungrounded-neutral system will indicate the existence of a ground fault but will not give its location (see Fig. 6.5). Several dt vices are available for determining the approximate location of ground faults. Such devices are admittedly helpful, hut they do not provide the complete answer. Some time is still required to locate and remove the faulty feeder from service for repair. These devices do nothing to prevent the occurrence of the fault.

SWITCHES

hll. G K g

FIG. 6.6 Ungrounded low-voltage system with single line-to-ground fault in one circuit.

SYSTEM GROUNDING

345

The problem of locating a fault on an ungroiuided-iieutral system is illustrated in Fig. R.G. While it is easy to see where the fault is iii the diagram, it is not easy to locate it i n the artual ungrounded system. The first step is to opeti the secondary feeders one at a time. This will tell on whirh feeder the fault is. After finding vhirh feeder the fault is on, then the branch cirruit,s are opeired one at, a time and finally the motors and loads taken off one at a time. If this is done during produrtion hours, it rail readily he seen how much production loss there may he just t o find a ground fault iir an ungrounded system. This is contrasted with a grounded-neutral system where only t,he motor A , Fig. 6.6, would have been tripped out and no other produrtion marhines iuterfered with. A second ground may occur on the same phase, but at a different location than the first. This is more diffirult t o find bemuse the operator must then open all circuits a t once and dose them one a t a time t o find the ground fault. Often it is argued that with an ungrounded system one ground fault cau be left on the system uutil it is convenient to locate it without interfering with production. Experienre has shoivn that double ground faults are rery common in ungrounded systems simply because the first ground is left on, hoping that the operator will find it before the serond ground fault occurs. ADVANTAGES OF SYSTEM NEUTRAL GROUNDING

The advantages of operating an industrial power system grounded compared with operatiug it ungrounded may be one or more of the following : 1. Reduced operating and maintenance expense a. Reduction in magnitude of transient overvoltages b. Improved lightning protection c. Simplification of ground-fault location d. Improved system and equipment fault protection 2. Improved servire reliability 3. Greater safety for personnel and equipment The relative weight of these advantages varies with system voltage classes and t o a lesser degree with installation conditions. Wheii the system is grounded at the neutral by a low value of impedance, grounded-neutral lightning arresters may be used which give hetter lightning protection, other things being equal, than do ungroundedneutral arresters required for ungrounded-neutral systems or for groundedneutral systems which are grounded through a relatively high neutral impedance. I n general, circuits below 15 kv are not exposed to lightning within the industrial plant, so that the advantage gained from better

346

SYSTEM GROUNDING

light,ning protection through the use of
347

SYSTEM GROUNDING

277-volt fluorescent lighting without lighting transformers. I n each case, the neutral is solidly grounded. It has been the experience of operators who have used both groundedand ungrounded-neutral systems that the failure rate is substantially lower and the time the system is out of service is less on the grounded system. This results from the fact that transient overvoltages are greatly reduced on a grounded-neutral system. Because grounding reduces these overvoltages, the life of electric insulation will he increased and service interruptions will be minimized. Even though the overvoltages of a n ungrounded-neutral system may not he high enough to cause multiple failures, every time a ground fault occurs, the repeated application of these overvoltages will weaken the insulation and cause a higher failure rate than in a grounded-neutral system. TABLE 6.1

Summq

f

of Advantages of the Grounded-neutral 480-volt System Grounded-neutral system

Safety.

... . .. . . ...

Service rdiobility.

.. Fin1 coil. . . .. ......

Maintenonce cost..

High-vollage Rvoresce lighting

I

Ungrounded system

SAFEST-Only 277 voits to ground Normolly 277 volts to ground when at m y time (assume good ground no around on sydem. 480 vdti 10 and 480 volts maximum line to line1 ground on two conductors when one phore i s grounded SAFEST-Voltage on system limited Voltage on recondory system may be to obout 277 volts when primary to as high as primary voltage for secondary failure OCCUR in Ironsbreakdown between primary and rocondory Ironsfarmer windings former supplying system SAFEST-Ground fault in ~ontroi Control circuit ground fault likely to wiring cmn put only 58 per ~ e n line f put full v ~ l t o g eon cantactor closing volloge on line-to-line connected coils LO"t.dO, closing coil. HIGHEST-Ground faults ore mod- Port or 011 of system must be taken out of service to Rnd ground faults ily located and repaired; syitem need not be taken out to Rnd Subject to severe transient overvoltground faults age* HIGHEST-Ground foults arc locd- Ground faulh if not removed may upon occurrence of a second ired and trip off immediately ground foult cause N o circuits to HIGHEST-h%nimizes Irondent overgoout atonce, thus causing 0 lossof voltages on the system twice (IS much production equipment HIGHTEST-Flooting grounds are Floating or arcing grounds likely very unlikely LOWEST-Ground faults arc easily Time must be spent hunting ground located fault, delta-connected About same (IS substation and ground detector Provides 277 volts for direct opera- Mud use step-down handormen lion of fluorescent lights. resulting from 480 to 277 volts or lower in a cod roving by lhe elimination of lighting Ironsformers and a reduction in copper

348

SYSTEM GROUNDING

A summary of the advantages of grounded-neutral operation on a 480volt system is given in Table 6.1. In general, the same advantages are applicable to other system voltages of 600 or lower. A summary of the advantages of grounded-neutral operation on systems of 2.4 to 15 kv is given in Table 6.2. TABLE 6.2

Summary of Advantages of the Grounded-neutral System (2.4 to 15 Kv) Grounded-neutral system

..... ....... Somice roiiobllity . .... Safely.

... First C o l t . . . .... ..... Maintenance cost..

I

Ungrounded system

SAFEST: Single line-to-ground faulh Subject lo severe troniient overare tripped off immediocly voltages HIGHEST: Ground faults are readily Part or 011 of system must be taken out of service to find fovltr located and repaired Ground foulh, i f not removed, may HIGHEST:Limited fault current C(IUSFI upon occurren~eof (I second ground a minimum of damage to equipment [with conventional resistonce fault C W S ~ two circuiti to go out at once, lhur couiing the 1 0 s of Wice grounding) as much production equipment HIGHEST: Minimizer transient over- High fault current assdated with two line-to-ground, faulh moy result in voltages on lhe system more damage to equipment LOWEST: Ground fauih ore easily Ground faulh ore more diticult lo locote locoled ABOUT SAME: Adds cost of resistor Requires ground-detector and faultlocator equipmentto be comparable and nwtral relaying

HOW TO OBTAIN THE SYSTEM NEUTRAL The best way to obtain the system neutral is to use source transformers or generators with Y-connected windings. The neutral is then readily available. Such transformers are readily available for practically all voltages. On new systems, 208Y/120 or 480Y/277 volts may be used to good advantage instead of 240 volts. For 2400- or 4800-volt systems, special 2400Y- or 4800Y-connected source transformers may be purchased or grounding transformers used.

GROUNDING TRANSFORMERS System neutrals may not be available, particularly in many old systems 600 volts or less and many existing 2400-, 48On-, and 6900-volt systems. When it is desired to ground existing delta-connected low-voltage systems (0-600 volts), grounding transformers are used to form a neutral which is then connected solidly to ground. I n like manner, 2.4- t o 15-kv

SYSTEM GROUNDING

34P

systems having only delta-connected equipment may he grounded by adding grounding transformers and neutral resistors. Grounding transformers may be either of the zigzag or Y-delta type. ZIGZAG GROUNDING TRANSFORMERS

The t,ype of gromrding transformer most rommonly used is a threephase zigzag transformer with no secoudary winding. The internal cotinection of t,his trausformer is illustrated in Fig. G.7. The impedance of the transformer to three-phase currents is high so that, when t,here is no fault on the system, only a small magnetizing current flo~vsin the transformer windings. The transformer impedance t o ground current,, h o w ever, is low so that it, allows high ground currents to floxv. The transformer divides the ground current iuto t,hree equal componeuts; these currents are in phase with each other and flow i n the three windings of the grounding t,ransformer. The met,hod of wiuding is seen, from Fig. 6.7, t o he such that when these t,hree equal currents flow the current i n one section of the winding of each leg of the core is i n a direction opposite to that in the other sertion flus \\-hich results from of the n$iding on that lea. The only magnetic the zero-sequence ground rurrents is the leal e field about earh XI-inding I

~

>

L I N E LEADS

1 t-

_ . )

1 1 7 1

I -

-

NEUTRAL LEAD

(0) WINDING SHOWN ON CORE

(b) SCHEMATIC OF

FIG. 6.7

DIAGRAM CONNECTIONS

Zigmg three-phore grounding transformer.

350

SYSTEM GROUNDING

section. This accounts for the low impedanre of the transformer to ground current. The short-time kva rating of a grounding transformer is equal t o rated line-to-neut,ral vokage times rated neutral current. h grounding transformer is designed to carry its rated current for a limited time only, such as 10 sec or 1 min. Hence, it is normally about one4enth as large, physically, as an ordinary three-phase transformer for the same rat,ed kva. Y-DELTA GROUNDING TRANSFORMERS

A Y-delta transformer ran also be utilined as a grounding transformer. I n this case the delta must he closed t o provide a path for the aerosequenre current, but the delta can be made up at any ronvenient voltage level. It may or may not be used to serve other loads. The Y nindiug must be of the same voltage rating as the circuit mhirh is to he grounded. The connections of the transformer are shown in Fig. G.8. APPLICATION

A grounding transformer should he connected t o the system in such manner that the system mill always he grounded. Figure G.9a shows a grounding transformer with an individual line breaker for connection directly to a main bus of the system. Figure G.9b shows a means of COIInecting a grounding transformer t,o a system without an individual line breaker. In this case, the grounding transformer is connected between the main t,ransformer bank and its hreaker. If grounding t,ransformers are connected as shown i n Fig. G.Yb, there should be one grounding transformer for each delta-ronnerted bank supplying power t o the system, or enough t o assure a t least one grounding transformer on the system at all times.

FIG. 6.8 Connections and current dirtribution in CI Y-delta grounding tranrformer when line-to-ground foult occurs on a three-phare system.

351

SYSTEM GROUNDING

A

A

A

Y

-

GROUNDING TRANSFORMER

-

GROUNDING R E S l STOR

i

{Jq3GROUNDlNG T R A N SFORMER

L

~

$

FIG. 6.9

GROUNDING RESISTOR

(b)

Methods of connecting grounding transformer to system.

SYSTEM GROUNDING

352

In applying grounding transformers, the first step is to review the system voltage and fault current level to determine whether the system should be grounded solidly or through a resistor. METHODS OF NEUTRAL GROUNDING

In grounding the neutral of a power system, the advantages outlined will be achieved provided that proper attention is given to the impedance of the circuit from system neutral t o ground. This circuit is illustrated in Fig. 6.10 for the commonly used grounding methods. These methods are referred to as solid grounding, rpsistance grounding, reactance grounding, and ground-fault-neutralizer grounding. Kote that each method is named in accordance with the nature of the external circuit from system neutral to ground. In each case the impedance of the generator, or transformer, whose neutral is grounded is in series with the external circuit. Characteristics of the various methods of system neutral groundiug are given in the following text and summarized in Table 6.3. Application limits and guides for the various methods are outlined with reference to the following: 1. Effect on development of transient overvoltages 2. Damage a t the point of fault due to magnitude of ground-fault current 3. Application of standard relays and circuit-interrupting devices for selective ground-fault tripping 4. Lightning protection SOLID GROUNDING

A power system is solidly grounded when a generator, power transformer, or grounding transformer neutral is connected directly to the station ground or to the earth, as shown in Fig. 6.11. Because of the reactance of the grounded generator or transformer in series with the neutral circuit, solid grounding cannot be considered a zero-impedance circuit. If the reactance of the generator or transformer is too great, the objectives sought in grounding, principally freedom from transient overvoltages, will not be achieved. Thus, it is necessary to determine how solidly the system is grounded A good guide in answering this question is the magnitude of ground-fault current as compared with the system threephase fault current. The higher the ground-fault current in relation to the three-phase current, the more solidly is the system grounded. For nearly all solidly grounded systems (also reactance-grounded sys-

SYSTEM GROUNDING

353

tems) it ia neecusary for the ground-fault current to he in the range of 25 t o 100 per cent of the three-phase fault rurrent to prevent the development of high transient overvoltages. This may mean symmetrical rms ground-fault currents in the order of 10,000 t o 40,000 amp. CIRCUIT

0 I

EOUIVALANT DIAGRAM

5?

UNGROUNDED

Y I

3.

RESISTANCE GROUNDED

5

GROUND F A U L T NEUTRALIZER

XG-REACTANCE

cp

OF GENERATOR OR TRANSFORMER USED FOR GROUNDING

XN-REACTANCE O F GROUNDING REACTOR RN-RESISTANCE OF GROUNDING RESISTOR

FIG. 6.10

System neutiol circuits and methods of grounding.

SYSTEM GROUNDING

SYSTEM GROUNDING

355

Direct grounding of a generator without external impedanre may cause the grouiu-fault current from the generator t o exceed the masimum three-phase fault rurrmt which the generator can deliver aud t,o e x r e d the short-rirruit rurreiit for which its vindings usually are hraced. Consequently, i n rases where solid grounding of a system is indirated, generators should be grounded through a react,or having a l o w ohmic value whirh d l limit fault current t o a value no greater than three-phase fault rurrent. 111 the case of three-phase four-wire systems, limitation of ground-fault current t o 100 per cent of the three-phase fault current, is usually practical n-it,hnut interfering with normal four-wire operation. Lightning arresters for grounded-neutral systems may be applied when the system is grounded through a low impedance to prevent displacement, of t,he system neut,ral d h respert t o ground beyond specified limits. In this rase, the maximum impedanre may he espresscd in terms of minimum ground-fault current. This cnrrent should be at least 60 per cent of the three-phase short-circuit current for appliration of grounded-neut,ral-type lightning arresters. I

GENERATOR SOLIDLY GROUNDED

FIG. 6.1 1

POWER TRANSFORMER SOLIDLY GROUNDED

T

GROUNDING TRANSFORMER SOLIDLY 7

Methods of solidly grounding the neutral of three-phase systems.

RESISTANCE GROUNDING

In resistance grounding, the neutral is connected t o ground through one or more resistors, as shown in Fig. 6.12. In this method, with resistor ohmic values normally used, the line-to-

356

SYSTEM GROUNDING

ground voltages which exist during a line-to-ground fault are nearly the same as for an ungrounded system (except transient overvoltages). This is illustrated in Fig. 6.13. A system properly grounded by resistance is not subject to destructive transient overvoltages. For resistance-grounded systems at 15 kv and

I

$9 Y

POWER TRANSFORMER NEUTRAL

GENERATOR NEUTRAL

FIG. 6.12

-

Methods of resistance grounding the neutral of three-phase systems.

NORMAL LINE-TO-, NEUTRAL

+' A) SYSTEM NEUTRAL UNGROUNDED

FIG. 6.1 3

GROUNDING TRANSFORMER NEUTRAL

-

POTENTIAL

+!

VOLTAGE DROP IN GROUNDED PHASE,DUE TO GROUND CURRENT

8) SYSTEM GROUNDED BY RESISTOR

System voltage diagrams during single line-to-ground faults.

a t operating frequency-transient voltages not shown.)

(All voltage5

SYSTEM GROUNDING

357

below, such overvoltages will not ordinarily be of a serious nature unless the resistance is so high as to limit the ground-fault current t o a small fraction of 1 per cent of the system three-phase fault current (i.e., to less than the system charging current). This much ground current (usually well below 50 amp) is far less than is uormally used with resistor grounding. Systems grounded Ghrough resistors as described in this chapter should use lightning arresters for ungrounded-neutral circuits, where lightning arresters are required. The reasons for limiting the current by resistance neutral grounding are as follows: 1. T o reduce burning aud melting effects in faulted electric equipment such as switchgear, cables, and rotating machines 2. To reduce mechanical stresses in circuits and apparatus carrying fault current 3. T o reduce electric shock hazards t o personnel, caused by stray ground-fault currents in the ground return path 4. To reduce the momentary line-voltage dip occasioned by occurrence and cleariug of a ground fault REACTANCE GROUNDING

The term reactance grounding describes the case in which a reactor is connected between the machine neutral and ground, as shown in Fig. 6.10. The magnitude of reactance in the neutral circuit determines how “solidly” the system is grounded and therefore what its characteristics will be. Since the ground-fault current which may flow in a reactancegrounded system is a function of the neutral reactance, the magnitude of ground-fault current is often used as the criterion for the various system Characteristics rather than referring to neutral reactance directly. In practice, reactance grounding is generally used only in the case cited under Solid Grounding, in which a generator neutral is to be connected directly to ground. In this event, it may be necessary t o add a low-value reactor to limit the available ground-fault current through the generator to a value no greater than the three-phase fault current contributed by the generator. The characteristics of a reactance-grounded system, which are dependent 011 the magnihde of ground-fault current, at any point in the system are summarized it1 Table 6.3. It will be seeu that, if a system is to be grounded through a reactor, the available ground-fault current should be a t least 25 per cent of three-phase fault current. This is considerably higher than the minimum fault current desirable in a resistance-grounded system; therefore, reactor grounding is usually not considered an alternative of resistance grounding.

358

SYSTEM GROUNDING

GROUND-FAULT NEUTRALIZERS

A ground-fault neutralizer is a reactor connected between the neutral of a system and ground and having a specially selected, relatively high value of reactance. A line-to-ground fault causes line-to-neutral voltage t o be impressed across the neutralizer, which passes a n inductive current, I , (Fig. 6.14). This current is 180' out of phase and is approximately equal in magnitude (when the neutralizer is tuned to the system) t o the resultant of t,he system charging currents from the two unfaulted phases l a and I , . The inductive and capacitive components of current neutralize each other, and the only remaining current i n the fault is due to resistance, insulator leakage, and corona. This current is relatively small, and as it is in phase with the line-to-neutral voltage, the current and voltage pass through a zero value at the same instant. Hence, the arc is extinguished without restriking and flashovers are quenched without removing the faulted line section from service. For systems on which faults in air are relatively frequent, ground-fault neutralizers may be very useful because they reduce the number of circuit-breaker operations required t o remove faults, thus improving service continuity. They have been used primarily on syst,emsabove 15 kv consisting essentially of overhead transmission lines. A f e n ground-fault neutralizers have been used t o limit ground-fault current t o substantially zero when a ground fault occurs in a large 6900-volt ungrounded-neutral system such as in steel mills. Overvoltages are reduced also in comparison with an ungrounded system. This, however, is secoiid choice to resistor grounding, which provides ground-fault relaying to disconnect the faulted circuit. It should be noted, however, that failures in solid insulation, such as paper, varnished cambric, and rubber, are not self-healing as insulator flashovers are and are not extinguished by use of t,he ground-fault neutralizer as flashovers on an open line would be. I n some cases where it has not been deemed desirable by the plant Transformer

-Ib

FIG. 6.14 Giound-fault-current pattern in ryrtem grounded b y means of neutralizer.

(I

ground-fault

359

SYSTEM GROUNDING

operators t o trip a circuit 011 the occurrence of a ground fault, special arrangements have been used t o limit the damage due t o the flow of charging current and yet he able to locate the faulted feeder easily. One scheme is to use a ground-fault neutralizer in the neutral t o limit the ground-fault current and t o reduce switching surges t o safe values. A resistor is arranged t o be connected in parallel with the neutralizer when it is desired t o pass enough ground-fault current t o rause relays t o give a signal or trip the breaker of the faulted feeder, as illustrated in Fig. 6.15. Because of the current t o be switched, a power circuit breaker should be used for switching the resistor. The resistor and relaying are selected as if the resistor only were used. Such a scheme is expensive arid is used only in very special rases.

POWER CIRCUIT BREAKER REOUIRED GROUNDING REACTOR)

FEEDERS

3 CT’S

FIG. 6.15 Three-current transformers and ground relay required for each circuit in special ground-fault-neutralirer application.

Ground-fault neutralizers alone have heen used t o a limited extent in systems having the following characteristics: 1 . Large existing systems having only two current transformers per circuit 2. Where the switchgear is such that the addition of a third current transformer and residual overcurrent relay involves considerable expense 3. Systems having heavy charging current, in which case damage t o machines may result in the event of a ground fault if the system is left ungrounded 4. Systems which are susceptible to arcing grounds, for example, overhead lines

SYSTEM GROUNDING

3 bQ

A ground-fault neutralizer plus grounding resistor is also applicable for systems having the above conditions except that three current transformers and a residual relay are mandatory. One of the characteristics of resonant-grounded systems is that care should he taken t o keep the ground-fault neutralizer tuned to the system capacitance to minimize the development of transient overvoltages. Thus, when sections of the systems are switched on or off, it may be necessary to adjust the neutral reactance by changing the neutralizer taps. This operation may be readily performed by providing in the powerhouse an ammeter and control switch for remote control of a motordriven tap changer on the neutralizer. Thus, when parts of the system are switched, the neutralizer may be readjusted a t the time. SUGGESTED GROUNDING METHODS FOR INDUSTRIAL SYSTEMS Various types of impedances for system neutral grounding have been used for many years. A review of the various methods and their features has indicated that the desirable practice for industrial plants is as follows: 1. Systems rated 600 volts and below: solid grounding 2. Systems rated 2.4 to 13.8 kv: resistance grounding in most cases; solid grounding in a few cases 3. Systems rated 22 kv and higher : solid grounding A summary of grounding methods a t various voltage levels is covered in the following text and tabulated in Table 6.5. SYSTEMS

Mx)

VOLTS AND LESS

The grounding of 208Y/120-volt systems has been almost universally adopted. Grounding the neutrals of 480- and 600-volt systems has previously been impractical because of the fact that most three-phase transformers supplying these voltages have heen historically delta-connected. Now standard load-center unit substations are available with Y-connected secondary windings rated 208Y/120,480Y/277, or 600Y/346 volts, which enables lowvoltage system neutral grounding to be used. This provides a safer, more reliable system with lower operating cost. These substations are available with delta-connected primary windings. The new standard Y-connected secondary windings for the 480- and 6OC-volt substations enable these units to he used for either groundedneutral or ungrounded 480- or 6OO-volt systems, respectively. The lowvoltage neutral is brought out through a hushing. When grounding is desired, the terminal of the neutral bushing is connected to the low-voltage ground bus. When grounding is not desired, this connection is left off. Therefore, one line of unit substation transformers now provides al!

SYSTEM GROUNDING

361

the voltage ratings commonly required, plus the advantage of being able to operate the low voltage either grounded or ungrounded. Substations rated 240 volts are normally delta-connected. To obtain groundedneutral operation in this voltage class, either the 208Y/120-volt substations or 240-volt delta substations with grounding transformers may be used. Relatively high ground-fault currents are usually required to operate the standard overcurrent protective devices in circuits 600 volts or less. Thus low-voltage systems are solidly grounded to assure sufficient fault current for operation of protective devices. Three levels of fault current are of interest in considering grounding methods for low-voltage systems. These are (1) ground-fault currents greater than three-phase fault current, (2) ground-fault currents equal to three-phase fault currents, and (3) ground-fault currents less than threephase fault current. I t is desirable to design toward condition 2. Therefore, systems supplied only by transformers are grounded solidly, in which case condition 2 will obtain. However, if generator neutrals are grounded solidly, condition 1 will obtain. Since the generator may not be braced for the values of ground-fault current which can flow, it is necessary to use a low-value neutral reactor t o bring the ground-fault current down to three-phase fault-current value. For all practical purposes. this is thought of as “solid” grounding rather than “reactance” grounding because the resulting ground-fault current remains high. Where grounding transformers are required to establish a neutral, it is usually desirable to limit the ground-fault current to less than three-phase fault current to keep the transformer size within practical limits. In this case it is necessary t o investigate the tripping characteristics of interrupting devices and establish a transformer reactance which will assure adequate current for tripping. Condition 3 romes about in actual practice, either with transformer solidly grounded or generator grounded through a neutral reactor, because of the resistance of the ground return path. In medium-voltage systems, this resistance has little effect. Referring to Table 6.4, 0.1 ohm in a 13.8-kv system neutral has practically no current-reducing effect. However, in low-voltage systems this resistance has a considerable effect. Thus, 0.1 ohm in the neutral of a 240-volt system reduces the current by as much as 95 per rent. Also, the system voltage is low enough that the hazard from shock due to the flow of ground current through thestructure of a building is minimized compared with higher voltage systems. All air circuit breakers and fused knife switches must have three overcurrent elements for application on three-phase grounded-neutral circuits. These devices will trip for the same value of current whether the fault is three phase or single phase t o ground. Hence, special ground relays are not

SYSTEM GROUNDING

362

TABLE 6.4

Approximate Symmetrical Rms Ground Currents with Solidly Grounded Transformer Banks

-

Approximate ground currenl

Ground

ramformer

bonk kva ~

4,160

2,400

240

Amperes ~

Per cent'

Amperes?

~

3,900 6,000 7,300

87 67 41

4,500 9,000 17,500

100

3,750 7,500 15,000

3,000 3.650 3,900

67 41 22

4,500 8.800 16,400

100

98 91

4,500 9,000 I8.000

2,500 5.000 7,500

2,160 2,320 2,360

43 23 16

4,900 9,200 12,700

98 92 85

5,000 10.000 I5.000

1,500

1.330 1,370 1,380

26

7

4,900 8.300 11,500

94 81 56

5,200 10,300 20,600

350 350 350 350

8 5 3 I

2,660 3.100 3,400 3,500

63 44 24 I?

4,200 7,000 14,000 28.000

5

1,000 2,000

280 280 280 280

3 2

2,450 2.680 2,750 2,770

47 31 16 8

5,200 8.700 17.300 34,600

150 300 500 l.000

I40 140 140 140

1,330 1,370 1.380 1.380

26

5,200

13

10.300

8 4

17.300 34,600

300 500

l.000 2.000 480

~

~

7,500 I5.000 30,000

3,000 6,000 600

0.1 ohm ~

Per cent'

Amperes

6.900

,&.lo"Ce.

~

-13,800

Ground

,e.i.lo"Ce, 1.0 ohm

300 500

13

I 3 1

1 0.5

100 97

4,500 9,000 18.000

-- ---

* Per cent means per cent

t For

three-r,

eero ground resistance, grounc

;e fault eurrcnt urrent is 100 per cent of three-phase fault

current in all cases. Norm: 1. Voltage assumed sustained on high-voltage side of bank. 2. 7 per cent reactance assumed for bank. 3. Bank assumed to he only source of fault current. 4. Ground resistance of 1 ohm is maximum recommended for main substation grounds. Well-bonded grounded return connections in buildings usually have resistance of 0.1 ohm or less.

SYSTEM GROUNDING

363

used in low-voltage circuits when the neutral is grounded. However, it should be noted that all three poles of a circuit breaker open simultaneously for a single line-to-ground fault but that in the case of the fused knife switch the faulted phase opens first and the remaining fuses may or may not open, depending upon the degree of overload imposed on them by the remaining single-phase circuits. MEDIUM-VOLTAGE SYSTEMS 12.4 TO 15 KVI

As noted in Table 6.5 medium-voltage systems are usually resistancegrounded. In this range of system voltages, limited ground-fault current is highly desirable, in the usual industrial power system. Connection of rotating machines directly a t these voltages is common; hence resistance grounding is used to reduce the damage which may result because of a grouud fault in the machine windings. In the case of small systems supplied by generators, reactor grounding (to produce not less than 25 per cent of three-phase fault current) may be used in the interest of economy and berause ground-fault currents are not too large. If the system is small and supplied by transformers, they may be grounded solidly for the same reasons. Arresters for grounded-neutral circuits discharge a t lower levels of surge voltage and, hence, limit overvoltages to values about 20 per cent lower than arresters for ungrounded-neutral circuits. The use of grounded-neutral lightning arresters, to obtain better lightriing protection for equipment connected directly to overhead lines, is the principal reason for ever selecting reactance or solid grounding in place of resistor grounding for an industrial power system of 2.4- to 15-kv class. Reactance grounding at this voltage level is preferred only when all the following conditions exist: 1. Overhead circuits directly connected to rotating equipment are snhject to lightning exposure. 2. This type of grounding does not result in excessive ground-fault current. In such cases the neutral reactor of a generator must be selected so that X , / X , is equal to or greater than 1 for the generator alone, and X , / X , for the system must be equal to or less than 3 to permit the use of groundedneutral lightning arresters. If high ground-fault current is acceptable and the systems include suitable power-transformer banks, solid grounding may be preferred in some cases because neutral reactors or resistors are not required and lineto-ground lightning arresters may be used.

SYSTEM GROUNDING

-

Do not ground solidly

Grounding reiistor-

[we tronsformen which supply power to

-_---__

_ _ __-

Resistor-

Y

formers on system

-

-to

bur

~

G!

F. If no Y-connected generators or tronr- Use one or more grounding transformers with resistors

-~

loads on the system)

the system, ovoid transformen which are

__

___-

Use resistonce grounding

Grounding practice

E. If Y-connected transformers on system Use redstance grounding

D. If Y-connected generators on system

Condition

a

<

v)

economy

I. In m o l l systems. where the r e d l i n g ground-fault current would not be exesrive, lhe growding transformer may be solidly grounded in the intorest of

-

P

1. Capacity of transformen used for grounding should be of odequote size (see Toble 6.61 2' . Smell systemr, where the resulting growd-fault cur- 0 c rent would not be excessive, may be solidly grounded. z 0 if desired, in the interest of economy 2 0

--

grounded, if desired, in the interest of economy

3. Small systems. where the resulting ground-fault currant would not be excessive, may be reactance

orrede.,

quate sire bee Toble 6.61 2. When severe lightning exposure is present, gen. erator may be grounded through 1ow-vobe reactomc to permit use of g'~vnded-nevtml-t~pOlightning

1. Ganaroton used for grounding should be of ode-

Remorks

366

SYSTEM GROUNDING

HIGH-VOLTAGE SYSTEMS (ABOVE I5 KV)

Systems above 15 kv are nearly always solidly grounded, because these are usually transmission circuits with open lines in which, in most cases, grounded-neutral-type lightning arresters are desirable for better overvoltage protection and lower cost. In addition, rotating equipment is seldom connected directly to these systems; hence, limited ground-fault current, to prevent burning of laminations, is a less important factor than in the medium-voltage systems. In addition, voltages above 15 kv are not usually carried inside buildings; hence shock hazards due to high fault currents are not a factor. Finally, the cost of grounding resistors at these voltages is high. TABLE 6.6 Minimum Ratings of Generators and Power Transformer Banks for Grounding Maaimurn System

Mi"im"m

Short-circuit

K"0

K"0

Rating

1,000,000 500,000 250,000

7500 3750 1750 1000 750 375 187

l50,OOO ioo,000 50,000 25,000

NOTES: 1. If a smaller rating is used, system may he subject to transient overvoltages during clearing of ground faults. 2. Tahle assume8 7 per cent for power transformer hank reactance or generator zero-sequence reactance. 3. System short-circuit kva is maximum value possible when generator or hank to he grounded is only source of ground current, all other grounded power sources being out of servioe. 4. Tahle iS based on the criterion that the reactance of the grounded apparatus dhould he no more than ten times the equivalent three-phase short-circuit reactance. For example. B system having 250,OWkva three-phaso short-circuit duty has B reactance of 100 per cent on a250.000-kva. base. The grounded apparatus may then have a reactance of ten times this or 1000 per cent on a 250,000-kva hxse. If the apparatus has 7 per cent reactance on its own base, then its kva rating may he 7 X 250,000 100o

=

1750 kva

INDUSTRIAL VS. UTILITY PRACTICE

The characteristics and operation of industrial power systems differ in some respects from those encountered in utility systems; therefore, it may be expected that grounding practices would also be different. Such is the case, although the basic principles of neutral grounding are followed in both types of systems. A comparison of the pertinent characteristics of the two types of systems is given helow and summarized in Tahle 6.7.

SYSTEM GROUNDING

367

Utility practice in recent years has favored solid grounding. This method permits the use of grounded-neutral-type lightning arresters with the resulting reduced lightning-arrester investment and improved level of protection. In addition, solid grounding offers savings in the use of graded insulation in transformers at 115 kv and above. A large percentage of ground faults on utility systems occurs by means of insulator flashovers; and the high ground-fault current due to solid grounding does not cause expensive damage to equipment at the point of fault. In the case of industrial power systems, resistance grounding is preferred for voltage levels from 2.4 to 13.8 kv. The principal reason for this practice is to ensure reduced magnitudes of ground-fault current and consequent reduction in possible damage at the point of fault. This is a particularly important factor in the case of ground faults in the windings of motors and generators. Although a ground-fault of limited magnitude and duration may cause sufficient damage to require the replacement of several coils, the desired result is obtained when the damage is confined to the coils and the machine laminations are left intact. TABLE 6.7

Industrial vs. Utility Grounding Practice Utility

lnduitriol

.............. ........... ........ ....................

Derimbiiity of high ~ o n t i n ~ i tof y power. Yes Predominant method of conducting power.. Cable Per cent of system subject to lightning hazard.. Smoil Investment in lightning orresten. Small Predominmce of voltage levels above 15 k v . . Few ~ y i t e m r Rotding equipment at diitribvlion and tranrmisdon leveis. Yes in most core!

Yes Overhead lines

........

SELECTION OF SYSTEM GROUNDING POINT GROUND AT EACH VOLTAGE LEVEI

As illustrated in Fig. 6.16, it is necessary to ground each voltage level to achieve the protection and advantages of neutral grounding. For example, if the 4.16-kv system in this diagram were not grounded, this level would have all the characteristics of an ungrounded system; at the same time, the 33-kv and 480-volt levels would have the characteristics of grounded-neutral systems. Each voltage level may be grounded a t the neutral lead of a generator, power transformer bank, or grounding transformer. Any generator or transformer used for grounding should, as far as possible, be one which is always connected to the system. Alternatively, a sufficient number of

SYSTEM GROUNDING

368

I10 K V SYSTEM

Y 0

-

GROUND R E Q U I R E 0 HERE TO GROUND NEUTRA L OF 33 K V S Y S T E M

83 K V S Y S T E M

hl

n

fi

GROUND REOUIRED H E R E TO GROUND N E U T R A L OF 4.16 K V SYSTEM

4 16 K V S Y S T E M

n "f N

GROUND REQUIRED HERE TO %ROUND N E U T R A L OF 4 8 0 V O L T S Y S T E M

480 VOLT S Y S T E M

FIG. 6.16

Each voltage level is grounded independently.

369

SYSTEM GROUNDING

generators or transformers should be grounded to ensure at least one ground on the system at all times. GROUND AT THE POWER SOURCE AND NOT AT THE LOAD

When a power system is grounded at the neutral of Y-connected motors, or at primaries of Y-delta stepdown transformers, it is necessary to ground a number of these points simultaneously to ensure that the system will remain grounded when one or more of these loads are out of service. Consequently, ground-fault current may he excessively high when all grounded points are in service. Since power sources are fewer in number than loads and are less likely to be disconnected, they are preferred as grounding points, as shown in Fig. 6.17. Other disadvantages of grounding a t the load are:

GROUND AT L O A D

FIG. 6.17

Grmnd at the source and not at the load.

370

SYSTEM GROUNDING

1. Standard load-center unit substations have delta-connected primaries; therefore special transformers are required if the primaries are to he used as grounding points. 2. Since the ground-fault current is dependent on the number of feeders or grounding points in operation, it may vary widely depending on system operating conditions. This makes selective relaying more difficult and may require additional directional ground relaying to avoid false tripping of healthy feeder circuits. GROUND EACH MAJOR SOURCE BUS SECTION

When there are two or more major source bus sections, each section should have a t least one grounded neutral point, since the bus tie circuit may be open, as shown in Fig. 6.18. If there are two or more power sources per bus section, there should be provisions for grounding at least two sources at each section. NEUTRAL CIRCUIT ARRANGEMENT When the method of grounding and the grounding points have been selected for a particular power system, the second question to consider is how many generator or transformer neutrals will be used for grounding arid whether (1) each neutral will be connected independently t o ground or (2) a neutral bus with single ground connection will be established. The factors involved in this determination and recommended practices are given below.

GROUNDING RESISTOR

GROUNDING

RESISTOR

(MAY BE OPEN)

FIG. 6.18 Grounding independently.

CI

system consisting of two or more sections which may operate

371

SYSTEM GROUNDING

SINGLE POWER SOURCE

When a power system has only one sourre of power (geuerat,or or transformer), grounding of the neutral of this sourre may he arcomplished as shown in Fig. 6.19. Provision of a switrh or rirruit breaker to opeu the neutral rirruit is not necessary berause (1) neutral rirvuits have prartirally zero poteutial with respect t o ground exrept duriug the short interval of a fault; heuw breakdowns are unlikely; (2) it is not desirable to operate the system ungrounded by haviug t,he ground ronnertiou open while the generator or transformer is in service; (3) uenbral si7-itrhing equipment greatly increases the cost of grounding. I n the event that some means of disrounertiug t,he grouud (.ounertion is required in a particular rase, a metal-rlad rircuit hreaker should he used rather than a n open disconnert s\vit,rh for indoor iustallatioirs. The latter is hazardous t o personnel if a ground fault should orcur at the time the switch is opened or closed. The merit of metal-rlad equipment over open wiring and open devices is generally rerognised for all types of electric equipment notwithstanding the iurreased rost for metal-clad switehgear. MULTIPLE POWER SOURCES

When there are only a few generators or power transformer banks at a station, individual neutral irnpedIndividual Neutral Impedances.

4% $ Y8 Y S I N G L E GENERATOR

SINGLE TRANSFORMER

(b)

( 0 ) FIG. 6.19

Grounding CI single power source.

SYSTEM GROUNDING

372

ances are frequently used. With this arrangement, the neutral of each generator or main transformer bank is connected directly toits neutral impedance without intervening switching equipment, as illustrated by Fig. 6.20 ( a ) t o ( c ) . No special operating instructions are required since each impedance is automatically connected whenever the corresponding power

EiTHER OF THESE TWO GENERATORS WILL ALWAYS,BE ON BUS WJiEN BUS IS ENERGIZED

9 $

Y

EITHER OF THESE TWO SOURCES WILL ALWAYS BE

Q$ v

FIG. 6.20

Grounding of multiple-source system with individual neutral impedances.

SYSTEM GROUNDING

373

source is in use and is deenergized mhenever this source is disconnected. Whrn oiily two sourres are invoived, this arrangement is preferahle to the use of a iieutral biis. When several sources are invoived, howrver, the ground rurrent is increased each time a source is added and may be raised t o levels whirh are undesirably high. In the case of resistance grounding, earh resistor must he rated for suffirient current to assure satisfactory relaying mhen operating independentiy. Consequentiy, the ground current mith several resistors will be several times the minimum required for effective relaying. Wheii individual resistors are used, cirruiation of harmonic current betweeii paralleled gerierators is iiot a problem, since the resistance limits the circulating current t o negligible values. Neutra1 Bus and Switchgear. When there are more than two or three generators or power-suppiy traiisformer banks at one station, it is commonly drsirable to use only oiie resistor. Each power source is then connected t o the resistor through a neutral bus and neutral switching equipment as showii by Fig. 6.21. This arraiigemerit keeps the grouiidfault current t o a practical minimum, since the ground current from the station is never greater than can be supplied through a single resistor. It also assures the same vaiue of ground current regardless of the numher of generators or transformers iii use and simplihes ground relaying. The primary purpose of the neutral breakers is to isolate the generator ar transformer neutral from the neutral bus mhen the source is taken out of servire, because the neutral bus is energized during ground faults. Breakers are preferred t o discorinecting switches for indoor installations GROUND BUS

A

A

A

v

v

v

$ Y

FIG. 6.21

GROUNDING RESISTOR

Y

(d)

$ Y

Neutra1 grounding b>i rneonr of neutrd bur and switchgear.

374

SYSTEM GROUNDING

to assure safety to personnel. If disconnecting switches are used (as with some outdoor installations), they should be elevated or metal-enclosed and interlocked in such a manner as to prevent their operation except when the transformer primary and secondary switches or generator line and field breakers are open. As shown in Fig. 6.21, it is necessary to provide only two neutral breakers, and only one of these is closed, although all generators may he in operation. This will eliminate any circulating harmonic zero-sequence currents. When the generator whose neutral is grounded is to be shut down, the second generator is grounded by means of its neutral breaker before the line and neutral breakers of the first one are opened. This procedure will assure that the system is grounded a t all times. I n the case of multiple transformers, all neutral breakers may he normally closed because the presence of delta-connected windings (which are nearly always present on at least one side of each transformer) minimizes circulation of harmonic currents between transformers. Selection of Arrangement. When total ground-fault currents with several individual resistors would exceed about 4000 amp, it is suggested that neutral switchgear and a single resistor be considered for reuistancegrounded systems. When only one source is involved hut others may be added to the station, it is suggested that space be allowed for neutral switchgear to be added if this will be necessary later. For similar generators with reasonably equal load division, circulating currents are negligible, and it is often found practical to operate with neutral breakers of two or more generators closed. This simplifies operating procedure and increases assurance that the system will be grounded at all times.

CALCULATION OF GROUND-FAULT CURRENT

The magnitude of current which will flow in the event of a ground fault on a solidly grounded system is usually determined hy the impedance of the grounded apparatus, plus the impedance of the lines or cables leading to the fault and the impedance of the ground return path. For interconnected systems, calculation of the rurrent may he rather complicated. For simpler cases, an approximation of the available fault current may be obtained from Table 6.4. This table applies only for faults near the transformer terminals when power is supplied by a single transformer hank wit,h its neutral directly connected to earth and with the primary connected to a system of relatively large short-circuit capacity.

S Y S T W GROUNDING

375

RESISTANCE GROUNDING

When a single line-to-ground fault occurs on a resistance-grounded system, a voltage appears across the resistor (or resistors), nearly equal to the normal line-to-neutral voltage of the system. The resistor current is equal t o the current in the fault. Thus the current is practically equal to line-to-neutral voltage divided by the number of ohms of resistance used. For example, consider a 13,800-volt three-phase system grounded by a 4-ohm resistor. Normal line-toneutral voltage for this system is 13,800/-\/3, or 8000 volts. The ground current is, therefore, very nearly equal to 8000/4, or 2000 amp. If two such resistors were used on the system, the ground current would be approximately 4000 amp. Resistors have a voltage rating equal to line-to-neutral voltage and an ampere rating equal to the current which flows when this voltage is applied t o the resistor. Thus, for example, a maximum ground-fault current of approximately 2000 amp will he obtained on a system when using a 2000-amp resistor. This very simple method of calculating the ground-fault current is not suitable except when the ground-fault current is small compared with the three-phase fault current for a fault a t the same location. Horucver, it is usually suitable for systems grounded by resistance of ohmir values normally used. The method just outlined applies to faults on lines or buses, or at the terminals of machines or transformers. If the fault is internal to a rotating machine or transformer, the ground-fault current will be less. The reduction in current is primarily due to the internal voltage of the apparatus. I n the case of Y-connected equipment, this internal voltage is a t full value a t the terminals and is zero a t the neutral. If the fault occurs a t the neutral of any apparatus, no voltage will appear across the system grounding resistor; so the fault current will be zero. At intermediate points in the winding between the neutral and a terminal, the fault current will he intermediate between zero and the current to a terminal fault, as shown in Fig. 6.22. For example, at a point 10 per cent of the winding length from neutral, the ground-fault current will bc approximately 10 per cent of the value for a terminal fault. For a fault anywhere between this point and a terminal, the current will be more than 10 per cent of the amount for a terminal fault. In the case of delta-connected machines the internal voltage to neutral may he considered to he 100 per cent a t the terminals and 50 per cent a t the mid-point of the windings, as shown in Fig. 6.22(c). The mid-points have the lowest potential with respect to the electric neutral of any d h e r

SYSTEM GROUNDING

376

UNGROUNDED GENERATOR NO.1 FAULT OCCURS ÇOMEWHERE IN WINDING E E TWEEN NEUTR AND L I N E TERMINAL

GROUNDED - N E U T R A L GENERATOR NO.2

GROUNDING RESISTOR _íGROUND)

( o ) FLOW OF GROUND-FAULT CURRENT FOR I N T E R N A L FAULT IN WYE-CONNECTED GENERATOR b

N0.I

b

-

N0.2

-

(GROUND P O T E N T I A L )

-' " " ' ~ " ~ ' ~ ""'~ ~ ' ~ t

t

PHASE

IF

VI Vr I F

R

I N T E R N A L VOLTAGE T E R M I N A L VOLTAGE FAULT C U R R E N T GROUNDI NG RESIÇTANCE

IF =

(0)

O F EUS

VT -(V,

-VI]

= -V I

R

R

( b ) MAGNITUDE OF CURRENT FOR INTERNAL GROUND FAULT I S PROPORTIONAL TO I N T E R N A L VOLTAGE

INTERNAL VOLTAGE

100%

@

VOLTAGE FROM ELECTRICAL NEUTRAL TO ANY POINT ON WINDING I S EETWEEN 50% AND 100%

INTERNAL VOLTAGE 50% (c)MINIMUM GROUND-FAULTCURRENT FOR DELTA CONNECTED APPARATUÇ IS 50% OF MAXIMUM GROUND-FAULT C U R R E N T

-

FIG. 6.22 Magnitude of currenl for interna1 g m m d faulb in maichinei connected to ryrtem having a rerirtance-grounded neutial.

(I

part of the windings. Therefore, a ground fault at any point in the winding w i l l produce a ground-fault current of 50 per cent or more of the resistor current rating. REACTANCE GROUNDING

In a rractance-grounded system with a single line-to-ground fault, the ground-fault current may he compiited from the formula

SYSTEM GROUNDING

377

(6.1)

(resistance may usually be neglected) where X I = system positive-sequence reactance, ohms per phase X , = system negative-sequence reactance, ohms per phase X o = system zero-sequence reactance, ohms per phase X , = reactance of neutral grounding reactor, ohms E = line-to-neutral voltage, volts I , = ground-fault current, amp An illustration of the method of calculating the ground-fault current in a reactor-grounded system is given under Selection of Reactor Rating (see page 381 of this chapter). SOLID GROUNDING

In a solidly g r o u n d 4 system with a single line-to-ground fault, the ground-fault current may be computed from the formula

RATING OF GROUNDING EQUIPMENT Grounding resistors, reactors, and transformers are normally rated to carry current for a limited time only. The standard time-interval rating usually most applicable for industrial systems, with relays arranged to protect the grounding equipment, is 10 sec. The voltage rating of a grounding resistor should be the line-to-neutral voltage rating of the system. The insulation class of a reactor is determined by the circuit line-toneutral voltage. The voltage rating may be less than line-to-neutral voltage, it being cakulated by multiplying the rated current by the impedance of the reactor. The voltage rating of a grounding transformer should be system line-toline voltage. Grounding resistors are rated in terms of the initial current which will flow through the resistor with rated resistor voltage applied. Conventional cast-grid or corrosion-resistant steel resistors will average approximately 7 per cent increase in resistance for each 100 C rise in temperature. The rated current of a grounding reactor is the thermal current rating. I t is the rms neutral current in amperes which the reactor will carry for its rated time without exceeding standard temperature limitations. The rating establishes an rms current which is assumed to be constant during

378

SYSTEM GROUNDING

rated time for purposes of design, ralrulation, and test. In service it is expected that the current may be greater than rated value during the initial cycles of the fault. If a grounding transformer neutral is solidly connected t o ground, the current which will flow during a ground fault is primarily determined by the reartance of the grounding transformer. When a resistor is used between neutral and ground, the current rating of the grounding transformer is based on the resistor rated current. I n either case the transformer israted t o carry the required current for rated time nithout exceeding its rated temperature limits. Ratings of neutral grounding equipment are summarized in Table 6.8. TABLE 6.0

Ratings of Neutral Grounding Equipment

Equipment

......................... ......................... ............

Reridor.. Reador.. Grounding transformer..

Time, sect

Reactonce

I

.............

I 10

* Insulation rlass is drtrrrnintd hy circuit line-to-neutral v a l t a g ~ . t Tcn sxonds is ntlrquate ior the conventional system. Standard ratings oi 1 mi", 10 mi", and continnous are svailablc.

SELECTION OF RATING OF GROUNDING EQUIPMENT RESISTOR RATING

The determination of the resistor ohmic value, thus the magnitude of ground-fault cnrrent, is based on (1) providing suflicient current for satisfactory performanre of the system relaying scheme and ( 2 ) limiting ground-fault current t o a value which will produce minimum damage at the point of fault. 111most cases, the ground-fault current may be limited hy the iieut,ral resistor t o a value from 5 t o 20 per cent of that which would flow for a three-phase fault. T o determine the minimum ground-fault rurrcnt required, a diagram of the system must be available giving ratings of current transformers and types of relays for each circuit. This diagram should include Consideration of future changes. The magnitude of ground-fault current innst, he sufficient for operation of all relays. In general, if the current is high enough t o operate the relays on the larger circuits, it will he adequate for the smaller circuits. The ground currents required for satisfactory operation of various types

379

SYSTEM GROUNDING

TABLE 6.9

Selection of Grounding Resistor

(Values given ore minimum recommended sround-fault current in per

cent

OF rrrred current of current

tronrformcr.)'

Type of relay Equipmsnt protected

per Cent

Y-connected generators. motors, ond transformer.. Delta-connected gcncraton, motors. ond t r m s ,

........................... former,... ......................... Foeden and tie liner.. .................. B",e* ................................

40 100

...

40

...

Pilot wire-100% Current b d m c e - l 0 0 %

501

* For further discussion and analysis of ground-fault rdsying. ser Chap.

0. ground differential is a d d d to the generator, the ground-fault current may be lirnitzd to lower values (if othw systzrn requircmmts permit). $ Based on current differential. If voltage differential is providcd, the groundfault current may be limited to lower values.

t If

of relays, expressed in terms of current-transformer rating, are given in Table 6.9. Note that the ground-fault current under all system operating conditions must equal or exceed the minimum required for relaying each circuit connected to the system. This value is established by selecting the highest of those currents which meet the requirements of the several conditions set forth in Table 6.9. An example of the proper use of Table 6.9 for the system shown in Fig. 6.23 follows: Determine from Table 6.9 the ground-fault current each generator must produce when it is the only pover source. The larger machine must produce a ground current of at least 1200 amp (100 per cent of the rating of the current transformers for differential overcurrent protection in the larger generator circuit). This ground current of 1200 amp is higher than is required by Table 6.9 for any other circuit, in the system. With the larger generator disconnected, the smaller machine must provide a ground current of only 800 amp for its own relaying requirements (again 100 per differential overcent of the rating of the current transformers for its 01~11 current relays). The 1200-amp circuit need not be considered under this operating condition, and the 800 amp needed in the smaller generator circuit is found from Table 6.9 to be adequate for relaying requirements in all the other circuits. If all sources are grounded, it can be show1 that

SYSTEM GROUNDING

380

there will always be sufficient ground current for relaying requirements as long as each source produces what is needed when it is the only supply source. For Fig. 6.23 it is correctly concluded that the larger machine needs a 1200-amp resistor and the smaller oue only ao 800-amp unit. If ground differential relaying were added to the two generators, a

P$

1200 AMPERE GROUNDING RESISTOR

1200/5 C T

FEEOERS T i € CIRCUIT

8 0 0 / 5 CT

I

$

T

400/5C.T.

FEEOER

FIG. 6.23 Selectim of grounding rerirtor bcired on cvrient-tronrforrner rotingr

SYSTEM GROUNDING

381

further analysis might be made. I n this case other system requirements may determine the rating of the resistor. The largest feeder circuit demands a t least, 400 amp, but 800 amp is required t o satisfy pilot-wire relaying on the tie circuit, if present. This would establish a minimum rating for both resistors. If this value also satisfies the requirements of adequate gcnerator differential protection of the larzer generator, the rating of both resistors may be 800 amp. REACTOR RATING

The reactance of a neutral grounding reactor should be chosen to limit theground-fault current and the current in the faulted phase t o the desired value. As previously stated, in order t o minimize transient overvoltages the ground-fault current must not he less than 25 per cent of three-phase fault current. This corresponds to a ratio of X o / X 1 equal to 10. For reartance grounding of generators the current in any winding must not exceed the three-phase fault current. This corresponds t o a ratio of X , / X , equal t o 1 . This establishes the criteria for maximum and minimum values of neutral reactance. It can be shown that under the condition of X , / X , equal t o 1 for any given generator on the system the current contribution in one phase winding of this generator t o a line-to-ground fault any place on the system (external t o the generator) cannot exceed the three-phase fault current of the machine. However, the neutral current may exceed this value, as shown later. The calculations concerning momentary duty (which is of interest for mechanical strength and transient overvoltages) are made using suhtransient values of machine reactance. The calculations concerning the thermal rating of apparatus are made using transient values of machine reactance. I n calculating the reactance of a neutral reactor, the positive-sequence reactance XI is taken t o equal the machine subtransient reactance. The calculation for determining the required reactance in the neutral t o limit the current in the machine winding to three-phase fault current becomes a very simple procedure, as illustrated below: lo =

3E

XI

+ x2 + xo + 3x8

I (three-phase)

=

E

x1

~

where I, = ground-fault current, amp (for a single generator this also equals the fault current in the machine winding) E' = line-to-neutral voltage, volts

SYSTEM GROUNDING

382

X , = positive-sequence reactance of generator, ohms per phase Xp = negative-sequence reactance of generator, ohms per phase Xo= zero-sequeiiw reactance of generator, ohms per phase 5,= reartance «f neutra1 reaitor, ohms If I,,

=

I (three-phase) and X ,

=

X,,

a

+ xo + 3x.v _- -xi 3E

ZX, 2x,

(6.5)

+ x o+ 3X." = 3x1 3xx=

x,-x.

The rated riirreiit of a iieutral groundiiig reactor is the thermal current rat,iiig. It is the rms iieiit,ral curreiit iii amperes which the reactor will rarry uiider staiidard conditioris for its rated time without exceeding staiidard ternperature lirnitatioiis. T h e rating establishes an rms current xhich is assumed to be eoiistant duriiig the rated time, for purposes of dcsigii, calculatioti, aiid test. I n service it is expected t h a t the current may he grcatcr thaii the rated value duriiig the initial cycles of the fault. The ixrreiit ratiiig of a iieutral groundiiig reactor is equal t o the rms symmctrical vurreiit (deulated i)y usiiig the t,ransient reactaiice t o represent syiii'hroiious ma(.hiiie positire-sequeiice reactaiice and the proper negative- aiid zero-sequenre reactanie values of the system. The mrreiit whirh will floiv throiigh a generator iieutral reactor is iiot iiidepeiidrnt of systrm coiistaiits, hut mil1 vary mith the number and siae

N0.I

N0.2

A

KVA

x'd =

x'd = 26% x'h = I3 % xo = 7 %

13%

x0 = 7%

& I000

KVA

-

480 VOLTS-60 CYCLE

-

TOTAL CONNECTEO SYNCHRONOUS MOTOR LOAD 1000 KVA

1) FIG. 6.24 volt ryrtem.

Xh = 31% Xd = 2 5 %

Reectance-grounded generotorr and rolidly grounded tranrformer on 480-

SYSTEM GROUNDING

383

of power sources. Thus the current rating of a neutral reactor is determined by the number and characteristics of system sources and whether they are grounded or ungrounded. The following example illustrates the calculation of ratings for generator neutral-grounding reactors to limit the fault current in generator windings t o three-phase fault current. Assume a system as show1 in Fig. 8.24. To determine the reactance of each grounding reactor from Eq. (6.6), XI is taken as the subtransient reactance X y of the related generator and Xuas the zero-sequence reactance of the related generator

X (ohms)

=

x(% __ kvz base kva

lo

(ohms per phase)

(6.7)

No. 1:

No. 2:

xu =

0'482 lo = 0.0129 ohm 1250 0.0239 - 0.0129 = o,oo37 ohm XN = 3 To calculate the current rating of each reactor, it is first necessary t o calculate the total ground-fault current le from Eq. (6.3). The positivesequence reactance of the system X , is calculated using the transient reactance X : of synchronous machines and the negative-sequence reactance of the system X 1 is calculated using the subtransieut reactance of synchronous machines.

No. 1:

SYSTEM GROUNDING

304

No. 2:

2'0

0'482 1250

3

lo = 0.0129 ohm

Transformer:

x,= x,= x o= 5'5

0'482 1000

lo = 0,0127 ohm

Connected load:

x2

=a

25

x

0.4s2 x 10 = 0.0575 ohm 1000

An equivalent circuit with values indicated is illustrated in Fig. 6.25. From Eq. (6.3)

3 (480/d%)

I'

0.00803

+ 0.0063 + 0.00705

*830 = 39,000 amp

0.0214

N0.I

XI

x2

XO

3% FIG. 6.25

N0.2

TRANSF

CONN. LO40

,0920

,0479

,0127

,0713

.00803

.0479

,0239

,0127

,0575

,0063

.02 5 8

,0129

,0127

:022I

.0111

.00705

Connection of positive-, negative-, and zero-sequence impedance networks for calculating ground-fault currents for system shown in Fig. 6.24.

SYSTEM GROUNDING

385

From inspection of the equivalent circuit it is evident that this total ground-fault current will divide through the paths to ground in inverse proportion to the impedance in the path.

No. 1:

I,,

= 0'00705 l o = ~

0.0479

0.147 X 39,000 = 5900

No. 2:

I,,

=

0.007051 o - 0.294 X 39,000 = 11,500 0.024

~~

To complete the picture, the ground-fault current a t the transformer will be 0.00705 ~ = 0.555I X 39,000 o 0.0127

IDT = ~

=

21,600

The reactor for generator No. 1 must be rated for a t least 5900 amp, and for No. 2 a t least 11,500 amp. This serves to indicate the method of determining the reactor current rating and proves that this rating is determined by system characteristics. The rating may be considerably greater than the three-phase short-circuit current of the related generator, as shown shove. GROUNDING TRANSFORMERS

The electrical specifications of a grounding transformer are as follows: 1. Voltage. The line-to-line voltage of the system. 2. Current. The maximum neutral curyent. I n a resistance-grounded system, this current is determined by the neutral resistor. I n a solidly grounded system, the current is determined by the grounding transformer impedance and the system impedance. 3 . Time. Usually designed to carry rated current for a short time, such as 10 see or GO sec. 4. Reactance. This quantity is a function of the initial symmetrical system three-phase short-circuit kva (use Fig. 6.26). The theory behind the determination of grounding transformer reactance is discussed in the following. When the grounding transformer is resistance grounded, the criteria for limiting transient overvoltages is either Xo/X, equal to or less than 10, or R o / X oequal to or greater than 2. It should he noted that Ro as it appears in this relationship is equal to 3 times the resistance of the neutral kesistor. When the grounding transformer is solidly grounded, the criterion for limiting transient overvoltages is X , / X , equal to or less than 10. The criterion for using groundedneutral-type lightning arresters is that X , / X , should be equal to or less

SYSTEM GROUNDING

386

than 3, and R o / X , should be equal to or less than 1 (see Chap. 5 ) . A summary of criteria for selecting neutral reactance is shown in Table 6.10. In a system having a grounding transformer, its reactance is the principal part of X, in the above criterion. Also, the positive-sequence reactance XI is equal to the reactance of the system to initial symmetrical rms three-phase short-circuit currents. Thus, the grounding-transformer reactance is a function of the initial symmetrical system three-phase short-circuit kva. On a system otherwise ungrounded, the groundingtransformer reactance required to provide any specified X o / X 1ratio is given by the following formula: 100

%

50

4

I

a W a

30 20-lf

10

\ 5

,

I

.5

,110

31 00

MAXIMUM SYSTEM SHORT CIRCUIT MVA CALCULATED USING SUBTRANSIENT REACTANCES OF ROTATING MACHINES FIG.6.26 Maximum allowable reactances of grounding transformers lo limit ground-fault current to 25 per cent of three-phase fault current.

SYSTEM GROUNDING

387

(-Y,,/.YJ x kv' x 1000 x,, = system symmetrical three-phase short-rircuit kva (masiniom ohms prr ]>haw)

((i.8)

Taking the specific case X d X , = 10, the desired grounding-transformer reactance may lie idriilated Ily thr fiirmula 10,000 X kv' xo, = system initial symmetrical three-phase sh;rt-rircuit

kva ~

(lj.!))

Curves shoiving typical values of groniidiiig-transformer rrartaiire for this condition are shown in Fig. 6.2(i. For example, it is desired t o apply a groundirig transformer i n thr folloning system: 2400-volt 50,000 itiitial s y m m e t r i d short-ririwit Iivn. The grounding transformer reactatice should be 10,000 x 2.4% - = 1.15 ohms per phase (mas) 50,000

XCT = -~

Grounding Transformer Grounded Solidly. The gri)rtiidiiig-traitsformer voltage, reactatwe, and time are determined as outlined al>ove. When grounded-neutr&type lightning arresters arc t o IIC applied, ttie grouiiding-transformer reactatice may tie determitied by 90, =

_ 3000 _ X _ kv? _ _~ system initial symmetrical three-phase short-rircuit k v s

(li.10)

When grounding transformers are solidly grounded, care should lie taketi that the reactanre is selerted at a value 1011- enough to provide sufficient fault current for tripping relays, Tuses, and circnit tjreakers. Grounding Transformer Resistance Grounded. I u this CBSP it is not necessary t o provide less groiiiiditig-tratisformer reactanre than t,hc values giveu in Fig. 6.26 siiice groiuidcd-tteotral-type lightning arrrsters are not applicable in resistailre-grounded systems. The values of reactatire given in Fig. (j.Z(i are equal t o ten times the system reavtatm t o threephase initial symmetrical rnis short-circuit current. This is cqnivalent to the ratio Xo/XI equal t o 10. Where the ratio of Ro/Xo is equal to or greater than 2 , the ratio of X , / X , may be greater than 10 without the dsirger of severe transient overvoltages. However, I?, must be low enough to permit sufficient current for good relaying. On systems of 600 volts or lo\\-er it is usually desirable to permit currents of magnitude considerahly greater than 25 per rent of initial symmetriral rms three-phase short-circnit current in order to assure positive tripping of protertive devices. 111 such systems the grouuding transformer is connected solidly t,o ground. The minimum current required for tripping is determitied by esamination of the system aud the ground-

SYSTEM GROUNDING

388

ing-transformer reactance selected t o permit at. least that much current t o flow in the event of a ground fault. TABLE 6.10

Summary of Criteria for Selection of Neutral Reoctavce

For limiting transient OVerYOltoge

XdX,

Ro/Xo

For application of grounded-neutral lightning arresterr

XdX,

_____

........................ ......... ......

R e o d m C e grounding.. Grounding transformer solidly grounded.. Grounding transformer resistance grounded'.

* Either criterion

10 or leis I 0 or leis

........ ........

10 or leis

2 or more

30, 11 .. 3 or less

is mtisfactory.

OTHER METHODS OF GROUNDING LINE GROUNDING

I n lorn-voltage systems (600 volts and below) which in the past have almost universally been connected in delta, it was sometimes advocated that one line be grounded, as illustrated in Fig. 6.27. This was done i i r order t o obtain some of the advantages of grounding at minimum expense. Because of its limitations and disadvantages it is strictly a compromise method and is rarcly encountered in modern industrial systems. Staudard load-center unit substations are now readily availahle with Y-connected secondaries at 480 and 600 volts i n all standard kva ratings. For

( A ) N E U T R A L GROUNDING

-( 8 ) CORNER- OF- THE- DELTA GROUNDING

FIG. 6.27

Two melhodr of grounding a low-voltage power system.

SYSTEM GROUNDHG

389

existing 480-volt delta systems dry-type zigzag grounding transformers provide a relatively inexpensive method of establishing a neutral. One of the outstariding disadvantages of corner-of-the-delta grounding is the necessity for positive identification of the grounded phase throughout the entire system. Instruments, meters, and overload relays should not be connected in the grounded phase. MID-PHASE GROUNDING

Where existing systems at 600 volts and below are supplied by three single-phase transformers with midtap available, it is possible to gain some of the advantages of neutral grounding by grounding the midtap of one phase. This method is illustratrd in Fig. 6.28.

FIG. 6.28

One phore of grounded ot the mid-point.

(I

delta system

THE INFLUENCE OF GROUNDING METHOD ON CONTROL-CIRCUIT SAFETY IN SYSTEMS 600 VOLTS AND BELOW Frequently the safety of a control rirruit is offered as a reason for a particular method of grounding. In all cases where motor-starter control eircnits are set up without control transformers, it becomes evident that there are problems with regard t o circuit arrangement which must be considered in order t o minimize operating difficulties and persolinel hazards. Accidental motor starting due t o faulted control circuits may be associated with ungrounded systems as well as most types of grounded systems. During such times as accidental motor starting may constitute a hazard, it should be standard practice to open the discomiiecting means whether the system is grounded or not, and regardless of the method of grounding. Analysis of the fault performance of motor control circuits from the standpoint of safety reveals that hazards may exist with all types of ungrounded and grouuded systems. Three methods quite commonly used are described. A similar analysis should be made with any other contemplated arrangement. Figure 6.29 shows a direct,ly connected control circuit on a n ungrounded system. A ground fault on any phase will remain unnoticed, and protective devices will not trip. Assume that a ground exists on either phase

390

SYSTEM G-ROUNDING

a L.

44 FIG, 6.29

Control circuit on ungrounded system without control power transformer.

-

.

A

L 6

L

B

0.L .

A

C

SYSTEM GROUNDING

391

2 or 3. A subsequent ground fault,at point R will impress full line-to-line voltage across the coil arid close thc contactor. A ground fault a t point C will pick up the coiitact,or, and the stop button will not stop the motor. Figure 6.30 shows a system wit,h solidly grounded neutral. A ground fault on any phase x i l l cause circuit tripping, and the fault mill be isolated. A ground fault a t point R or C will impress liiie-to-neutral voltage (58 per cent) amass thc coiitactor roil. This will usually not pirk up the contactor, but it will prohably burn out the (.oil. If the “start” button is closed during t,his period, full fault curreiit Xi-ill flow until interrupted by a protert,ire devirc. .Iground fault a t C ii-hilc t,hc motor is running ivill prevent stoppitig the motor from the stop button if the contactor fails t o drop out on 58 per rent voltage. Furthermore, the stop button may be called upon t o interrupt a fault, current in excess of its capability. Figure 6.31 shows one method of connecting a control circuit on a line grounded system. Here a ground fault on any phase except 1 d l cause circuit tripping. A ground fault a t R or C ivill not pick up the contactor and remain unnoticed. Closing the start hutton under this condition will cause full fault current t o flow through the start button. SPECIAL PROBLEMS AUTOTRANSFORMERS

Poiver autotransformers are quite frequently used in public-utility poiyer transmission and distrihutiori systems; however, their use in industrial power systems as a part of the power distribution system is relatively infrequent. Autotransformers are quite common, however, in control and utiliaatioii equipment. Systems using autot,rausformers may be subject t o dangerous fundamental frequency overvoltage during system faults or from high-frequency or steep wave-front transient overvoltages on the lines, originating from lightning or switching surges. Since the magnitude of these overvoltages depends in part upon the method of grounding the system and autotransformer, the nature of these overvoltages will be explained. Consider the case of a n ungrounded system using an autotransformer as represent,ed in Fig. 6.32. Lines a , b, c represent the loiv-voltage system normally operating at line-to-line voltage and points d , e , f represent the terminals of a step-up autotrarisformer normally operating a t line-to-line voltage E2. 111the event of a line-to-ground fault on the line connected t o terminal d, thc loiv-voltage phases b and c are elevated aboveground by the amount

392

SYSTEM GROUNDING

8:dc=

1

Ed,

= -dE,2

4 3

+ E,2 + h,,E*

(6.11)

For example, in the case of an autotransformer rated 13.8/34.4 kv operating ungrounded on an ungrounded system, a line-to-ground fault on one of the high-voltage lines will impress a voltage t o ground on two of the loii.-voltage lines of

0.58

m+ 34.4' + (13.8)(34.4) = 25 kv

Obviously, this is an undesirable situation and cannot be tolerated. Solid grounding of the autotransformer neutral eliminates this type of overvoltage. Another type of overvoltage called transient inversion can occur in a n autotransformer, as illustrated in Fig. 6.33. Steep wave-front transient overvoltages produced by lightning or slyitching surges coming in over lined and arriving a t point, .J are impressed across a portion of the aut,otransformer winding .IK, point K remaining a t it,s normal frequency value until C , can he rharged. The result is that the port,ion of the ivinding J K has impressed upon it practically the entire voltage disturhance. Since the port,ion of the winding K N is closely coupled t o J K , the voltage ivill be stepped up in K N by t,he turn rat,io of K.V t,o K J . Since the initial disturbance may he several times normal voltage, and since this may be stepped up tivo t o twenty times or more by inversion (depending upon the winding ratios), it, is evident that a serious overvoltage may be experienced. The hazard due t o transient inversion is greatest for autotransformers in which the high- t o lowvoltage ratio approaches unity. This type of overvoltage can be eliminated by solidly grounding the neutral. I n cases where this is not feasible, a lightning arrester or Thyrite* resistor connected between the neutral and ground can he used t o minimize this voltage. The presence of a tertiary delta on the autotransformer also tends t o minimize transient overvoltages of this hature. Another system autotransformer connection which is subject to both normal frequency inversion and transient inversion is operation with the supply system neutral grounded and the autotransformer neutral isolated, as shown in Fig. 6.34. A line-to-ground fault on the high-voltage line 2onnected t o terminal h forces the voltage of point h t o that of N , . This inverts the phase of winding hd by impressing voltage N , , from point h t o d . The hd portion of the winding induces in the d N , portion of the winding a voltage of corresponding phase and of a magnitude depending upon the turn ratio

* Registwed

trademark of General Electric Company.

SYSTEM GROUNDING

393

T O HIGH VC LTAG E UNGROUNDED SYSTEM

a

FIG. 6.32

Ground fault on ryrtem with autotransformer connecting ungrounded systems.

(1

;ca

TO UNGROUNDED HIGH VOLTAGE SYSTEM

I

FIG. 6.34 Autotransformer neutral isolated, supply-system neutral grounded.

of the two parts of the winding. This results in a shift of point Nz,as shown in Fig. 6.35. Note that phase voltages N P jand N 2 k are far above normal for the case

SYSTEM GROUNDING

394

illustrated, where the step-up ratio was 2 : l . If the step-up ratio had heen 1 . 1 : 1, that is, the autotransformer normally boosting the low voltage hy 10 per rent, the faulted phase would be overexcited by ten times normal, resulting in a much more severe shift of N 2 and overexcitation of the other phases. That is, the closer the autotransformer ratio is t o unity, the more severe is the overvoltage from this type of fault. Overvoltages from this cause can be prevented by solidly grounding the neutral of the autotransformer. The resultiirg voltage magriitudes are given by the following relations:

For example, in the case of a n autotransformer stepping up 10 per cent = E.V%h=

E Z = 1 . 1 Der unit 1.1 = 6.35 per unit d T ( l . 1 - 1)

1

= 7.32 per unit

S o t e also in this rase that only the high-voltage lines and connected apparatus are subject t o overvoltage. The lorn-voltage lines are not subjected to any ahnormal voltages in this case. The foregoiiig examples illustrate the nature of the overvoltages which ran be obt,ained with autotransformers. I n general, solidly grounding the neutral of t,he autotransformer is a satisfactory means of eliminating

h

t

EZ b

t El

1 -

FIG. 6.35 Vector diagram illustrating normal frequency inversion of clutotmnrformer.

SYSTEM GROUNDING

395

overvoltages. The disadvantage of solid neutral grounding is that thirdharmonic currents aiid telephone interfereiice may heromc excessive iii rertaiii cases. These harmonic problems ran usually he eliminated h?, use of a tertiary delta 011 the ant,otransformer. See referetire 5 for a more romplet,e discussion of this s n b j w t . SYSTEMS WITH PUBLIC-UTILITY SUPPLY

Some iiidustrial systems are directly roiinerted at t,heir operat,iiig voltage t o public-utility systems. The scheme of grounding the industrial system should be properly coordinated \\-it,ht.hat, for the utility system. If two systems are interconnected by means of a transformer bank, at least one \\-inding of t,he bank will normally be roiinevted i t 1 delta, and t,his delta-rotiiierted i~iiidingwill make grorindilig of earh systrm itidepetident of grounding of the other. GENERATOR-TRANSFORMER UNIT INSTALLATIONS

Figure (i.36 shorn an arrangemelit using a distrihut,ioti-t,ype t,raosformer, loading resistor, and relay in the gciierat,or neutral. This scheme may he provided x i t h a 5- or 10-miti ratiiig t o permit time for traiisferriiig load off the atrected mavhiiie before it is takeii out of service. The distribut.ioii transformer will usiially have a rating of 25 t o 50 h a , aiid the relay may be connerted t o operate on resistor current, or volt,age, depeiidiiig 011 the particular illstallation. This system is used bei,aose sometimes the rost of the resistor and distribntioii-type transformer is less than the vost of + high-voltage Ionrurreiit resistor roiiiiertrd dirwtlv hetween the neutral and ground. DISTRIBUTION TRAN 5 F O R M E R

RELAY

Y

-I-

Grounding the neutral of distribution transformer.

FIG. 6.36

(I

generator-tr~nrformerunit with resistance-loaded

396

SYSTEM GROUNDING

This results in a n effective high-resistance ground which, because of the limited system and the absence of switching devices, is satisfartory from the standpoint of transient overvoltages, and since no problem of relay coordination is involved, the relaying problem is simple. THREE-PHASE FOUR-WIRE SYSTEMS

I n these systems, single-phase loads are connerted between phase cotiductors and the neutral conductor. The neutral conductor is insulated over its entire length except where it is grounded at its source of supply. The neutrals of such systems should be grounded so solidly that during a ground fault the voltage between any phase conductor and ground does not appreciably exceed normal line-to-ground voltage; otherwise, abnormally high voltage t o ground mill be impressed on the unfaulted circuits. T o be adequately grounded, therefore, four-wire systems must use solid or reactance grounding with ground-fault currents approximately equal to three-phase fault currents. This is usually accomplished by direct connection of transformer bank neutrals t o ground. FAULT DUTY MAY BE INCREASED BY SOLID GROUNDING

Solid grounding of the service transformer neutral can be responsible for fault currents exceeding the three-phase values. This may i n some cases necessitate larger circuit breakers than would be dictated by threephase faults. Here is another advantage of limiting the ground-fault magnitude. A specific example (Fig. 6.37) incorporating a representative arrarigement will serve to explain what factors contribute t o a greater line-toground duty. The positive-, negative-, and zero-sequelire impedance diagrams for the system in Fig. 6.37 are shown in Fig. G.38. Three-phase fault-momentary duty :

I,,

=

x- (IJ(l.5) El

1.0 (1040)(1.5) 0.0832

=

~

18,750 amp asymmetrical

=

Interrupting duty:

I,,,

=

$ (Id(1.0)

=

~

0.0985

(1040)(1.0)

=

10,570 amp symmetrical

Line-to-ground fault-momentary duty : 3E,

I",,

=

Xi'

+ x,+ xo

(IB)(l.5)=

3

0.0832

+ 0.0832 + O.OF (1040)( 1.5) =

20,650 amp asymmetrical

397

SYSTEM GROUNDING

Interrupting duty: lint

=

X;

3 +3XEI, + Xu (zB)(l'o) = 0.0985 + 0.0832 + 0.06 (1040)(1.O) =

12,900 amp symmetrical

Ratio of line-to-ground t o three-phase fault duty: Momentary:

Interrupting :

The key t o this problem is the fact that three-phase fault current is coutrolled by the factor l / Z 1 while line-to-ground fault is controlled hy thc factor 3/(Z, Z, Zo). If Z,, Z?, and ZU\yere all equal, the two fault currents would be equal. Any system condition uhich acts to reduce

+ +

A

IND MOTORS 3000 K V A X- 25 PERCENT

X" = 6 2 . 5 PERCENT ON 7500 KVA

7 5 0 0 UVA X = 6 PERCENT

SYN MOTORS 2000 KVA X " / X ' = 2 0 / 2 5 PERCENT

X ' ? X ' = 7 5 / 9 3 . 7 5 PERCENT ON 7 M 0 K V A

FIG. 6.37 Typical system where ground-fault current may be greater than three-phore fault current.

SYSTEM GROUNDING

398

-

, 1 6 1040 ~ bMP BASE CURRENT(IB)= 4 7500

POSITIVE SEQUENCE X"

if

N

I /O.Il 9.09 110.75 1.33 1 / 0 . 6 2 5 = ~ 12.02

1/12.02= 0 . 0 8 3 2

0.625

X" EQUIVALENT

*

0.0832 N

IF

.eF

x ' ( FOR

INTERRUPTING DUTY )

I/O.lI 1/0.9375= '9.09

10.16 0.9375

I/I0.16=0.0985

X ' EQUIVALENT

El

N

-

F

0

NEGATIVE SEQUENCE X 2 (SAME AS POSITIVE SEQUENCE X" EXCEPT OMlTTlhO E l ) X,

EQUIVALENT N

2NLh/L.

IF

0.0832 ZERO SEQUENCE XO N

-2AN\r

IF

0.06

FIG. 6.38

Sequence impedoncer expressed in per-unit on 7500-kva 4160-volt threephase bare, for circuit shown in Fig. 6.37.

SYSTEM GROUNDING

399

Z, or Z, or any condition which tends to increase Z, mill make the line-toground fault current greater than the three-phase value. In the example the utility service line coutaiiis a fair amouut of reactance ii-hic,h becomes iiicluded i n the positive-sequence uct\vork but, riot in the zero-sequence netn-ork. Thus, referred to the 4I6O-volt bus Xo is smaller than X , and XI. Had the iiwomiug h i e shortkiri.uit dut,y been 500 mva instead of I50 mva (lower X , slid X2),there wor~ldbe scarcely auy difference hetween line-to-ground and three-phase fault-currrnt values. 111 the case of load-renter suhst,ations for inataiii,e, the highvoltage supply system reactanre \ d l he very small compared il-ith the transformer rcactauce; thus solid ueutral groundiug i n prarhically all cases results i i i 110 iircreased short-riruuit duty. In passing it is iuterest,ing t o note that grountliug any other ueutrals of 4160-volt equipment i n the esample would redwe %,, and (.&useline-togrouud fault rurrciit to be elevated.

EXAMPLES

OF PRACTICE

Example 1. Consider the syst,em of Fig. 6.39. ;\pplirat,ion proredure is as follows: I . All necessary data are giacit oii the diagram. 2. Select groundiug method. From Table (i.5, voiiditioti d , resist,anre grounding is suggested. 3. Select all three generator urutrals as grounding point,s, to assurr that the system \\-ill alij-ays he grounded. 4. Review system rclayiiig. a. Ground-fault curretit required for relaying is as follo\vs, from Table 6.9: Generators.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 amp Feeder irsiug (iOO;5 current tmisformers.. . . . . . . GOO amp Smaller freders.. . . . . . . . . . . . . . . . . . . . I.ess than (i00 amp

h. Ground-fault protection is available 011 every circuit esvcpt the bus, which can be protected by neutral back-up rdays. 5 . Select neutral circuit arrangement. Resistors should be rated at least GO0 amp. Three iudividual rcsistors would provide a maximum total ground-fault curretit of only 1800 amp. Therefore, individual resistors are selei.ted, as suggested uuder Seiitral (:iri,uit Arrangements. 6. Select neutral grnuilditlg equipment. Since gtwerator breakers may be used for back-up protwtion, a IO-sei. timc interval rating may hc used. Resistors for iudoor mounting are suggested. 7. The oiily additional devices needed arc three iirwtral current transformers and three neutral overr.urrent relays.

SYSTEM GROUNDING

51

PHASE OVERCURRENT RELAYS(TH0SE ON GENERATORS HAVE VOLTAGE RESTRAINT)

@) 07

@

GROUND OVERCURRENT RELAYS GENERATOR PHASE DIFFERENTIAL RELAYS(NEUTRAL CT'S NOT SHOWN) GROUND OVERCURRENT BACK-UP RELAYS FIG. 6.39

Example 2.

Circuit diagram tor Example 1.

Consider the system of Fig. 6.40. Procedure is as follows: 1. All necessary data are found on Fig. 6.40. 2. Select grounding method. From Table 6.5, condition R. 3. Select location of grounding point, as the neutral of the main bank. 4. Select neutral circuit arrangement, as a direct connection to earth (from Table 6.5). 5 . Review system relaying. We shall consider in turn each breaker used for system protection. a. Transformer primary breaker. This breaker will operate on trans-

SYSTEM GROUNDING

401

former secondary faults. For such faults the ground-return path will be so short that its resistance will be negligible, if the transformer case is properly bonded to the system neutral. From Table 6.4,a ground-fault current of about 17,300 amp may flow to a terminal fault. This is over ten times the circuit rating and hence is sufficient for operating phase overcurrent relays in the primary circuit. 21. T r a n s f o r m e r secondary breaker. This breaker is primarily for bus faults. Since the maximum ground-fault current is only about ten times the circuit rating, fast tripping may not be 1600 AMP .) I 480 V provided by the breaker, but the primary breaker will give satisfactory protection, as discussed for transformer faults. G. Feeder breakers. Since the FIG. 6.40 Circuit diagram for Example 2. current maximum . mound-fault , (17,300 amp) is a t least twenty times the rating of the largest feeder breaker, these provide adequate ground-fault protection. 6. No neutral grounding equipment is required. The transformer neutral must, of course, be available for grounding. OPERATING EXPERIENCE Case 1. The following is quoted from theexperience of an engineer of a large glass-manufacturing company as related to an AIEE group recently. “ A few years ago in one of our large plate glass plants two feeder grounds occurred on two different phases about 2000 feet apart in two departments. When the fireworks and excitement had subsided the two departments involved were shut down for several hours until repairs could be made. Within a few hours after this trouble happened a number of motors, seven altogether, were brought into the electric shop with burnt out or grounded windings. This, we believe, is an important effect of the system surges that occur during very bad fault conditions such as this., The loss of production and damage on that occasion amounted to several thousand dollars. “With continuous process operations the hunting of ground faults is very difficult, and two grounds on the same phase but on two different feeders are exceedingly difficult to trace. This is because all the feeders must be opened a t once and closed one at a time to find the trouble. Our

402

SYSTEM GROUNDING

experience is that the first ground remains on the system because we cannot open the feeder breakers to hunt it. The result is that the system operates with two phases a t line-to-line voltage to ground, and the operating electrician hopes that no other grounds occur before he has an opportunity t o find the first one. “It was because of our experiences, such as I have mentioned, and the need in our operations for the highest possible service continuity, that we began to seriously consider the use of grounded neutral low voltage distribution systems. “The cost of a grounded neutral low voltage system is slightly higher than an ungrounded system. The additional transformer neutral bushing and connections, the neutral bus and wiring are items that add t o the cost. These are first costs that do not add more than one per cent t o the total cost of a unit substation. “Two of our plate glass plants are now operating 100 per cent with 600 volt grounded neutral systems, and t.wo other works are about 50 per cent cut over. Two window glass plants, operating a t 460 volts, are completely changed over t o grounded neutral and a third is in the process of being changed. Several new plants, one paint plant and two fabricating plants were built with 460 volt grounded neutral systems. Our total transformer capacity operating a t 600 volts or 460 volts grounded neutral is now 40,000 kva, consisting of 30 units. “Our experience with these systems has been very satisfactory. There is no question that the service reliability has greatly improved. A majority of the faults occur on branch feeders and are cleared by the local branch protection devices such as fuses. Troubles are localized and promptly repaired. As the electricians become used t o the new systems they are more enthusiastic and quickly learn, for instance, that a single blown fuse probably indicates a ground. None of them has expressed any desire t o return to nongrounded systems.” Case 2. An engineer from a large steel company reported as follows on experiences with a grounded-neutral 6900-volt system which was placed in operation in 1947: “The operating record of the system since the grounded neutral was installed is most gratifying. The ground faults experienced show a marked reduction in number and severity. For instance, during the year 1944, the number of ground indications recorded totaled 34. Of these 34 indications, 19 resulted in equipment failures such as grounded motor coils or flashed-over bushings. During the year 1951, there were two ground relay operations resulting in one equipment failure, and the first fifty weeks of 1952 show a similar record. Particular attention has been paid t o the severity of the damage caused by these ground faults. I n each instance i t appears that the relaying has been fast enough t o clear the fault before any destructive burning resulted.”

SYSTEM GROUNDING

403

REFERENCES 1. Concordia., C.., and H. A. Peterson. Arcine Faults in Power Svstems. Trans. A I E E . vol. 60, pp. 340-346, 1941. 2. Concordia, C., and W. F. Skeats, Effect of Restriking on Recowry Voltage, Trans. AIEE, vol. 58, 1939. I

3. AIEE Standards No. 32, “Neutral Grounding Devices.” 4. National Electrical Code. 5. Blume. L. F.. “Tranaformer Eneineerins.” John Wilry & Sons. Inc., 1951. 6. Allen, J. E., and S. K. W a l d o r f , k n g Ground Tests i n a Normally Ungroundcd 13-Kv, 3-Phase Bus, Tmns. AIEE, 1946, p. 298. 7. Shott. H. S., and H. A. Peterson, Criteria for Neutral Stability of Y-Grounded Primary, Broken Delta Secondary Transformer Circuits, Tiann. AIEE, 1941. p. 997. 8. Clarke, Edith, “Circuit Analysis of A-C Power Systems (Symmetrical Components),” John Wiley &Eons, Inc., New York, vol. I, 1943, vol. 11, 1950. 9. Application Guide for the Grounding of Synehronoua Generator Systems, AIEE Committee Report, Power Apparatus and Systems KO. 6, June, 1Y53, pp. 517-526. 10. Quinn, R. F., Should the High Voltage Neutral of a Wye-delta. Stepdown Transformer Bank Be Grounded? Gen. Elm. Rev., June, 1945.

chapter 7

by L. J. Carpenter, Shelby C. Cooke, Jr.,

R. H. Kaufmann, and David Stoetzel

Equipment Grounding STATIONARY EQUIPMENT, BUILDINGS, AND STRUCTURES OBJECTIVES OF EQUIPMENT GROUNDING

Equipment grounding consists of the connecting to ground of noncurrent-carrying metal parts of the wiring system or apparatus connected to the system. This includes all metal conduits, metal raceway, metal armor of cables, outlet boxes, cabinets, switch boxes, motor frames, transformer cases, switchgear enclosures, metal enclosures of motor controllers and other frames, and metallic enclosing cases of all electric equipment and electrically operated equipment. One objective of equipment grounding is to limit the potential between non-current-carrying parts of the plant and between these parts and earth to a safe value under all conditions of normal and abnormal system operations. To accomplish this objective, a plant grounding system is required. The purpose of this is t o seek to achieve a uniform potential in all parts of the structure and apparatus, as well as to provide that operators and attendants shall also be a t the same potential a t all times. By achieving more nearly a uniform potential throughout the grounding system, the chances of large differences of potential within reasonable reaching distance of a person, great enough to shock or injure an attendant when short circuit or other abnormal occurrences take place, are reduced. A grounding system is very likely called upon to function very infrequently, and inadequacy may become evident only a t that time. It is like the gun that nobody thought was loaded until someone pulled the trigger. When a ground fault occurs on an electric system, lives may depend on an adequate equipment-grounding installation. 404

EQUIPMENT GROUNDING

405

A second objective of equipment grounding is a low-impedance return path for ground-fault current. The hazard to personnel exists a t the time a ground fault occurs. Forcing the current to flow through a highimpedance grounding connection may create a dangerous potential difference. Also, high impedance a t joints and connections or insufficient cross section in grounding circuits may cause arcing and heating of sufficient magnitude to ignite nearby combustible material or explosive gases.

IMPORTANCE OF EQUIPMENT GROUNDING

The published data of the Division of Industrial Safety, Department of Industrial Relations, State of California, states that in the year 1952 there were 909 recorded electrical work injuries, of which 40 were fatal. For comparison, similar figures for several previous years are listed in the accompanying table. Iniuries Year

1946 1947 1948 1949 1950

Rdio Total

Fatal

305 572 755 693 690

28 57 48 42 33

10.9 10.1 15.7 16.5 20.9

I I I ;;:;

:;:: :;

Of the 909 recorded injuries, 153 could be related directly to contact with frame case or non-current-carrying metal parts. It was found that in these 153 recorded injuries either no grounding or inadequate grounding could have been responsible for the injury. Typical injury descriptions are as follows. “Refrigerator Repairman. Electric drill shorted out-severe shock; employee knocked out for about fifteen minutes.” “Carpenter. Operating portable electric hand-saw on wet groundreceived shock and dropped saw. Laceration, severe, dorsal surface a t base of distal phalanx, third finger left hand.” “Cabinetmaker. Ground wire broken off in drill. Hot wire grounded to frame, took hold of another grounded drill-unconscious about one minute.”

406

EQUIPMENT GROUNDING

Inasmuch as adequate equipment grounding tends to keep the potei tial difference between equipment frame and ground within safe limits, it can he safely said that these 153 accidents, or approximately 17 per cent of the total, were attributable to inadequate equipment grounding. The National Fire Protection Association consistently reports that about 10 per cent of all fires, representing about 10 per cent of losses from fires, are specifically attributed to “Electrical, fixed services, fires due to misuse or faulty wiring and equipment.” They also report that another 10 per cent, representing about 30 per cent of losses from fires, is of unknown origin. I t is inevitable that many of these fires were caused by inadequate equipment grounding and insufficient attention given to return paths for ground-fault currents. The increased use of system-neutral grounding has focused attention on the necessity for good equipment grounding systems to obtain lowimpedance return paths for ground-fault current. For safety to personnel, it is generally recognized that equipment grounding is required but is often provided as an afterthought and, consequently, may or may not he adequate-for the purpose intended. With a little careful consideration, it becomes apparent that a well-planned equipment grounding system must be provided whether the system neutral is grounded or not. Ungrounded neutral systems often operate for extended periods with a single-phase faulted to ground. During such periods, a contact between another phase conductor and a metallic enclosure raises the enclosure to full-line potential aboveground. Failure to provide a suitable connection between enclosure and ground presents a serious hazard to personnel. Also, the flow of fault current through a high-impedance connection during a double line-to-ground fault may create differences of potential of dangerous proportions. COMPONENTS OF AN EQUIPMENT GROUNDING SYSTEM

Definitions. For the purpose of further explanation of the grounding system the following definitions are established (see Figs. 7.1 and 7.2). Grounding electrode is a conductor embedded in the earth, used for maintaining ground potential on conductors connected to it and for dissipating into the earth currents conducted to it. Ground bus is a protective ground network used to establish a uniform potential in and about the structure. I t is tied solidly to the grounding electrodes. h typical ground network is illustrated in Fig. 7.3. Grounding conductor is a conductor used to connect equipment frames or wiring-system enclosures to the ground bus. The distinction between a neutral conductor (white) and a grounding conductor (green) is illustrated in Fig. 7.4.

FIG. 7.1

Typical grounding system

for on outdoor substation.

c Y 0

EQUIPMENT GROUNDING

408

FIG. 7.2

pical grounding system for a building and heavy electric opparatvr in the

building.

GROUNDING

ELECTRODES

CONNECTlON TO WATER PlPiNG

JOINT

CONNECTIONS TO BUlLDlNG STEEL E A C H SlDE OF EXPANSlON .IOlNT

FIG. 7.3

Typical ground bur.

410

EQUIPMENT GROUNDING

Types of Grounding Electrodes. A continuous underground waterniping system provides a very satisfactory grounding electrode (Fig. 7.5). Consideration should be given t o the size of pipe arid the extent of the system if this is t o be the sole means of connection t o earth. Table 7.1 tahulates the size of wat,er pipe in terms of equivalent grounding conductor or bus. Artiiicial grounding electrodes should also be used. Such electrodes may be rods, pipes, plates, or conductors embedded in the earth. They should be of noncorrosive metal, such as copper or copper-hearing steel. They are embedded in the earth by bring driven or by burial. The Ground Bus. The importance of a continuous metallic circuit of low FIG. 7.5 ~ ~ method ~ of grounding i ~ i~npedarice ~ l in the returo path for to o large w o t e i pipe. ground-fault currents is illustrated in Fig. 7.G. Fig. 7.Fa shows a 120/240-volt single-phase system with transformer neutral connccted to ground through a grounding electrode ivhich measures 10 ohms resistauoc to earth. The conduit is connected t o earth through a separate grounding electrode which measures 20 ohms t o earth. h fault oc(?urs lictween conductor B and the conduit.

Ground-fault current =

20

120 in

+

=

4 amp

Voltage drop from conduit to earth equals 4 X 20, or 80 volts. Figure 7.iib shows the same system with both transformer neutral and conduit connected to a common ground network which is connected t o earth through a single grounding electrode which measures 25 ohms resistance. A fault occurs between coilduotor B and t h e conduit. A high fault current will flow through the low-resistance ground-return path causing fault interrupting devices t o operate. Little, if any, current flows through the 25-ohm resistance, and therefore the conduit will remain very close t o earth potential. It should not be inferred from the above t h a t 80 volts potential is necessarily fatal. T h e example used simply ,illustrates t h a t appreciable resistance in the ground-return path results in a difference in potential during a ground fault which may be great enough t o be fatal t o a person

EQUIPMENT GROUNDING

41 I

stepping or reaching from one point t o another. A continuous path of low impedance i5 effeoted by means of a properly designed ground hus. The size of the ground bus is determined by the magnitude of current and the time of flow, based on the maximum allowable temperature rise. For bolted joints the temperature rise should be limited t o 250 C , and for brazed joints t o 450 C:. While ground buses and connections must be adequately braced t o withstand the mechanical stresses due t o the initial asymmetrical line-to-ground fault current, the heating effect of such current can generally be disregarded because of its short duration. The following equations may be used in determining the size of ground bus when copper is used for conductors. For bolted joiuts with initial temperature of 26 C and temperature rise of 250 C,

For brazed joints with initial temperature of 26 C and temperature rise of 450 C,

1

0 VOLTS

Illustrating the importance of a continuous metallic ground-return path of low imc-edance.

FIG. 7.6

412

EQUIPMENT GROUNDING

where A = cross section, circular mils I = ground fault current,, amp S = time of floy, see For ungrounded, impedance-grounded, arid solidly grounded systems it is usually easy t o determine the magnitude of fault current t h a t could flow in the ground hus. For uiigrouiided arid impedaiice-grourided systems this will approximately equal line-t,o-line fault current,, a i d for solidly grounded systems i t will approximately equal three-phase fault current. For the average grounded or ungrounded power distribution system wit,h adequate (auk protective devices, a time of flow of current of 10 sec is conservative atid may be used in thc above calculation. Aside from the theoretical considerations t,here are practical limits which may finally determine the maximum or minimum size of ground bus. For mcclianical st,rength the ground bus should not be smaller t h a n S o . 2/0-.4~vgconductor. It is not usually necessary t o exceed 500 MCM or equivalent for large generating stations and substations, or S o . 4/0 Awg for small stations and industrial plants. However, it may he desirable t o exceed these values where exceptional precaihon is required or where extremely high ground-fault currents are expected. A ground bus of adequate size for the installation should be run completely around the periphery of the building (see Fig. 7 . 3 ) . Grounding conductor material should he soft-drawn or medium-hard-draxn copper wire or copper bar. For steel-frame buildings the ground bus should be conneoted t o each outside building column (Fig. 7.7). In large buildings a network should he provided t o include internal buildiiig columns. The around bus should he connected t o electrodes a t int,ervals of 200 f t or less. If the building consists of more than one floor, each floor should have its own ground bus, these floor ground hnses in turn should be connected by a number of condiictors t o the main ground bus 011 the first floor. In buildings having no steel frameivork, a grouiid network equiualent t o the above should be provided. Where no steel framework is available, all grounding conductors must he taken directly t o the ground bus. Better aocessihility is obtained if an exposed bus is provided in the upper structure, and often it is inore economical t o install it in this FIG. 7.7 connection of ground bur to building column. manner.

EQUIPMENT GROUNDING

413

T h e gronnd hos may be installed in the form of a loopoiilsidetlie boundaries of the buildings, biiried in the backfill for the columii footings arid foundation wall. The loop ground bus should be installed a minimum of 18 in. orit,side &hebuildiiig wall and 18 in. below the finished grade. Where exposed to mechanical injury the conductor should be suitably prot.ectcd by pipes or other Siibstaritial guards. If guards are iron pipe or other magnetic matcrial, the conduct,or should be electrically coiiiiected t o both erirls of the guard t,o prevent, inductive choke ef- S,G,7,8 Codwe,d-type buried groundfcct. ( h i d a c t o r s laid iiiidergroii~id ing" connection, should, urilcss other\\ protected, he laid slack t o prevent their being readily broken. 811 hiiritd grounding corinectioiis should be made by brazing or (:nd\\-eld-t,ype joiiit (Fig. 7.8). All ot,her grouridirig comicctions may be ma& by brazing, Cad\reld, or with approved pressure terminals. Steel-to-copper coririectioiis should he made abovegrouiid wherever possible. The ground bus should he coiiiie , at least at two points. to a continuous uiidcrgroiirid vater-piping ciri or t o suitable grouiiding e k trodes (Fig. 7.9).

FIG. 7.9

Cadweld connection to water pipe.

414

EQUIPMENT GROUNDING

Grounding Conductors. Grounding conductors should he large enough to carry the ground fault current safely. Where the grounding conductor is insulated (green wire), it should be the same size as phase conductors. Where the conductor is bare, the temperature rise is limited to the same as the ground bus except that where exposed or adjacent to inflammable materials the total temperature is limited to 100 C. Although neutral conductors may be grounded a t the source, they should not be used for equipment grounding. Separate conductors should be used for neutral conductors and for equipment grounding. Equipment-grounding conductors must be identified with green color code. This is distinguished from neutral conductors which should be white color code. An illustration of this requirement is shown in Fig. 7.4. limiting Values of Resistance from Ground Bus to Earth. In large stations the resistance of the ground bus to earth should not exceed 1 ohm and should be made as much lower as can be realized economically. In small substations a resistance from ground bus to earth of higher values than that in large stations is generally permissible because the ground-fault currents are relatively smaller and they are in general only accessible to qualified personnel. Preferably, however, it should not exceed 5 ohms and should be as much lower as can be realized economically. For residential customers it is common practice to ground one side or the neutral of electric services on the premises. In cities this connection is ordinarily made to a water pipe, which usually provides a very lowresistance grounding connection. I n rural locations water systems may not be accessible, requiring that a driven pipe or rod must be installed for the grounding connection. The National Electrical Code requires that such grounding connections shall have a resistance not to exceed 25 ohms. Methods of measuring resistance 60 ground are discussed later.

TABLE 7.1

Minimum Size Water-pipe Electrodes

Size of Grounding Conductor

Sire of water Pipe,

or Bus, Awg

Inches I.P.S.

8 6

$5

A

I 1 >i 1% 2

2 1 /o 2/0 4lQ

$/a

2%

EFFECT OF IMPEDANCE IN EQUIPMENT-GROUNDING CIRCUITS

Of course, no one would contest the fact that reactance as well as resistance influences the return path taken by ground-fault currents. Recent tests, however, indicate that reactance his a much more marked effect than has been previously appreciated. This is particularly true

EQUIPMENT GROUNDING

415

where circuit conductors are enclosed in magnetic materials such as steel conduit or husways. I t was found that, when the enclosing conduit and a return coiiductor (of about equal resistance) outside the conduit were paralleled, the current divided approximately 20 parts in t.he conduit to 1 in the conductor at low current,s and 10 parts in the conduit t o 1 in the conductor at high rurreiits. About the same ratio also held true when the rurrent was allowed to divide hetween the conduit and a very low-resistauce steel-frame huildiirg structure. When the path by may of the eonduit was opened, a substantial voltage appeared between conduit and ground. With this same ret,rirrr conductor iriside the conduit., the current tended to divide about equally hetween conductor and corrduit. This leads to the conr~lusionthat the 60-cycle reactance of any groundreturn circuit remote from the outgoirig circuit conductor will likely be high compared ivith its resistance and limit the magnitude of groundreturn current which i t will carry. It may also he concluded that t,he conduit or enrlosing metallic structure will tend to carry an appreciable port,ion of thc fault rurrent and that failure to provide a continuous path will result i n arcing and heating, which may cause fires in combustible materials whirh may he rrear. This may account for the many fires that are reported tiy insurance statistics as caused hy faulty electrical circuits or of unkiioirii origin. POWER PLANT AND DISTRIBUTION EQUIPMENT

The frames of stationary or permanently located rotatirig electric equipment and the frames and enclosures of static equipment such as transformer t,anks and associated equipment permanently lorated should he grounded by dirert roniieotion t o the building grouiid b u s through a grouudiug condurtor equal in size t o the largest conductor in the line connected t o the equipment hut not less than No. 6 Amg nor greater than S o . 4/0 Awg (Figs. 7.10, 7.11, and 7.12). Driven ground electrodes should be employed a t earh outdoor substation. To provide a vorrvenient method of grounding switchgear, a ground hus should be provided as part of the equipment for structures or panels contaiiring such primary apparatus as current transformers, potential transformers, pon-er circuit breakers, and disconnecting switches and such other apparatus as relays, instruments, and meters which require grounding. Each of these metal structures, metal panels, or metal supports should lie individually connected to the switchgear ground hns, which must not he smaller in current-carrying capacity than 25 per cent of the highest rontinuous-current rating of any piece of primary apparatus t o which it is connected. Usually a 2- by >&in. bar is used.

416

EQUIPMENT GROUNDING

FIG. 7.10 Grounding connection-utdoor circuit breaker

FIG. 7.12

FIG. 7.1 1

Grounding connection-tronr-

former tank.

Grounding connection-motor

frame to building column

EQUIPMENT GROUNDING

417

This switchgear ground hus should, in turn, be connected t o the common station ground bus by suitable conductors having a current-carrying capacity equal t o t h a t of the switchgear ground bus (Pig. 7-13). I n many cases ( f o r example, i n metal-clad switchgear or other metal strue,ture) apparatus may be considered adequately grounded through their mountiiig on the structure. In some suhstatioii installations, where all connect,ioris are underground and there is no possibility of energizing t,he enclosing FIG. 7.13 Switchgeor ground bur. fence by falling overhead wires, it i s desirable t o keep <.he fence serrarated from the stat,iori ground bus. The reason for t,liis is t h a t during a ground fault the suhstatioii grouiid bus may he elevated in potential above true earth and may constitute a hazard t o personnel who come in contact with the fence.

STATIONERY UTILIZATION EQUIPMENT

The frames arid metallic eiiclosiiig cases of all electric equipment a n d electrically operated equipment not grounded through bus drop cable may he considered adequately grounded if bolted or welded t o the st,eel framework of a structure which has been suitably grounded. If this condition does not exist, aii individual grounding conductor should be run from the equipment t o t.he ground bus. A rigid conduit system may riot, coiistitut,c an adequate grounding circuit. However, because of the high reactauce of pat,hs provided outside the conduit, the conduit will carry a large percentage of t,he current, arid therefore i t should be made adequate if possible. If it is necessary t o run a conductor for equipment grounding, it sliould he inside the conduit which carries the corresponding phase conductors. A grounding coririectiori run inside the conduit or raceway through which poGer is brought t o the equipment must be an insulated conductor equal in size to the largest, conductor in the line feediiig the equipment but riot larger than No. 4/0 hwg. r l 1he frames and metallic eiiclosiiig cases of all electric equipment and

EQUIPMENT GROUNDING

419

The lead sheaths, shields, and armor of power cables should he grounded a t both ends, with the exception of single-conductor cables as noted helow. I n long cables it is sometimes desirable that sheaths also be grounded a t several intermediate points. The lead sheath, shield, and armor of large single-condurtor cables (500 MCM and above) should he grounded a t one end only to prevent rirculating currents. The sheath, shield, and armor of such a cable should he insulated from ground throughout the remainder of its length unless the cable is too long, in which case insulating joints must he provided to permit grounding a t a sufficient number of points to keep sheath voltagrs down t o desirable limits. For examplc, the mutual reactance bo neut,ral X , for a 500-MCM standard-strand cable, varnished cambric insulated, lead sheathed (approximate outside diameter, 1.5 in,), wit,h equivalent, spacing between cables of 3 in., is approximately 0.0525 ohm per 1000 ft. Assuming a current I , of 400 amp in t,he conductor, the induced vokage to neutral per 1000 ft can be calculated from the formula P, = I , X , (7.3) = 400 X 0.0525 = 21 volts

For lead-sheathed cable in duct, sheath voltage should be limited to 12 to 15 volts. For jacketed cable, sheath voltages of 40 to 50 volts may be permitted, because the jacket acts as an insulator. Conduits, wireways, busways, junction boxes, etc., should be grounded by adequate connection to the ground bus. The minimum requirements for these connections are established by the National Electrical Code. SMALL APPARATUS AND DEVICES

Small apparatus and devices present a very serious problem because of their diversity and because they are usually associated with low voltage and relatively small blocks of power which are inclined to be considered as innocuous. The need for proper grounding is even greater on these applications, however, because equipment in this category is widely used by personnel unfamiliar with the potential hazards, and it is usually not so carefully protected. This is emphasized with radher startling clarity by the figures from the Division of Industrial Safety of the State of California. Of the 153 reported injuries attributable to inadequate equipment grounding, 82 were associated with portable electrically operated tools and 33 more with “cords, plugs, portable extensions, etc.” In other words 115 out of 153, or nearly 75 per cent, were directly related to relatively small, so-called “harmless” devices. Incidentally, 699 nut of the total of 909 reported electrical work injuries, over 77 per cent, weredirectly related to circuits and equipment of 600 volts or less.

420

EQUIPMENT GROUNDING

In many cases where portable tools and equipment are connected through a plug and receptacle, a grounding connection can be made by using special three-pole outlets with a third, or grounding, conductor in the connecting cable. It i s now standard practice in many plants to use only three-pole outlets throughout the factory area where portable equipment is apt to be plugged in. This has not become an effective solution to the problem as yet, because portable tools are often supplied with a two-conductor cord and a two-pole plug, and i t is difficult to convert to a three-conductor cord and get the third conductor connected to the metallic case of the tool. Furthermore, the three-pole plug will not enter the conventional two-pole outlet, which tends to discourage its use until such time as a complete conversion can be accomplished. Other methods have been used to a limited extent, for example, a three-conductor cord with a clamp-type terminal dangling a t the plug end. This method does not present an effective solution because there is seldom a good place to fasten the terminal to ground, and it is too much bother to fasten it even if there is a good place. There appears to be a trend toward the adoption of the “green” conductor in wiring systems and the three-pole plug and receptacle, as standard. However, there are practical objections which must be overcome by education and manufacturing standardization. One of the important objections to the adoption of the green conductor in wiring systems is the danger of making a wrong connection in the box or in the tool, and by so doing connect the exposed metallic case to the ungrounded circuit conductor. This condition is more likely to occur in the small shop or the home where the inexperienced amateur electrician does the job than in a large industrial plant where experienced labor, adequate supervision, and intelligent safety practices are employed. LIGHTING FIXTURES

The minimum requirements for grounding of lighting fixtures should be those specified in the National Electrical Code. The Code states that every metal fixture shall be grounded unless all the following conditions exist: 1. The fixture is on a circuit operating a t 150 volts to ground or less. 2. The fixture is on an outlet wired with knob and tube work or nonmetallic sheathed cable. 3. The fixture is not mounted on a metal or metal lath wall or ceiling, or if so mounted is insulated from its support and from the metal lath by the use of insulating joints or fixture supports and canopy insulators. 4. The fixture is not installed within 8 ft vertically or 5 ft horizontally of laundry tubs, bathtubs, shower baths, plumbing fixtures, steam pipes, or other grounded metalwork or grounded surface.

420

EQUIPMENT GROUNDING

In many cases where portable tools and equipment are connected through a plug and receptacle, a grounding connection can be made by using special three-pole outlets with a third, or grounding, conductor in the connecting cable. It i s now standard practice in many plants to use only three-pole outlets throughout the factory area where portable equipment is apt to be plugged in. This has not become an effective solution to the problem as yet, because portable tools are often supplied with a two-conductor cord and a two-pole plug, and i t is difficult to convert to a three-conductor cord and get the third conductor connected to the metallic case of the tool. Furthermore, the three-pole plug will not enter the conventional two-pole outlet, which tends to discourage its use until such time as a complete conversion can be accomplished. Other methods have been used to a limited extent, for example, a three-conductor cord with a clamp-type terminal dangling a t the plug end. This method does not present an effective solution because there is seldom a good place to fasten the terminal to ground, and it is too much bother to fasten it even if there is a good place. There appears to be a trend toward the adoption of the “green” conductor in wiring systems and the three-pole plug and receptacle, as standard. However, there are practical objections which must be overcome by education and manufacturing standardization. One of the important objections to the adoption of the green conductor in wiring systems is the danger of making a wrong connection in the box or in the tool, and by so doing connect the exposed metallic case to the ungrounded circuit conductor. This condition is more likely to occur in the small shop or the home where the inexperienced amateur electrician does the job than in a large industrial plant where experienced labor, adequate supervision, and intelligent safety practices are employed. LIGHTING FIXTURES

The minimum requirements for grounding of lighting fixtures should be those specified in the National Electrical Code. The Code states that every metal fixture shall be grounded unless all the following conditions exist: 1. The fixture is on a circuit operating a t 150 volts to ground or less. 2. The fixture is on an outlet wired with knob and tube work or nonmetallic sheathed cable. 3. The fixture is not mounted on a metal or metal lath wall or ceiling, or if so mounted is insulated from its support and from the metal lath by the use of insulating joints or fixture supports and canopy insulators. 4. The fixture is not installed within 8 ft vertically or 5 ft horizontally of laundry tubs, bathtubs, shower baths, plumbing fixtures, steam pipes, or other grounded metalwork or grounded surface.

EQUIPMWT GROUNDING

421

Grounding of the fixture is permissible in a11 cases. TheNationalEleetricalCodealsos b t e s that fixtures shall he considered as grounded when mechanically connected in a permanent and effeetive manner to metal raceway, the armor of armored cahle, the grounding conductor in a nonmetallic sheathed cable, a separate grounding conductor not smaller than No. 14 Awg, or to gas piping, provided that the r a c e way, armor, grounding conductor, or gas pipe is grounded in an approved manner. Grounding of hook-suspended uuits supplied through a disconnecting plug should be accompliihed by means of a separate grounding conductor in the connecting cable equal in currentcarrying capacity to the liie conductors. Connection shouid be made through separate grounding contacts in the plug aud receptacle. The grounding of high-intensity mercury lighting fixtures and fluorescent lighting fixtures should he accomplished hy running an individual grounding conductor to the ground hus or ik equivalent. A continuous row of ktnres may be considered as one fixture if the mechanical connection hetween a11 sections is such that electrical continuity is assured. LIGHTNING AND LIGHTNING-ARRESTER GROUNDS

For lightnjng arresters, a local grounding connection should be made by driving rods into tbe earth near the arresters (Fig. 7.15). In addition, the lightning-arrester grounding conductor should be connected into the common station ground bus. For the average case, an arbitrary upper limit of 5 ohms resistance to ground has been estahlished. Lower values may be desirable, depending on the degree of proteotion required. The eonnection from arrester to ground should be as ehort and as straight as possible. The National Electrical Code states that a lightniagarrester ground wire shaii not be maller than No. 6 Awg. A larger siee conductor is required as thesystem voltage increases. For instance, a minimum of No. 2 is suggested for 4160-, 6900-, and 13,800-volt distrihution circuits. Properly made ground connections are an ewential feature of a lightpiug-rod system for protection of buildings. It appears to be more RG. 7.15 nectim

Lighhing-arreiter p r m d m-

422

EQUIPMENT GROUNDING

important to provide ample distribution of metallic contact in the earth than to provide low-resistance connection. Low-resistance connections are desirable, however, and should be provided where practicable. Ground connections should be made at uniform intervals shout the building, avoiding as much as possible the grouping of connections on one side. Each down conductor should have a ground connection. Ground electrodes should be at least 2 f t away from and should extend below the building foundation. They should make contact with the earth from the surface downward to avoid flashing at the surface. Interior metal parts or structures of a building should he grounded independently, and if they are within 6 f t of metallic roofs, walls, or down conductors, they should be securely connected thereto. METHODS OF MEASURING RESISTANCE TO EARTH

For new installations of grounding electrodes it is desirable that test electrodes be placed at the site for 60 to 90 days before tests are made in order that the earth around the electrode may become stabilized. By so doing, a more accurate indication of resistance to earth will he obtained. Theoret,ically it is possible to calculate the resistance of any system of grounding electrodes. However, soil resistivity is dependent on soil material, moisture content, and temperature. It has been found that the range of soil resistivity usually encountered varies between 500 and 50,000 ohms per cu cm. Also, a considerable variation in soil resistivity at a given location may he expected because of normal seasonal changes. Obviously, formulas for calculating the performance of grounding connections hecome so romplicated and involve so many indeterminate factors that they are of little value. Many such formulas have been developed, and they are useful as general guides, but the resistance of any given installation can be determined only by tests. Several methods of testing have been devised, all of which result in approximations of varying degrees of accuracy. It is important, that the measurement of grounding connection resistanre be made a t the time of installation and at periodic intervals thercafter to determine the adequacy and permanence of the grounding connection. Usually preiision in such measurements is not required, as it is necessary to know only whether the resistance is of the order of 1, 10, 100, or 1000 ohms. These values are indicative of whether the ground is satisfactory for the particular inst,allation or whether improvement is necessary. The commonly used methods of measuring and testing the resistance of a grounding connect,ion make use of two auxiliary electrodes in addition to thc one under test. The rcsist,ance may b i measured by the use of a voltmeter and ammeter, a Wheatstone bridge &h a slide-wire potenti-

EQUIPMENT GROUNDING

423

ometer, or hy self-contained instruments which give direct readings in ohms. Portahle ground-testing instruments provide the most convenieut and sstisfactory means for measuring the resistance of grounding connections. The common instrument used for measuring insulation resistance is not suitahle for measuring grounding-connection resistance, however, becanse it does not measure the iow values. Three methods of measuring and testing grounding connections are descrihed helow. Three-point Method. The connections and measurements of the resistance of grounding connections 5y the three-point method are illustrated in Fig. 7.16. Either alternating curreiit of commercial frequency

A U X I L I A R Y NO.1

A U X I L I A R Y N0.2

'11.

f.y

{Rz

RI = R x t R y Rp = R x + R z

Ra = R y t R Z Ry

i

R 3 - RZ

RX = RI RX

i

Rp

- Ry - RZ 2Rx

= RI

- Rg t R Z

= Rp

-RZ

i

RX =

+ Rp - R 3 RI + R p - A 3

RI

7

FIG. 7.16

Three-point method of mearuring rerirlonce of earth connoction.

425

EQUIPMENT GROUNDING

potential across the resistance. The connections for this method using alternating current for test are illustrated in Fig. 7.17. A current of known magnitude is passed through the electrode under test and one of the auxiliary electrodes. The drop in potential betweeu the electrode under test and a second auxiliary electrode is then measured, and the ratio of this voltage drop to the known current will indicate the resistance to earth. By using a voltage-measuring device which has a high impedance, the resistance of the auxiliary potential electrode will have no appreciable effect on the accuracy of the measurements.

L -

DETECTOR FOR MEASURING P O I N T OF BALANCE

S L I D E WIRE POTENT1 O M E T E R

R X t R y I S MEASURED B Y MEANS OF A WHEATSTONE

IS

BRIDGE OR OHMMETER

DETERMINED

RAtRB

FIG. 7.18

Ratio method of m e o w i n g resistance of earth connection.

426

EQUIPMENT GROUNDING

Ratio Method. The ratio method of measuring and testing the resistance of grounding electrodes is illustrated in Fig. 7.18. In this method the resistance of the electrode under test in series with an auxiliary electrode is measured by means of a Wheatstone hridge or ohmmeter. A slide-wire potentiometer is connected across the same two electrodes with the sliding contact connected to a second auxiliary electrode through a detector for determining the point of balance. The point of balance on the potentiometer fixes the ratio of the resistance of the test electrode to the total resistance of the two in series which was determined in the first measurement. CHARACTERISTICS OF EARTH AS A CONDUCTOR

The characteristics of earth as a conductor are generally variable and unpredictable. They are even difficult to measure with any degree of accuracy. The various methods of measuring resistance previously described are valuable to the extent that they give a magnitude of resistance in the earth circuit which is relatively accurate and will indicate whether ground connections are satisfactory or not. It should not he expected that current magnitude in earth circuits can he calculated by conventional formulas. Current values in earth circuits have been found to vary with the frequency of applied voltage, in some instances inversely. Earth resistance varies with applied voltage. A-C and d-c resistances of driven ground rods have heen found to differ greatly. Furthermore, the resistance of an earth connection varies with earth composition, moisture, temperature, season of the year, depth and diameter of rod, and other reasons. Because of the nature of earth circuits, it is difficult to determine the shock hazard from measured resistance in event of a ground fault. Likewise, the effectiveness of ground-fault protective devices which depend on current return in the earth is not easy t o determine. Metallic return paths for ground-fault currents ensure a safe and adequate means of providing a level of safety which is predictable. They should be used in industrial systems wherever possible. GROUNDING LARGE PORTABLE MACHINERY FOR SAFETY

Large portable machinery, such as electric shovels and machines of a similar nature, are not located on permanent foundations and normally receive power by flexible trailing cable a t a relatively high voltage. Voltage levels of 2400 or 4160 volts are common. Such machines as these present serious problems with respect to the safety of operators and

426

EQUIPMENT GROUNDING

Ratio Method. The ratio method of measuring and testing the resistance of grounding electrodes is illustrated in Fig. 7.18. In this method the resistance of the electrode under test in series with an auxiliary electrode is measured by means of a Wheatstone hridge or ohmmeter. A slide-wire potentiometer is connected across the same two electrodes with the sliding contact connected to a second auxiliary electrode through a detector for determining the point of balance. The point of balance on the potentiometer fixes the ratio of the resistance of the test electrode to the total resistance of the two in series which was determined in the first measurement. CHARACTERISTICS OF EARTH AS A CONDUCTOR

The characteristics of earth as a conductor are generally variable and unpredictable. They are even difficult to measure with any degree of accuracy. The various methods of measuring resistance previously described are valuable to the extent that they give a magnitude of resistance in the earth circuit which is relatively accurate and will indicate whether ground connections are satisfactory or not. It should not he expected that current magnitude in earth circuits can he calculated by conventional formulas. Current values in earth circuits have been found to vary with the frequency of applied voltage, in some instances inversely. Earth resistance varies with applied voltage. A-C and d-c resistances of driven ground rods have heen found to differ greatly. Furthermore, the resistance of an earth connection varies with earth composition, moisture, temperature, season of the year, depth and diameter of rod, and other reasons. Because of the nature of earth circuits, it is difficult to determine the shock hazard from measured resistance in event of a ground fault. Likewise, the effectiveness of ground-fault protective devices which depend on current return in the earth is not easy t o determine. Metallic return paths for ground-fault currents ensure a safe and adequate means of providing a level of safety which is predictable. They should be used in industrial systems wherever possible. GROUNDING LARGE PORTABLE MACHINERY FOR SAFETY

Large portable machinery, such as electric shovels and machines of a similar nature, are not located on permanent foundations and normally receive power by flexible trailing cable a t a relatively high voltage. Voltage levels of 2400 or 4160 volts are common. Such machines as these present serious problems with respect to the safety of operators and

428

EQUIPMENT GROUNDING

a voltage difference between the frame and ground due to the contact resistance between the two. In the case of a portable machine, the electric connection between the frame and the surrounding earth is a poor one, and, hence, a dangerous voltage can he developed when ground current is flowing. This voltage, approximately equal to the product of the ground current and the frame-to-ground resistance, produces a definite shock hazard to personnel. Personnel in the vicinity of portable machines may come in contact with this frame-to-ground voltage in various ways. Workmen during idle moments frequently sit on the ground in the shade of the machine with their backs against the “cats.” In making adjustments or repairs to the dipper, the workmen usually will be standing on the ground a t some distance from the base of the machine and will be touching the metal parts of the dipper as shown in Fig. 7.20. The frame-to-ground voltage will be transmitted along metal, such as compressed-air, fuel, or water lines, drag chains, tow cables, or other metal parts connected to the frame. FACTORS IN SHOCK HAZARD

The factors which affect the shock-hazard problem are illustrated in Fig. 7.20. The physical arrangement is shown in Fig. 7.20A and the equivalent electrical diagram in Fig. 7.20B. The two most important factors with respect to the shock hazard are 1. Magnitude of ground-fault current 2. Resistance from machine frame t o ground (R,) The ground-fault current which will flow upon the occurrence of a ground fault is determined by the power-supply voltage acting on the fault circuit impedance. The normal power-system impedance composed of the substation-transformer, transmission-line, and drag-cable impedances Z,, Zf, and Z, are of necessity relatively low t o permit proper performance of power equipment connected to the system. The fanlt current might be in the order of thousands of amperes if i t were limited only by these impedances. A protective system designed to handle such large ground currents without creating dangerous shock voltages would require a very extensive and costly low-resistance grounding circuit for the portable machine. However, another means of limiting the shock voltage is by limiting the current which will flow on a ground-fault occurrence. This can be done with no sacrifice in the power-machinery performance by introducing a neutral grounding impedance Zo,Fig. 7.20. A resistor, rather than a reactor, is used as the neutral grounding impedance for two reasons: (1) to avoid the twice normal transitory current magnitude encountered with high reactance, which would be required to

EQUIPMENT GROUNDING

429

limit the ground-fault current to 50 amp and (2) to avoid the relatively high switching and arcing ground transitory overvoltages which would be permitted by such high-reactance grounding. The portable-machine frame-to-ground resistance is very indefinite. Even under favorable circumstances, the ground resistance will not be low, and it may be 100 ohms or more on rock formations. Even a gr0u.p of driven ground rods at the line end of a trailing cable may have a ground resistance of 10 ohms or more unless the installation is carefully made and maintained. (Kote that a 10-ohm ground and a ground-fault current of 100 amp would produce a shock hazard potential of 1000 volts between the machine and ground.) Available shock-hazard information and data indicate that the possible voltage from frame to ground should not be allowed to exceed 100 volts and that adequate relaying should be provided to deenergize the circuit immediately upon the occurrence of a ground fault. A ground-fault current of 50 amp and a portable-machine ground-return-path resistance of 2 ohms have proved successful as a satisfactory compromise. A reliable ground-circuit resistance of 2 ohms requires a ground-return conductor from the frame of the portable machine back to the electric-supply substation, as represented by impedances Z, and Z, in Fig. 7.20B. This ground-return circuit alone is designed t o come within the 2-ohm limit since other ground paths are so variable. This practice assures adequate safety even if the ground resistance of these other ground paths is very high. For example, even if the resistance of the other path is infinite and the ground-fault current is limited to 50 amp, the maximum voltage which can occur between the frame and ground is 100 volts. In most cases it will be found that the ground-return-conductor resistance is appreciably under 2 ohms, thereby limiting the maximum frame-toground voltage to a value under 100 volts. CIRCUIT PROTECTION

To remove circuit elements which have experienced insulation failure and the resulting ground fault, circuit breakers should be located t o isolate the faulty circuit elements. Since it is not normally desirable t o shut down the entire system because of a ground fault in one of the shovels, circuit breakers should be located so that the system can be switched in smaller sections. A common arrangement is that of having circuit breakers at the main substation feeding individual pole-line circuits which go out to the portable-machine areas. Along the pole lines at the point a t which feeders are tapped off to supply portable equipment, portable switch houses are commonly used which include circuit breakers. From the portable switch houses, or cable skids, feeder cables are run to

.

A

SUB.

MAIN

GROUND

PROTECTIVE CIRCUIT

GROUND

L

I

:

-

?

S

b

\

A.

V

HIGH TENSION

STEP TRANSFORMER DOWN BANK

PHYSICAL ASPECTS OF CIRCUIT

1

-

1

1

-

--

=

FIG. 7.20

-

-

EQUIVALENT ELECTRIC CIRCUITARROWS INOICATE CONTINUITY OF CIRCUIT ONLY

25

z2

B

-

-

-

R2

SHOVEL FRAME IS AT T n E POTENTIAL O F rnis POINT

T

-

Simplified circuit showing the factors involved in the shock hazard of portable machines.

8.

-

-

0

0

-z

$2

t

t

, h

23A

-

L I N E TO GROUNO FAULT SHOVEL # I

-

-

ZPA

-

I

TRANSFORMER

GROUND FAULT SHOVEL U 2

m

EQUIPMENT GROUNDING

433

individual shovels. Therefore, by proper detecting means, selectivity in circuit-breaker tripping can be attained so that upon the occurrence of a grouiid fault only the faulted portion of the circuit is deenergized and the remainder of the system continues to operate as normal. As mentioned before, upon the occurrence of a ground fault the possibility of shock hazard makes it essential that the grounded feeder be disconnected from the system immediately. If the grounded feeder cannot be isolated, then this entire system should be immediately deenergioed. This can be accomplished by means of back-up relaying equipment operating the circuit breaker located a t the substation. SlMULTANEOUS GROUND FAULTS

Should two line-to-ground circiiit faults evist simultaneously on different phases a t different, locations, current of line-to-line short-circuit magnitude will flow in the protective ground circuit between the two fault locations. This current flow is not limited by the neutral grounding impedance (Fig. 7.21) and could produce unsafe voltages on the protective ground circuit. To avoid this possibility, automatic switching equipment must be applied which will function immediately if a ground fault occurs and isolate that part of the system. Thus each individual ground fault is automatically removed as soon as i t occurs. On an ungrounded system, ground-detector equipment can be employed to detect the first ground fault and to operate to clear the entire system since with such operation there is no way of telling upon which feeder circuit the ground fault has occurred. This in most cases is not acceptable as an operating procedure.

II/

POWER SUPPLY UNGROUNDED

I 1 1 FEEDER TO PORTABLE MACHINES

FIG. 7.22

Ground-detector system for an ungrounded power supply.

434

EQUIPMENT GROUNDING

For maximum service reliability and safety, large portable machines should not he operated on power-supply systems which do not have their neutrals grounded for the reason mentioned in the previous paragraph and also because : 1. If for any reason the feeder breaker is not tripped open upon the occurrence of a ground fault (relays sometimes are blocked), the development of a second ground would permit dangerous voltages to appear where they could produce shock hazards. 2. Ungrounded systems are subject to higher transient overvoltages, as a result of circuit switching or restriking faults, which increase the possibility of simultaneous line-to-ground failures. When the source transformer is not Y-connected, it is still possible to ground the neutral of the system by the use of a zigzag connected grounding transformer in combination with the resistor. When applying a grounding transformer, it is necessary that care be used in its selection, as outlined in Chap. 6. Large portable machines often receive power from the same system which supplies the mill load. Because of relaying limitations, it is necessary that the ground-fault-current level for the mill power system be considerably higher than that recommended for the portable-machine power system. I n order to meet the requirements for both of these types of load, a 1: 1 ratio delta-Y connected isolating transformer is recommended for installation in the feeder supplying the portable machines. The use of such a transformer permits grounding each system separately at the proper level. RELAYING

The portable-machine power system should incorporate relay protection against phase and ground faults. Relays should be applied a t all circuit-breaker locations and in the grounding connection, as outlined in Chap. 9 and shown in Fig. 7.24. Because of the low value of groundfault current involved, extra-sensitive relays are often required. Careful attention must he given to the selection of current transformers to be sure that their characteristics are satisfactory a t the operating burden imposed by the protective relay. The current transformer and relay combination should ensure positive operation a t currents well below the 50-amp level. It is recommended that all relaying equipment used in connection with ground-fault protective system for large portable machinery and shovels be sensitive to current magnitudes of about one-third the solid groundfault value (15 amp when a 50-amp neutral resistor is used) so &s to assure adequate current for relay operation.

EQUIPMENT GROUNDING

435

LIGHTNING PROTECTION

The light,nirig protective equipment at the main substation should consist of the protective devices which would normally he applied t,o a substation of the same size and voltage ratings (see Chap. 5 ) . It is recommended that lightning arresters be applied at the junction of the trailing cable and the pole-line feeder to limit the maximum linr-to-ground voltage. Surge protective capacitor equipment should be applied on the load side of the circuit breaker, within the portahle machine, to reduce the slope of the wave front for the protection of rotating machines. h good driven ground should be established a t the tap-off station, to which the protective ground circuit as well as the lightning-arrester ground leads should be connected (see Fig. 7 . 2 0 A ) . The shork hazard associated with a direct lightning stroke to a portable machine is a real one and extremely difficult to eliminate. Sumerous possible protective systems have been investigated, but all are hopeless to attain in practice. Often a thunderst,orm means a temporary shutdown of operations anyway. Thus during the danger period of the storm the personnel should seek shelter in Ihe metal cab of the machine or at a spot well removed from the machine. To stand on the ground near a portahle machine is about as dangerous as st,anding under a h e . GROUNDS AT SUBSTATION

One important point concerning the design of the main suhstation should be mentioned here. The ground-fault current, which may be supplied by the high-voltage supply system, is not cont,rolled by the l(ica1 electric-system design and may often be quite large. Any line-to-ground flashover at the main substation will allow the high-voltage-system ground current to flow into the main substation ground in the manner illustrated in Fig. 7.23. This current would persist for a time interval governe:l by the switching time of the high-voltage protective circuit breaker. A 5000-amp ground-fault current (which might easily be equaled or exceeded) in combination with the main substation ground resistance of 2 ohms (which would be considered as a good station ground) would cause the entire main substation structure to be elevated 10,000 volts with respect to ground and remain at this potential until the high-voltage-system protective circuit breaker operated (probably % to I see). If the portablemachine circuit protective ground were physically interconnected with the main substation ground, it is obvious that this high potential would he distributed t o the frames of all portable machines and thus constitute a serious hazard.

EQUIPMENT GROUNDING

A36

STATION STRUCTURE-

r---- --I I

-__

1

I

I

TRANSFDRMEF

RSC HT

EC

POTENl'IAL

R,

2. FIG. 7.23 Fault diagram for high-tension system.

As a result of these considerations, it is recommended that the protective ground circuit, originating a t the step-down transformer low-voltage neutral, be insulated from the main substation ground system with the same insulation level as applied t o the low-tension line circuit, and grounded a t an adjacent separate grounding connection. The voltage gradient in the earth surrounding the main substation diminishes very rapidly as one moves away from the substation. In general a 50-ft separation between the two grounding terminals is sufficient to avoid any substantial coupling between the two ground beds. It is important to avoid any direct interconnection between these two grounds such as would be produced by buried metal pipelines, etc. TYPICAL POWER SYSTEM

A typical power system for supplying portable-machine loads and illustrating the points discussed in this chapter is shown in Fig. 7.24.

T T - - - MAIN

----- --- 1 SUB STATION STRUCTURE FEEQER CB

ER FEEDERS

NEUTRAL)

I I I

i C T TO SWITCH :H nousE

@Y+!+2L FIG. 7.24

SURGECbP PROTECTIVE TRAILING CbBLt SURGE PROTECTIVE CbP

Typical distribution system for supplying power to portable machines.

430

EQUIPMWT GROUNDING

REFERENCES

1. Eaton. J. R.. Grounding- Electric Circuits Effectively. . , Gen. Eke. Rev.. June.. July. ., ~ U g u s t ,i94i. 2. Groundinv- Princides and Practice. a consolidated remint of five srticles Dublished in Eleelrical Efi'nginewing.January to May, 1945. 3. Safety Rules for the Installation and Maintenance of Electric Supply and Communication L i n q National Bureau of Standards Handbook H32. 4. Safety Rules for the Installation and Maintenance of Electric Utilization Equipment, National Buresu of Standards Handbook H33. 5. National Electrical Code. 6. MeCall, M. C., and I,. R. Harrison, Some Characteristics the Earth as a Conductor of Electric Current, U.S. Bur. Mines Repl. Inuesl. 903,September, 1952. 7. Electrical Work Injuries in California Industries Year Ended December 31, 1952. Depsrtment of Industrial Relations, State of Califomin. 8. Code for Protection Against Lightning, National Bureau of Standards Handbook

Is'

46. 9. Accident Facts, National Safety Council, 1952 edition. 10. Kaufmsnn, R. H., Some Fundamentals of Equipment, Grounding Circuit Design, AIEE Paper 54-244, 1954.

Ch-apter 8

by W.

C. Bloornquist

Power-factor Improziement * Why are power engineers interested in plant power factor, what causes low power factor, and how can it be improved? The objectives of this chapter are to answer these questions briefly and to include handy application information for power-factor problems. More complete information on this subject can be found in the book from which parts of this chapter were reproduced. * In summary, the effects of low plant operating power factor may be any or all of the following: overloaded cables, transformers, etc.; increased copper losses; reduced voltage level, resulting in sluggish motor operation; reduced illumination from lighting, especially where incandescent lamps are used; and increased power costs where a power-factor clause, or its equivalent, is part of the rate structure and is enforced. Generally, low power factor is due to partially loaded induction motors. Frequently drives are “overmotored,” i.e., the motor is selected to handle the largest load but usually operated at less than full load. There are also other factors contributing t o lower power factor, such as replacement of incandescent lamps with fluorescent lamps; use of rectifiers instead of synchronous motor-generator sets for d-c power supply; and increased installation of various induction devices, electronic eqrripments, air-conditioning units, etc. Most of these changes or replacements are in the interest of worker comfort and efficiency, lower manufacturing cost, and technological advances; the fact that they contribute to lower plant power factor is of secondary importance. As plants become motorized it can be expected that the plant power factor will become poorer unless some corrective measures are taken. Improvement of power factor can reduce power costs, release electrical capacity of the power-distribution system, raise the voltage level, and

* P a t s or this chapter are reproduced, with permission. from “Capacitors f o r Industry,” hy W. C. Bloomquist and R. C . Wilson, copyright General Electric Company, John Wilcy & Sons, h e . , XPWYork, 1950. 439

440

POWER-FACTOR IMPROVEMENT

reduce the system losses. However, the two main reasons for improving the power factor are (1) to reduce the power bill when there is a powerfactor incentive in the rate clause and (2) to increase or release electrical capacity of the power-distribution system. Although the first is still of primary importance, the second is becoming more important as engineers recognize the economics. This is especially true when capacitors are used for power-factor improvement because the electrical capacity released is valued at several times the cost of capacitors. The two most common methods for improving power factor are shunt capdcitors or synchronous motors. Each has its own application; usually the capacitor method is most economical and practical for existing plants, while the synchronous motor finds its main application when a new and large motor drive is added.

POWER-FACTOR FUNDAMENTALS The usual definition of power factor in terms of the phase relationship of voltage and current in a sine wave is intentionally avoided because it is abstract and difficult to translate into a simple physical concept. The concept used here-hased on the fact that there are two types of current in an a-c circuit--is particularly helpful in understanding the effect of power factor on system operation and understanding capacitor applications. Although the following discussion on fundamentals is written around the use of capacitors because they generally are the most practical and economical means for improving the power factor, these fundamentals also apply to other met,hods, such as synchronous motors and condensers. The current required by induction motors, transformers, fluorescent lights, induction heating furnaces, resistance welders, etc., may be considered to be made up of two separate kinds of current: magnetizing current and power-producing current. Some loads, such as incandescent lights, require only power-producing current. Power-producing current (or working current) is that current which is converted by the equipment into useful work such as turning a lathe, making a weld, or pumping water. The unit of measurement of the power produced is t,he kilowatt (kw). Magnetizing current (also known as wattless, reactive, or nonworking current) is that current which is required to produce the flux necessary to the operation of induction devices. Without magnetizing current, energy could not flow through the core of a transformer or across the air gap of an induction motor. The unit of measurement of magrietizing volt-amperes is the IMovar (kvar). Total current is the current that is read on an ammeter in the circuit. It is generally made up of both magnetizing current and power-producing


441

current. The unit of measurement of total volt-amperes or “apparent power” is the kilovolt-ampere (kva). Most a-c power systems require both kilowatts and kilovars. 2 2 Does Not Equal 4! The arithmetic applicable to everyday life 2 = 4. It is unfortunate that instead of follows the simple rule that 2 following such a simple rule the addition of kilovar current and kilowatt current follows a principle of geometry. If the kilowatt and kilovar components of current are each 2 amp, Fig. 8.1, the total current may be found from the right-triangle relationship as follows:

+

+

(Kilovar current)2

+ (kilowatt

=

22

(total current)z

+ 22 = (total current)?

4

+ 4 = (total current)2

Total current =

v‘‘%= 2.83 amp

+

Therefore, 2 2 does not equal 4. The following useful formulas apply when kw, kvar, and kva are substituted for their respective currents:

+

kva = d ( k w ) * (kvar)z kw = 4 ( k v a I 2 - (kvar)z kvar = 4 ( k v a I 2 - (kw)*

(8.1)

(8.2) (8.3) 2 AMP.

KVAR CURRENT

TOTAL

CURRENT 2 AMP.

FIG. 8.1

Diagrom showing cornponent currents in a-c circuits.

2.83 AMPS.

WHAT IS POWER FACTOR?

Power factor may be expressed as the ratio of power-producing current in a circuit t o the total current in that circuit. Another definition of power factor, which is generally more useful, is the ratio of kw or working power t o thb total kva or apparent power. Thus

POWER.FACTOR IMPROVEMENT

442

kw kva kw = kva X pf kw kva = Pf

Power factor

= -

(8.4) (8.5)

Stated another may, the power factor is that factor by which the apparent power must be multiplied in order to obtain the working power. For the case illustrated in Fig. 8.2 the power factor is SO/lOO, or 0.8, or, as it is commonly expressed, 80 per cent. The angle included between the kva and the kilowatt components is called the power-factor angle and is designated by the symbol 8. The cosine of this angle (cox e) is the power factor.

\

6 0 KVAR

100 K V A

\

FIG. 8.2

Right-triongle relationship for power-factorcoldotions in a-c circuits.

The actual calc.ulation of power fact,or is illustrated hy the following example. Example 1. What is the power factor of the load on a 460-volt threephase system if the ammeter indicates 100 amp and the wattmeter reads 62 kw? Since in a three-phase circuit

4 3 volts x amperes 1000 1.73 X 460 X 100 = 79,6 -_ -

kva =

1000

Power factor = kw/kva 78 per cent.

=

62/79.6

=

0.78, or, as it is often expressed,


443

LEADING AND LAGGING POWER FACTOR

The terms leading and lagging power factor are apt to be confusing, and they are meaningless unless the direction of both kilowatt and kilovar flow is knoivn. Generally, however, in industrial plants only the load power factor is considered, in which case the following rule may be helpful in differentiating between leading and lagging power factor: " The power factor is lagging if the load requires kilovars and leading if the load furnishes kilovars." Thus, an induction motor has a lagging power fartor TABLE 8.1

Power Factor of Load and Source Direction of flow I

A t load Figure

8. a 8. b 8. c

Type of load

I

Induction Synchronous motor (overexcitedl Synchronous motor

Kw

I I Kwr

Power factor'

Kw

Kvar

Power factorl

-___

In

In

Log

Out

Out

Lag

In

Out

Lead

Out

In

Lead

In

out

Out

Lag

In

lundereidtcdl

I

lag

* Power factor measured at the load.

t Power factor measured at the generator

z gAG) LOAD

LOAD

INO. MOTOR

[ (LAG1 LOAD

SYN. MOTOR (UNDER EXCITED 1 (C)

FIG. 8.3

Diagram for use with Table 8.1.

444


berause its magnetizing requirements must be supplied by the power source or other sources. On the other hand, an overexcited synchronous motor can supply kilovars (from t,he motor d-c field action); so such a synchronous motor has a leading power factor. Table 8.1 and associated Fig. 8.3 indicate the power factor for common operating conditions for both loads and supply sources based on the direction of kilowatt and kilorar flow. It is obvious from this table that the terms leading and lagging are apt to be confusing. I n order to avoid this confusion, varmeters are replacing power-factor meters. The varmeter has a zero center point with scales on either side, one labeled “ i n ” and the other “out.” In most industrial circuits the kilowatt flow is in only one direct,ion, e.g., to a motor load; so single-scale wattmeters are customarily used. However, in a t,ie line or transfer circuit a wattmeter with a center point should be nsed. Kilovar readings are generally more useful than power-factor readings as they indicate t,he actual magnitude of the magnetizing components. However, if the power-factor value is needed, it can he computed from the kiloir.at,t and kilovar values or read directly from the alignment chart, Fig. 8.9. POWER FACTOR OF A GROUP OF LOADS

The power factor of an individual load is generally known or can he estimated quite closely. The poiver factor of a group of different loads should generally be calculated. This can he done quite simply by means of t,he relations explained previously. Example 2. Figure 8.4 shows a substation supplying three different kinds of loads: incandescent lights, synchronous motors, and induction motors. The substation power factor is obtained from the total kilovars and kilowatts of the various loads, and from these the total substation k r a and poiver factor may be found as follows: 1. Find the kilowatts and kilovars of each load

a. 5 0 - k ~ alighting load: Since incandescent lights are primarily a unity poiver-factor load, all the current is kilowatt current; so kva = kw. b. 300 hp of connected induction motor loads: Assume kva load = 0.75 X (connected motor horsepower) with an operating power factor of 80 per cent lagging. kva = 0.75 X 300 = 225 kw = (0.75 X 300) X 0.8 = 180 kvar = 4 ( 2 2 5 i z - (180)2 = 4.5 0 . 6, 2 5 - 32.400 =

4/18,225=

135

(8.5) (8.3)

445


c. 75 hp of 0.8 leading power-factor synchronous motor: At full load assume kva = motor-horsepower rating = 75 kva

X (0.8) = 60 kw = 75 kvar = -\/(75)* - (60)2= 4 5 6 2 5 - 3600

(8.5) (8.3) 2. Find the kilowatts and kilovars that the substation must supply, =

-\/m 45 =

SUBSTATION

INCANDESCENT LIGHTS 50 KVA 1.0 PF

-

INDUCTION MOTOR LOADS 225 KVA 0.8 PF (LAG)

SYNCHRONOUS MOTOR LOADS 15 KVA 0.8 PF (LEAD)

/t

1 5 KVA

45 KVAR

180 KW

60 K W

SO KVAzKW

135 KVAR

(b) 290 K W 90 KVAR

FIG. 8.4

Construction of load diagram for Example 2.


44b

a. Kilowatts: = 50 Lights Induction motor load = 180 Synchronous motor = 60 Total = 290 kw

b. Kilovars: Lights Induction motor load Subtotal

=

o

= 135 = 135 kvar

(Since an overexcited synchronous motor has the ability t o supply kilovars, the net kilovars that must be supplied by the substation is therefore the difference between the kilovars supplied by the synchronous motor and the kilovars required by the induction motor loads.) Induction motor loads require Synchronous motor supplies Substation must supply

135 kvar 45 90 kvar

3. Find the substation kva and power factor. kva

+

d ( k i v ) 2 (kvar)2 d(290)' (90)' = 484,100 8100 = d 9 m O = 303 290 Power factor = - = 0.956 lagging 303 = =

+ +

(8.1)

(8.4)

Since the substation must supply some of the kilovar requiremei1t.s (the synchronous motor is riot large enough to supply all the load kilovar requiremeuts), the over-all p o w r factor is lagging. The various loads are added diagrammatically as shown in Fig. 8.4. HOW TO IMPROVE POWER FACTOR

When the kilovar current in a rircuit is reduced, the total current is reduced. If the kilowatt current does not change, as is usually true, the power factor will improve as kilovar current is reduced. When the kilovar current becomes zero, all the current is kilowatt current and therefore the power factor will he 1.0 (unity) or 100 per cent. For example, in Fig. 8.2, if a capacitor is installed t o supply the total or 60 kvar, the line power factor xi11 he 1.0. Thus, the power factor may be improved hy supplying the load kilovar requirements by a capacitor.


447

This is shown pictorially for another example in Fig. 8.5a and b. The working load requires 80 amp, but because oE the motor magnetizing requirements of GO amp, the supply circuit must carry 100 amp. After a capacitor is installed to supply the motor magnetizing requirements, the supply circuit needs to deliver only 80 amp to do exactly the same work. The supply circuit is now carrying only kilowatts; so no system capacity is wasted in carrying nonworking current. From the right-triangle relationship the following important fact can be drawn: the simple subtraction of kilowatts from total kua never equals the kilovars ezcept at unity power factor. In actual practice, it is generally not necessary or economical to improve the power factor to 100 per cent; capacitors or synchronous motors are used to supply part of the load kilovar requirements and the supply system the remainder.

INDUCTION MOTOR LOADS

(b)

(0)

FIG. 8.5

Schematic arrangement showing how capacitors reduce total line current by supplying magnetizing requirements locally.

Example 3.

In the example of Fig. 8.2, suppose that the power factor is to be improved from 80 to 90 per cent with capacitors. How much of the load magnetizing requirements is furnished by capacitors? (See Fig. 8.6 for diagram construction.) Without capacitors a t 0.8 power factor km = 80 kvar = 60

448


IVith capacitors and 0.9 power factor kw = 80 same 80 kva = = 88.9 0.9 Line kvar = ~ ‘ / ( 8 8 . 9 )~ (SO)z = 47903 - 6 4 5 = 38.7 ~

(8.6)

(8.3)

Since the line supplies 38.7 k m r hiid the load requirement is 80 kvar, the capacitor supplies tho difference, or 80 - 38.7 = 21.3 kvar.

FIG. 8.6

Diagram for Example 3.

CONVENIENT CALCULATION METHODS

FOR


The calculating method described previously mas primarily intended t o show how kilorars influence the power factor and that in a-c circuits the total kva is obtained by using the right-triangle relationship and not just by arithmptical addition of the kilowatts and kilovars. It is evident from these ralrulations that the right-triangle method is rather laborious for pover-factor calculations. From the right-triangle relationship several simple and useful mathematical cxpressions may he written:

kw

cos 0 = pf = kva kvar tan 0 = ~kw kvar sin 0 = kva Because the kilowatt component usually remains constant (the kva and

44v

POWER4ACTOR IMPROVEMENT

kvar components change with power factor), Eq. (8.8) involving the kilowatt component is the most convenient to use. This expression may be rewritten as kvar = kw X tan e (8.10) For example, assume that it is necessary to determine the capacitor rating to improve the load power factor.

kvar a t original pf kvar a t improved pf

= =

kw X tan 8 , km X tan O2

Therefore, the capacitor rating required to improve the power factor is

- tan

(8.11)

ckvar*

=

kw X (tan 8 ,

For simplification (tan O1

-

tan Sz) is often written asA tan. Therefore,

ckvar

=

kw X A tan

02)

(8.12)

All tables, charts, and curves which have a “kw mu1tiplier”for determining the capacitor on synchronous motor kilovars are based on the above expression. Table 8.2 lists the “kw-multiplier” values for a wide range of operating conditions; various trigonometric functions useful in power-factor applications are included in the Appendix. Example 4. Determine the capacitor rating for Example 3 by using Table 8.2. The “kw multiplier” or A tan as read from the table is 0.266. Substituting in Eq. (8.12), ekvar

=

80(0.266) = 21.3

KllOVAR GENERATORS

Capacitors. The concept of a capacitor as a kilovar generator is helpful in understanding its use for power-factor improvement. A capacitor may be considered a kilovar ‘generator because it supplies the magnetizing requirements (kilovars) of induction devicen. This action may be explained i d terms of the stored energy. When a capacitor and an induction device are installed in the same circuit, there will he an exchange of magnetizing current between them, i.e., the leading current taken by the capacitor neutralizes the lagging current taken by the induction device. Because the capacitor relieves the supply line of supplying magnetizing current to the induction device, the capacitor may be considered to be a kilovar generator, since i t actually supplies the magnetizing rkquirements of the induction device.

The prefix “ c ” in ckvar is used to designate the capacitor kvar in order to differentiate it from load kvar.

Desired improved power factor, COI 92 Originol power factor cos 8 ,

~~

0.80 0.81 0.82 0.83 0.84 0.85 0.86 0.87 0.88 0.89 0.90 0.91 0.92 0.93 0.94 0.95 0.96 0.97 0.98 0.99

1.0

0.50

0.982 1.0081.034 1.0601.0861.112 1.1391.165 1.192 1.220 1.248 1.2761.3061.3371.3691.403 1.440 1.481 1.529 I 589 1.732

0.51 0.52 0.53 0.54 0.55

0.9370.9620.989 1.015 1.041 1.0671.0941.120 l.1471.1751.2031.231 1.261 1.2921.3241.358 0.8930.9190.9450.971 0.9971.023 1.050 1.0761.1031.131 1.1591.1871.217 1.2481.280 1.314 0.8500.8760.9020.9280.9540.9801.0071.033 1.0601.0881.116 1.1441.1741.2051.237 1.271 0.8090.8350.861 0.8870.9130.9390.9660.992 1.019 1.0471.075 1.1031.133 1.164 1.196 1.230 0.7690.7950.821 0.8470.8730.8990.9260.9520.979 1.007 1.035 1.063 1.093 1.124 1.156 1.190

0.56 0.57 0.58 0.59 0.60

0.7300.7560.7820.8080.8340.8600.8870.9130.9400.9680.9961.024 1.054 1,085 1.1171.151 1.188 1.229 1.2771.337 1.480 j: 0.6920.7180.7440.7700.7960.8220.8490.8750.9020.9300.9580.9861.0161.0471.0791.1131.1501.191 1.2391.2991.442 2 0.6550.6810.7070.7330.7590.7850.8120.8380.8650.8930.921 0.9490.979 1.0101.042 1.0761.1131.lS41.202 1.262 1.405 0 0.6190.6450.671 0.6970.723 0.749 0.7760.802 0.8290.8570.8850.913 0.943 0.974 1.006 1.040 1.077 1.118 1.166 1.226 1.369 ? 0.583 0.609 0.635 0.661 0.687 0.713 0.740 0.766 0.793 0.821 0.849 0.877 0.907 0.938 0.970 1.004 1.041 1.082 I. I30 I. I90 1.333

0.61 0.62 0.63 0.64 0.65

0.5490.575 0.601 0.6270.653 0.6790.70610.732 0.7590.7870.8l50.8430.8730.9040.9360.970 1.007 1.048 1.096 1.156 1.299 2 0.5160.5420.5680.5940.6200.6460.6730.6990.7260.7540.7820.8100.8400.8710.9030.9370.974 1.015 1.063 1.123 1.266 0.483 0.5090.535 0.561 0.5870.6130.6400.6660.6930.721 0.749 0.7770.8070.838 0.8700.904 0.941 0.982 1.030 1.090 1.233 f 0.451 0.4740.5030.5290.5550.581 0.6080.6340.6610.6890.7170.7450.7750.8060.8380.8720.9090.9500.998 1.058 1.201 0.4190.4450.4710.4970.5230.5490.5760.6020.6290.6570.6850.7130.7430.7740.8060.8400.8770.9180.9661.0261.169

0.66 0.67 0.68 0.69 0.70

0.388 0.414 0.440 0.466 0.492 0.518 0.545 0.571 0.598 0.626 0.654 0.682 0.712 0.743 0.775 0.809 0.846 0.8870.935 0.995 1.138 0.358 0.384 0.410 0.436 0.462 0.488 0.515 0.541 0,568 0.596 0.624 0.652 0,682 0.713 0.745 0.779 0.816 0.8570.905 0.965 1. 108 0.3280.3540.3800.4060.4320.4580.4850.511 0.5380.5660.5940.6220.6520.6830.7150.7490.7860.8270.8750.935 1.078 0.299 0.325 0.351 0.377 0.403 0.429 0.456 0.482 0.509 0.537 0,565 0.593 0.623 0.654 0.686 0.720 0.7570.798 0.846 0.906 1.049 0.270 0.296 0.322 0.348 0.374 0.400 0.427 0.453 0.480 0,508 0.536 0.564 0.594 0.625 0.6570.691 0.728 0.769 0.817 0.877 1.020

0.71 0.72 0.73

0.242 0.268 0.294 0.320 0.346 0.372 0.399 0.425 0.452 0.480 0,5080.536 0.566 0.597 0.629 0.663 0.700 0.74 I 0.789 0.849 0.992 0.21 4 0.240 0.266 0.292 0.3 I8 0.344 0.37 I 0.397 0.424 0.452 0.480 0,508 0,5380.569 0.60 I 0.635 0.672 0.71 3 0.761 0.82 I 0.964 0,186 0.21 2 0.238 0.264 0.290 0.31 6 0.343 0.369 0.396 0.424 0.452 0.480 0,510 0.541 0.S73 0.607 0.644 0.685 0.733 0.793 0.936

I'

1.395 1.4361.484 1.5441.687 1.351 1.392 1.4401.5001.643 1.308 1.3491.3971.4571.600 1.2671.308 1.3561.4161.559 1.227 1.268 1.3161.3761.519

5

<

F

5

3 2

I

'

I

'

I

0.74 0.75 0.76 0.77 0.78 0.79 0.80 0.81 0.82

0.83 0.84 0.85

0.159 0.185 0.211 0.2370.263 0.289 0.316 0.342 0.369 0.397 0.425 0.453 0.483 0.514 0.546 0.580 0.617 0.658 0.706/0.7660.909 0.1320.1580.1840.2100.2360.2620.2890.3150.3420.3700.3980.4260.4560.4870.5190.5530.5900.6310.6790.7390.882 0.1050.1310.1570.1830.2090.2350.2620.2880.3150.3430.3710.3990.4290.4600.4920.5260.5630.6040.6520.7120.855

0.0790.1050.131 0.1570.1830.2090.2360.2620.2890.3170.3450.3730.4030.4340.4660.5000.5370.5780.6260.6860.829 0.052,0.0780.1040.1300.1560.1820.2090.2350.2620.2900.3180.3460.3760.4070.4390.4730.5100.5510.5990.6590.802 0.0260.0520.0780.1040.1300.1560.1830.2090.2360.2640.2920.3200.3500.381 0.4130.4470.4840.5250.5730.6330.776 0.0000.0260.0520.0780.1040.1300.1570.1830.2100.2380.2660.2940.3240.3550.3870.421 0.4580.4990.5470.6090 750

..... 0.000 0.026 0.052 0.078 0.104 0.131 0.1570.184 0.212 0.240 0.268 0.298 0.329 0.361 0.395 0.432 0.473 0.521 0.58110:724 ..... .....0.0000.0260.0520.0780.1050.1310.1580.1860.2140.2420.2720.3030.3350.3690.4060.4470.4950.5550.698 ..... ..... .....0.0000.0260.0520.0790.1050.1320.1600.1~~0.2160.2460.2770.3090.3430.3800.421 0.4690.5290.672 ..... ..... ..... ..... 0.0000.0260.0530.0790.1060.1340.1620.1900.2200.251 0.2830.3170.3540.3950.4430.5030.646 ..... . .... ..... ..... ..... 0.000 0.027 0.053 0.080 0,108 0.136 0.164 0.194 0.225 0.257 0.291 0.328 0.369 0.417 0.477 0.620 05 ..... ..... ..... ..... ..... ..... 0,000 0.026 0.053 0.081 0. I09 0. I37 0. I67 0.I98 0.230 0.264 0.301 0.342 0.390 0.450 0.593 ..... ..... ..... ..... .......... ..... 0.0000.0270.0550.0830.1110.I4l0.1720.2040.2380.2750.3160.3640.4240.567 & ..... ..... ..... ..... ..... ..... ..... .....0.0000.0280.0560.0840.1140.1450.1770.2110.2480.2890.3370.3970.540 5 ..... ..... ..... ..... ..... ..... ..... ..... .....0.0000.0280.0560.0860.1170.1490.1830.2200.261 0.3090.3690.512 m ..... ..... ..... ..... ..... ..... ..... ..... ..... .....0.0000.0280.0580.0890.121 0.1550.1920.2330.281 0.341 0.484 5 ..... ..... , ..., ..... ..... , .... ..... , .... ..... ..... ..... 0,0000.030 0.061 0.093 0.127/0.1640.205 0.253 0.313 0.456 0 ..... ..... ...., . .... , .... ..... ..... . .... ..... ..... ..... ..... 0.000 0.031 0.063 0.097 0.I34 0.I75 0.223 0.283 0.426 3 ..... ..... ..... ............... ..... ..... ..... ..... ..... ..... ..... 0.0000.0320.0660.103,0.1440.I920.2520.395 9 ..... _......... _......... ..... ..... ..... ..... ..... ..... ..... ..... ..... 0.0000.0340.0710.1120.1600.2200.363 ..... ..... ...., ..... ..... ..... ..... ..... ..... ..... _......... .._....... ..... 0.0000.0370.0790.1260.1860.329 ..... ..... ...., ..... ., ... , ., .. ..... ..... ..... .., .. ..... ..... ..... ..... ..., , ..... 0,0000.041 0.089 0.149 0.292 0.000~0.048~0.1080251 ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... _......... ..... 0.0000.06~0:203 ..... ..... ..... ............... ............... ..... ..... ..... ..... .....,.......... ..... .....0.0000.143 ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... .......... ~

m

0.86 0.87 0.88

0.89 0.90

io

0.91 0.92 0.93 0.94 0.95 0.96 0.97 0.98 0.99

+

1 1

ckvar = kw X multiplicr = 5on x 0.691 = 345.5

10.000

IS?

POWER-FACTOR IMFROVEMENT

Synchronous Motors and Synchronous Condensers. Synchronous motors and synchronous condensers may also act as kilovar generators. They generate kilovars in the same manner as a conventional generator does. Their ability to generate kilovars is a function of excitation and, in the case of synchronous motors, it is also a functioii of load. When underexcited, they do not generate sufficient kilovars to supply their own needs and consequently must take additional kilovars from the system. When overexcited (normal operation), they can supply all their own kilovar requirements and in addition can supply kilovars to the system. Thus, they may be considered as kilovar generators. Synchronous motors are widely used for power-factor improvement. The kilovar output that t,hey are capable of supplying to the line is a function of excitation and motor load. The curves of Fig. 8.7 show the

SYNCHRONOUS MOTOR LOAD

FIG. 8.7 excitation.

Curves showing approximate kilovan supplied by synchronous motors with rated


453

kilovars that a synchronous motor is capable of delivering under various load ronditions with normal excitation. A t high overloads (not shown on these curves) a synchronous motor may take magnetizing current from the line. The two powr-factor ratings of synchronous motors most commonly used in industry are unity power factor and 0.8 power factor. These ratings refer t o t,he operating power factor at full load and with normal field excitat,ion. In the case of the 0.8 power-factor motor, this always means 0.8 power-factor leading. Synchmnous condensers are rarely economical for industrial plants; so no further reference will be made to them. INSTRUMENTS AND MEASUREMENTS FOR POWER-FACTOR STUDIES

When power-fartor studies are made, it is essential that sufficient and useful data be availahle or taken in order t o select the proper value and loratioti of capacitors. If the study is for rate purposcs, t,hen the power bills usually furnish sufficient information to determine the kilovars rcquired. Most rates are based on a “billing” demand which is determined from the actual dcmaiid and power factor. Actual kilowatt demand is usually obtained from a demand register attachment on the watthour meter or by rerording- or prititing-type instruments. Power fartor may he measured directly or obtained from other indications such as from kilowatt, kilovolt-ampere, or kilovar values or from kilon-atthours and kilovar-hours. If power factor is measured directly, it is almost, always measured at the time of average or normal conditions. Measurements by recording or graphic instruments are most desirable and useful because t,hey provide a permanent record. Indicating instruments are satisfactory for spot checking, such as for indiridual feeder circuit,s or loads. They can be used also to good adrantage in place of rerording instruments if readings are taken at frequent intervals. The preferred measurements are kilo&atts, kilovars, and volts; from these the kva and pover factor can be calculated. Vokage readings are especially desirable if automatic capacitor control with a vohge-responsive master elemcnt is contemplated. There are many t,ypes of meters and instruments available for pomerfactor studies. Generally, portable devices are preferred because of their convenience. Accuracy, while desirable, is not as important as portability, and accuracy within a few per cent is generally acceptable for the purpose.

454


There is now availahle for operatiori on circuits up to ti00 volts the hook-on t,ype of ammeter, voltmeter, wattmeter and varmeter, and powerfactor meter. These instruments provide a convenient means for obtaiuing the data normally required for plant studies; their main advantage is that the circuit does not have to be broken t o measure t,he current component-the instrument is merely hooked over the conduct,or. Figure 8.8 is useful for obtaining the power factor wheri wattmeter readings are available. Fig. 8.9 is handy when kw and kvar or kmhr and kyar-hr values are known; this latter method is frequently used by utilities to obtain t,he power factor of small industrial plank where the

R VALUES

FIG. 8.8

-

R,V' W, Curve for obtaining power factor from ratio o f wattmeter readings (applicable

only for balonced loodl.



456

more expensive types of power-factor measuremeirt are not economically jnstilird. The power factor can be read directly from Fig. 8.9 by suhstituting kwhr for kw and kvar-hr for kvar values. LOCATION OF CAPACITORS AND SYNCHRONOUS MOTORS All the benefits that shunt capacitors and synchronous motors provide stem from the redurtion of kilovars. This is true of power-bill savings, releasr of system capacity, voltage improvement, and reduction of losses. 3Iauimum henefit is obtained when they are located at the load. The concept of a kilovar generator as described previously is particularly helpful in understanding this point. Whenever possible, capacitors should be located at or near the load on 180- or GOO-volt systems III order t o obtain the minimum cost and maximum benefits. These locations are shown in Fig. 8.10. The most effcctive locatioii is a t the load, as shown by C1, next ‘22, etc. In the case of synrhronous motors the same freedom of electrical location is not always practical or economical. Vsually synchronous motors INC SUPPLY

c3

&(

DISTRIBUTED LOADS 0

I

c2

.

c2

FIG. 8.10 Electrical location of shunt capacitors on indurtriol power systems. copacilori 01 load, such (IS CI or C2, for maximum over-all benefits.

Locate


457

arc of larger horsepon.cr rating than is ecooomical for operation at 2-10 or 480 volts, the common utilization voltages of industrial plants. 1 1 0 ~ cver, the samr principle applies, i.c., connection t o t,hr load bus whose power fact,or is to be impro\wl. Typical iris(al1:rtions of capacitors and synchronous motors arc shown in Figs. 8.11 to 8.13.

FIG. 8.1 1

Installation showing capacitor locoted near the motor ond connected (15 shown

in C1 of Fig. 8.10,

FIG. 8.12 Installation rhowing copocitor Flex-A-Plug and connected as shown in c2 of Fig. 8.10.

ECONOMIC CONSlDERATlONS

Although maximum o v e r 4 operating bendita are obtained when eapacitom are located at the load, it is not always practical or economical to locate capacitors at each load. M& industrisl plenta contain a number of small loads;Since capacitors are made in etandard sises it would be imprsetical to apply the m m t capacitor Hovara at each load. Then, too, in the genersl caae all thwe I d sre not on sll the time, EO it is poesible to take advantage of the divemity by instslling a single capacitor equipment at ~ o m ecentral loek tion. For example, if only 60per cent of the total motor load is in o p e n tion at one time, then a group capacitor need be only half the kilovar ra&g of the total number of kilomra connected at individual loada T h e syetem operating voltage infIuenced the economic Considerations aesoeisted with location of apaeitora and motora For example, 23O-vdt capscitor equipmta c& more than twice lls much es 460-or 575-volt equipmenta. Ah, economic c0m-m should include a enitable switching device, For example, although 24Wvdt capacitor units are the most economical, yet theae equipmenta with the proper switching


459

device usually cost more than 460- or 575-volt equipments for prartically all industrial applirations because of the higher cost of siritrhing devires for 2400-volt service. ADVANTAGES OF IMPROVED POWER FACTOR Improved power factor may give economir or system advantages or both. The largest economic advantages are usually obtained where power rates include a monetary incentive for improved power factor. This advantage ran be readily determined by ralculating power costs a t various plant power factors. Power rates are too varied to permit more than mention of that point here. The system advantages of improved power factor usually are (1) released system capacity, (2) reduction of power system losses, and (3) improved voltage conditions.

POWER-BILLING SAVINGS

The main use of capacitors in industrial plants and often a deciding factor in the selection of synchronous motors is to reduce purchased power costs when the rate contains a power-factor clause or its equivalent. Generally, the return on these investments is many, many times the return obtained from straight business investments. It is common for capacitors to pay for themselves in 56 to 3 years. This represents an annual gross return of 200 to 3355 per cent. The estimated mean return is in the neighborhood of 65 per cent. This figure is based on a study of the power-factor clauses of a number of utilities distributed throughout the country. The rate of return will depend upon the cost of capacitors, which in turn depends somewhat upon the voltage class. However, the above figures are applicable to most industrial installations, the major exception being 230-volt installations, for which the return will be about half of the above values. The amount of power-factor improvement depends upon the original power factor and the type of rate structure, and each case should he studied individually. However, it is generally economical to improve the power factorJo take advantage of the full amount of the penalty and bonus. A rough rule that has been used is improvement to 90 to 95 per cent. The best way t o determine the capacitor kilovars t o use is t o calculaGe the rate of return and actual dollar savings for various final power-factor values.

460


Practically every size of plant having the usual type of power-fartor rate structure cau justify capacitors. I t is the rafe of retzcrn 011 the capacitor investment which is the important factor. Actually, the rate of return will usually be greater for small plauts herause of the higher demand and energy charges for small loads. RELEASE O F POWER-SYSTEM CAPACITY BY POWER-FACTOR IMPROVEMENT

What is meant by the expression “release of system capacity”? When capacitors or synchronous motors are in operation in a plant, they deliver kilovars, furnishing magnrtizing current for motors, transformers, etc., thus reducing the current from the power supply. Less current means less kva or load on transformers and main branch feeder circuits. This means capacitors and synchronous motors can be used to reduce overloading of existing facilities or, if the equipment is not overloaded, permit additional load to he added. Release of system capacity by power-factor improvement-and especially with capacitors-is becoming more important as plant engineers appreciate their economic advantages. World War I1 demonstrated this quite clearly; many of the new plants incorporated capacitors as part of the poner-system layout, and existing plants found they were a quick and economical means of caring for increased loads. Thermally Overloaded Apparatus. Many cases arise where a cable circuit, transformer, or generator is overloaded (kva or amperes) because of ION power factor. The procedure for determining the value of kilovars required to reduce the load to a sperified value can best be explained by an example. Example 5. A 460-volt cable circuit is rated a t 235 amp but is carrying a load of 300 amp at 0.7 power factor. What kvar of capacitors is required to reduce the current to 235 amp? 1.73 X 460 X 300 = 240 kva = 1000 kw = 240 X 0.7 = 168 (8.5) The kva corresponding to 235 amp is 188; so the operating power factor corresponding to the new load is 168 cos e1 = K g = 0.895 say 0.90 (8.4) The capacitor kilovars required ckvar = 168 X 0.536 ckvar = 90 Kilowatt-limited Apparatus.

(8.12)

Some apparatus such as turbine-genera-


461

tor and engine-generator sets have a kilowatt limit of the prime mover as well as a kva limit of the generator. Usually the kilowatt limit corresponds to the generator kva rating, and the set is rated a t that kilowatt value a t unity power-factor operation. Intermediate kilowatt values, such as those between 0.8 and 1.0 power-factor operation, are determined by the power factor and kva rating a t the generator, so neither the kilowatt nor kva load exceeds the generator kva rating. Improvement of the power factor can release both kilowatt and kva capacity. Example 6. A 1000-kw turbine-generator set (turbine capability of 1250 kw) is already operating a t rated load, 1250 kva and 0.8 power factor. An additional load of 170 kw a t 0.85 power factor is to be added. What value of capacitors is required so that neither the turbine nor the generator will be overloaded? Original load : kw = 1000 kvar = 1000 X 0.75 kva = 1250

=

750

(8.10)

Additional load: kw = 170 kvar = 170 X 0.620 = 105 170 kva = - = 200 0.85 Total load: kw kvar

= =

1000 750

(8.10) (8.6)

+ 170 = 1170

+ 105 = 855

The minimum operating power factor for a load of 1170 km and not exceeding the kva rating of the generator is

cos e

1170 1250

= -=

0.935

The maximum load kvar for this condition is kvar = 1170 X 0.379 = 444

(8.10)

where 0.379 is the tangent corresponding to the power factor of cos 0 = 0.935. Therefore, the capacitors must furnish the difference between the total load kvar and the permissible generator kvar, or ckvar = 855

- 444

=

411

461

POWECFACTOR IMPROVEMENT

AMOUNT OF CAPACITY RELEASED BY POWER-FACTOR IMPROVEMENT

The foregoing showed how plant kva load could be reduced by powerfactor improvement. This will show how additional load can be added to a fully loaded circuit, transformer, distribution system, etc., by improving the power factor of the existing load, and also the magnitude of the additional load for a given improvement in power factor. Among power engineers this is termed release of capacity-thermal, electrical, or system capacity-and is usually abbreviated as T c or S,. The determination of the amount of system capacity released by porverfactor improvement is a somewhat arduous procedure; so several useful curves have been included to eliminate the need for calculations. Since any additional load added because of release of capacity by power-factor improvement may be at any power factor and different, from the original load power factor, there is 110 single convenient curve which will give the permissible load that can be added, although there is a rather involved equation which expresses this relationship. One approach, and a conservative one which also permits rather easy form of expression, is to assume that any additional load is at the same power factor as the original load. Figure 8.11 shows the basic diagram which applies to all the following expressions for T, where T , = capacity released, in terms of kva or per cent. load cos 0, = original power factor cos e2 = improved power factor of original load cos O3 = final power factor of combined load Since the total kva must not exceed the original load OR, the circle BB' will establish the limits. To satisfy this requirement, OE must equal OB

OC

+ Tc

=

OE

=

OB

+

where OC = origirial load O B capacitors OF or BC. Although the follo\ving data and figurrs ha1.e bren prepared specifically for caparitors, they a,lso apply t o synchronous motors for the same output of leading kilovars. These data are prcseiited in various forms, Figs. 8.15 to 8.17, for convenience. Figure 8.15, which includes an example for its use, shows the electrical capacity released for a specific value of capacitors. For estimating purposes and when there is not a large charige in powerfactor improvemerit, i.e., cos 0, to cos &, the followirig equation may be used: T C (in kva) = ckvar X sin 0, (8.13) Figure 8.15 is also valuable in that it may be used to shew the incremen-


463

tal gain in capacity released per kilovar of capacitors added. If, in the example in Fig. 8.13, t,he value of capacitors is doubled, the capacity released will increase from 28.5 t o 42.5 per cent, or a gain of 50 per cent. However, this 50 per cent gain in capacity requires a 100 per cent increase in capacitors. Even t,hough there is a gain in the total capacity released, the incremental gain may not necessarily be economically attractive. T, in Terms of Power-factor Improvement (cos 8 , to cos &). Often it is more convenient t o work in terms of the original and improved power factor rather than actual values of ckvar. In such cases Fig. 8.l(i, which also contains additional useful information, should be used F

A 8 : LOAD KVAR BC: OFzCKVAR

\

Basic load diogrorn for determining r e l e ~ r eof system electrical capacity by power-factor improvement.

FIG. 8.14

10

20

30

40

SO 60

70

SO

90

CAPACITOR W A R I N PERCENTAGE OF INITIAL KVA

100

I10

120

130

140

FIG. 8.15 Per cent electrical capacity released by capacitors and the approximate Rnol power factor (COI of the combined load.

081

IF A PLANT HAS A LOAD OF 1000 KVA 4 T 70 PERCENT POWER FACTOR AND 480 KVAR OF CAPACITORS ARE ADDED. THE SYSTEM ELECTRlC CAPACITY RELEASED 15 APPROXIMATELY 28.5 PERCENT: T H A T IS. THE SYSTEM CAN CARRY 2 8 . 5 PERCENT MORE LOAD (AT 10 PERCENT POWER FACTOR1 WITHOUT EXEEDING THE KVA BEFORE THE POWER FACTOR WAS IMPROVED T H E F I N A L POWER FACTOR ICOS 831 OF THE ORIGINAL LOAD P L U S THE ADDITIONAL LOAD I S APPROXIMATELY 90 PERCENT

EXAMPLE

0


465

=E IF THE LOAD POWER FACTOR (GO5 e,1 OF A SUBSTATION IS IMPROVE0 FROM 70 TO 95 PERCENT, THE SVSTEM ELECTRICAL CAP&CITY RELEASED IS 28.5 PERCENT; THAT IS, THE SYSTEM CAN CARRY 28.5 PERCENT MORE LOAD IAT 70 PERCENT POWER FACTOR)

WITHOUT EXCEEDING THE KVA BEFORE THE POWER FACTOR WAS IMPROVED. THESE CURVES ALSO SnOW THAT IT IS MORE ECONOMICAL TO OBTAIN THIS AMOUNT DF SVSTEM CAPACITV 81 POWER-FACTOR WPROVEMENT WITH CAP4CITORS THAN BY ~ D D I T I O N A LSUBSTATION AND DISTRIBUTION FACILITIES FOR A L L SIC VALUES GREATER THqN 1.7 IOBTAINED BV INTEPPOLATION. THE FINAL POWER FACTOR (GOSe,I OF THE ORIGINAL LOAO PLUS THE ADDITIONAI. LOAD IS 90 PERCENT.

FIG. 8.16 Per cent electrical capaciv released by improvement of load power factor (COI81 to COI BJ and the relative economics of capacitors for increasing the electrical capacity.

EXI\*PLL IF THE ORIGIN#.L LOAD POWER F4CTOR ICOS S,I 15 70 PERCENT b,ND THE FIN4L POWER FACTOR lCOS 93) OF THE COMBINED LOAD (ORIGINbL LOLID PLUS THE ADDITIONAL LOAD ALSO 4 1 70 PERCENT POWER FACTOR1 15 90 PERCENT, THE SYSTEM ELECTRICAL CAPACITY RELEbSED IS 2 8 5 PERCENT, THAT 15,TWF SYSTEM GAN CARRY 2 8 5 PERCENT MORE LOAD (AT 70 PERCENT POVER FACTOR1 WlTHOUT EXCEEDING THE K W BEFORE THE POWER FACTOR WAS IMPROVED THESE CURVES ALSO SHOW TH41 IT 15 MORE ECONOMICAL TO OBT4IN THIS AMOUNT OF SYSTEM CAPACITY BV POWER-F4CTOR IMPROVEMENT WITH CAPaCITORS TH4N BY 4DDITION4L SUBSTATION by0 OlSTRlBUTlON F4CILITIES FOR ALL S/C WLUES GREATER THAN.17 (OBTAINED BV INTERPOL4TlONL

FIG. 8.17 lcos 8 , to capacity.

Per cent electrical capacity released by improvement of load power fador BJ and the relative economics of capacitors for increasing the electrical

COI


467

Example 7. (Original load power factor improved to a specified value. See Fig. 8.16.)

Substation rating = 500 kva and operated a t full load Original power factor = 70 per cent (cos 8 , = 0.70) Average kilon-att load = 0.7 X 500 = 350 km (8.5) System voltage = AGO volts Improved pover factor of original load = 95 per cent (cos = 0.95) From Fig. 8 . l F it is found that improving the power factor of the load from 70 to 95 per cent releases 28.5 per cent capacity; i.e., 28.5 per cent of 350 kw or 99.8 km load at the original 70 per cent power factor can be added without increasing the load on the 500-kva substation; or, in terms of kva. this is 0.285 X 500 = 142.5 kva. .oo j ~

AS.. -.=I

"

.!:: ! .:: , .. .. .. .. .. ..

:: : : : -:..: --

...... D E T E R M I N A T I O N OF CKVAR TO G I V E D E S I R E D I M P R O V E M E N T OF W O V E R A L L POWER

S"

SAPACITOR

K V A R I N T E R M S OF O R I G I N A L

KW L O A D

FIG. 8.18 Graph for determining the value of capacitors required for improvement in load power factor (COI 61to cos 6a). There doto are for use only with Fig. 8.17.

460


From Fig. 8.16 it will also be noted that. the resultant power factor (cos &), i.e., the over-all power factor of the original load and capacitors, plus the additional load, is 90 per cent. Thus, this curve gives both the per cent load that can be added (or capacity released) and the final or over-all operating power factor. Tc in Terms of Final Power Factor (cos 8 , to cos OS). I n those cases where it is desirable to know the capacity released for a specific final ouer-all power factor cos R3, use Fig. 8.17. The value of capacitors required to release T c for this condition must be obtained from Fig. 8.18: the required ckvar value cannot,be obtained from any of the expressions, tables, or curves given elsewhere in this chapter. Use of Figs. 8.16 and 8.17. The difference between Figs. 8.16 and 8.17 should be clearly understood for a proper interpretation of T c . For example, Fig. 8.16 is used when the power factor of the original load is improved to a definite value, cos 82. The resultant power factor cos B3 is of secondary importance. On the other hand, Fig. 8.17 is used when the final power factor of the combined load, i.e., original load plus the additional load, is required to be a definite value, cos 03,as is often true in connection with power-factor rate studies. The examples in Figs. 8.16 and 8.17 show that the per cent capacity released is the same; the reason is that in both cases the original and final power factors (cos 03)are also the same. These charts can be used interchangeably only when the original and final power factors are the same for each case. ECONOMICS O F CAPACITORS FOR RELEASING SYSTEM CAPACITY

System load current-carrying capacity can he increased by power-factor improvement or by additional substation and distribution facilities. Actually there are several other factors in favor of power-factor improvement which should be given consideration. I n addition to releasing capacity, improvement of the power factor reduces losses and raises the voltage. Then too, in the case of capacitors, their installation need not he so permanent as substation and distribution facilities; so they can he more easily moved as changes occur in system loading or arrangement. If the allied benefits are neglected, the comparison is dependent upon the cost relationship between new substation and distribution facilities (8)and the cost of capacitors ( C ) . I n order to get a true comparison, all costs should he on an installed basis. Where actual installed cost data of substations and capacitors are not available, Table 8.3 may serve as a guide.


TABLE 8.3

469

Installed Costs (1954)

IS1 System "oltoge

240 480 600

-

* Instslled cost

Substation and distribution facililie,. dollars per kvo*

40-50 30-40 25-35

(Cl Capacitors, dollars per k v m t

SIC

18-20 8-10 8-10

2-2.8 3-5 2.5-4.4 I

Id-ccntcr system ineludina primary and secondary switchpear. cable. and transi rncrs. (The load-center svstem 1s thc most economical method of distrihution.) ' t Installed cost of capacitors with a manually operated switching device. of a I

,ewer

These data are also applicable to synchronous motors; for thc C value use the cost of the leading kilovars output in dollars per kilovar. Once the ratio S / C has been obtained, it is a simple matter t o det,crmirle the relative economics since, in Figs. 8.16 and 8.17, lines of various S / C ratios have been plotted. The cost relationship applies t o anu case, whether it is for an entirely new substation or for addit,ional feeder circuits, etc., for an existing substation or plant. In all cases, the appropriate S costs should be used. I t should be specifically noted that the values to use for S arid Care their unit costs, i.e., dollars per kva and dollars per kvar, respectively, and not the total dollar cost; to use total dollars in Figs. 8.16 and 8.17 irill give erroneous answers. For all points under an S/C curve (or for all S/C values greater than that established by the power-factor points) it is more economical to release system capacity by power-factor improvement, with rapacitors than by expanding substation and distribution facilities. Example 8. Are capacitors economical for releasing the amount of system capacity stated in Example 71 Assume that new substation and distribution facilities cost $30 per kva installed and capacitors cost $10 per kvar installed. The S/C ratio is 30/10, or 3. From Fig. 8.16 i t will be found that t,he intersection of the power-factor points (cos = 0.70 and cos .Q2 = 0.95) lies undcr an S,'C value of 3. Therefore, the capacitor method is more economical than expanding substation and distribution facilities for handling this amount of load. This can be checked in actual dollars as follows:


470

The value of ckvar required t o improve the power factor from 70 t o 95 per cent is ckvar = kw X multiplier (8.12) =

350 X 0.691

=

242

Installed cost of capacitors = 242 X $10 = 52420 Substation capacity released = 142.5 kva Installed cost of substation and distribution facilities = 142.5 X $30 = $4275 I n this case, additional system capacity obtained by use of capacitors costs only 56 per cent as much as new substation anddistrihutionfacilities. Not only can capacitors be used economically t o release the capacity of existing facilities hut, where the load power factor is low, it is economical t o use capacitors in new installations. This fact was recognized during World War 11, and many plants incorporated capacitors as part of the power-system design. POWER-SYSTEM LOSSES

The reduction in electrical losses due t o power-factor improvement results in an annual gross return of as much as 15 per cent on the capacitor investment. Although the return from loss reduction alone is seldom sufficient to justify t h e installation of capacitors, it is an attractive additional benefit. I n most industrial plant power-distribution systems, the kilowatt ( P R ) losses vary from 2.5 t o 7.5 per cent of the load kilowatthours, depending upon hours of full-load and no-load plant operation, wire size, and length of main and branch feeder circuits. Capacitors are effective in reducing only that portion of the losses that is due t o the kilovar current. Losses are proportional t o current squared, and since current is reduced in direct proportion t o power-factor improvement, the losses are inversely proportional t o the square of the power factor. kw losses

cc

original pf improved pf

Loss reduction = 1 -

original pf

(8.14) (8.15)

The capacitors have losses, but they are relatively small-only onethird of 1 per rent of the kvar rating. Example 9. Determine the savings in losses due t o improved power factor for the following conditions:


471

Total kilowatthours per year = 150,000 Origiiral power factor = 0.75 Improvcd power fartor = 0.95 Assume losses as 5 per cent of the total kilowatt hours, or 7500 1,oss reduction

(E)*

=

1 - -

=

1 - 0.623

=

0.377 or 37.7 per cent

(8.15)

ltedurtioii i n losses = 0.377 X 7500 = 2828 kwhr Assuming a net realization of GO per rent, then the actual reduction in losses = 2828 X 0.60 = 1697 kwhr. VOLTAGE IMPROVEMENT* The disadvantages of low voltage are so well known that they will not be rcit,cratcd here exrept to state t h a t they are economic and operational liabilities. Akhough caparitors raise voltagelevels, it is rarely economical t o apply them in industrial plants for voltage improvement alone. Voltage improvement may therefore be regarded as an additional benefit of capacitors. HOW CAPACITORS RAISE VOLTAGE LEVEL

'l'hc follo\ving simplified expression is usually used to ralculate the voltage drop of a circuit: e

=

R I ros 8

X I sin 8

(8.16)

where thc ( f ) value is used for a lagging power factor and the ( - ) value for a leading power factor. When the total current, and power factor are known, the component rurrents may he 0t)taiired from the right-triangle relationship.

k w current

=

I,,,,,, cos 8

(8.17)

kvar current

=

I,,,,,, sin 8

(8.18)

Thus the above equat,ioir for voltagc drop may b e rewritten: e

=

IT'X (kw current)

X X (kvar current)

(8.19)

From this expression it is evident that kilorar rurrent operatcs only on * Also

S
Clrap. 4.

472

POWER-FACTOR IMPROVEMEN1

reartanre, and sinre raparitors reduce t,he kilovar current they reduce t,he voltage drop and by an amount equal t o the rapacitor rurrent times the reartance. Therefore, it is neressary only t o know the capacitor rating and system reartanre t,o predirt the voltage change due to caparitors. There is some misunderstanding regarding voltage replalion when shunt rapacitors are used. Capacitors do not redure the voltage regulation uiiless they are automatirally switched. The voltage regulation wit,h unswitched rapacitors is prartirally the same as without capacitors, hut the voltage leuel is raised. MAGNITUDE OF VOLTAGE RISE DUE TO CAPACITORS

There is generally a good deal of misapprehension regarding “high voltage” due to caparitors. Actually, in most cases where the voltage a t light load is high it will be found t h a t most of the voltage variation is in t,he incoming supply source and is not primarily due to the variation or regulation mithin the plant. As far a s the voltage rhangedue t o capacitors is concerned the ralculations are simple; generally Eq. (8.21) or (8.22) is acrurate enough for this purpose. The volt,age rise due t o capacitors in industrial plants with modern poiver-distrihotion systems and one t,ransformation is not very great and is rarely more than 4 or 5 per cant. The greatest gain in voltage improvement, will be i n plant-distribution circuits having high reactance and low system voltage, such as is true of 240-volt systems having long runs of open wire, and where the spacing between phase wires is large, such a s 0 t o 10 in. When the load-center method of distrihution is used for 480- and 600-volt systems, the voltage improvement will be small. WHY POWER-FACTOR IMPROVEMENT IS SO EFFECTIVE IN

REDUCING VOLTAGE DROP

From Eq. (8.19) above it is seen that voltage drop is made up of two parts, one part due t o the kilowatt current and resistanre, which is usually called the resistance component, and the other part due t o the kilovar current and reactance, which is usually called the reactance component. I n industrial power systems the reactance is usually much larger than the resistance and in the order of two t o four times for distribution circuits and five t o ten times for power transformers. It is evident, then, that a change in the kilovar current has a larger effect than the kilowatt current in reduring the total voltage drop. Also, since the resistance component is tixed for a given kilowatt load, and the rircuit reactance is fixed, the practical way t o reduce the voltage drop is by reducing the kilovar cur-


473

rent. The kilovar current is reduced when the power factor is improved. For example, if a circuit has a resistance of 1 per rent, a rcartatire of j per cent,, and au operating power factor of 0.70, t,he voltagc drop is 4.3 per cent. The major part of t,he drop is due to reactanre mid is 3.(i per cent. If the power factor is improved to 0.90, the voltage drop is 2.4 pcr cent; if improved to unity (no kilot-.ar current): the drop is niiiy 0.7 of 1 per cent. CALCULATION OF VOLTAGE RISE DUE T O CAPACITORS

There are many calrulating methods for determining the voltage rise due to capacitors (or voltage change due to stvitchiirg blocks of capacitors in and out of service), but the per cent method is perhaps the most practical and simple; occasionally data in terms of actual values of volts, amperes, and ohms are used. The following few expressions are gcnerally sufficiently accurate, for practical purposes, to determine the voltage change due to capacitors or the leading kilovars from a synchronous motor or condenser. ,i few formulas from other sections are repeated here for convenience. Formulas for Voltage Change

v, = ICX

(8.20)

%

vc =

ckvar X (% X,) base kva

(8.21)

%

vc =

ckvar X yo transformer reactance transformer kva rating

(8.22)

Other Useful Formulas

% ohms

=

% x*=

% xs

=

X X base kva 10 X (kv)z base kva X ( % X,) kva, base kva X 100 system short-circuit kva

Nomenclature

ckvar = capacitor rating, kvars

I,

=

capacitor current, amp

kv = line-to-line voltage, kv kva = kva selected as a base for calculations

(8.23) (8.24) (8.25)

474


kva,

=

kva rating of equipment, such as a t,ransformer, on its own base

I',

=

voltage change, volts (line-to-neutral)

% Vc

=

per cent volt,age change or rise due to capacitors

X

=

reactance, ohms

70X b = per cent reactance of equipment, such as that

of a transformer,

on selected common base

%X,

=

per cent react,ance of system. Actually, Eq. (8.25) gives per cent impedance, but reactance may be assumed equal to impedance in computing voltage rise for this purpose

%X,

=

pcr cent reactance of equipment such as that of a transformer on its own kva rating

Reactance data necessary for calculating the voltage rise or change in transformers, cable, open wire, or busway are included in Chap. 1. It is generally sufficiently accurate to assume that the transformer reactance is the same as the impedance for this type of approximation. Equation (8.22) is derived from Eq. (8.21) and is based on the assumption that the primary-system reactance is negligible as compared with the transformer reactance. The error introduced by this assumption is generally small. When the available short-circuit kva at the transformer primary is less than 100 times the transformer rating, the system reactance should be added to the transformer reactance. The following example illustrates the use of the simplified voltage-rise equation. Example 10. A 360-kvar bank of capacitors is connected t o the secondary of a 1000-kva transformer having an impedance of 5.5 per cent. The short-circuit duty of the primary system is 100,000 kva. What is the approximate per cent voltage rise due to the capacitors? (8.22)

SELECTION A N D APPLICATION OF M O T O R S A N D CAPACITORS INDUCTION-MOTOR CHARACTERISTICS

Power-factor improvement of induction-motor loads by means of shunt capacitors has been a common practice for many years. A recent development is the practice of connecting the capacitors directly at the motor

POWER-FACTOR IMPROVEMM

475

terminals in order to permit switching the capacitors and motor aa a unit. A typical installation is shown in Fig. 8.19. This unit arrangement is desirable because the capacitors are always on when the motor is in operation. The power factor of an induction motor is quite good a t full load, usually between 80 and 90 per cent, depending upon the motor speed and type of motor. At light loads, however, the power factor drops rapidly, as illustrated in Fig. 8.20. Generally, induction motors do not operate at fullload (often thedrive is “overmotored”), resulting in a low operating power factor. Even though the power factor of an induction motor varies materially from no load to full load, note that the motor kilovars are essentially constant. This characteristic makes theinduction motor a particularly attractive capacitor application; with a properly selected capacitor the operating power factor is excellent over the entire load range of the motor, aa shown in Fig. 8.20, generally 95 to 98 per cent at full load and higher a t partial loads. The capacitor rating for the motor data of Fig. 8.20 is 5 kvar, which also corresponds to the motor no-load magnetizing requirements. Therefore, since the capacitor supplies all the motor no-load magnetieing current, the no-load operating power factor is unity. The reaaon the power-factor curve with capacitors is so flat over the entire motor-load range is that the

;~~ FIG. 8.19

5k &

Installation of a dkvar 4W-volt three-phase capacitor Iacated on the motor starier. Capacitor is switched with the motor and el&ically connected as shown in (A) in Fig. 8.21.

476


net kvar is low and varies only a little with load. For example, the net kvar at no load is zero and at full load only 2.0: kvar, which compares with j and 7.6 kvar, respectively, for operation without capacitors. Therefore, since the kvar load with qapacitors is so small in respect to the kw load, the kiv and kva are almost eyual, which means that the operating power factor is almost unity. The addition of capacitors for power-factor improvement does not change the motor performance characteristics, as the operating speed and shaft output depend upon the motor load and applied voltage. 100

90

KW

KVAR KVA

80

16

70

14

60

12

0:

0

a LL

10

: 5 0

: P I-

40

8

30

6

20

4

10

2

0

0 -

z Y

0 0:

w n

0

-I 4

-I 2

3 -

4-

4 4

MOTOR LOAD

FIG. 8.20

Motor characteristics for a typical medium-rile and speed induction motor.


477

LOCATION OF MOTOR CAPACITORS

Capacitors may he connected t o each motor and switched with the motor, as shown in Fig. 8.21A or B , or capacitors may bepermanentlyconnected to the feeder circuit a t selected starters for convenience, as shown in Fig. 8.21C. The preferred and most advantageous electrical location from an over-all standpoint is that of Fig. 8.21.4 or B. In either easethecapacitor and motor are switched as a unit by the motor starter; so the capacitor is always in service when the motor is in operation. The connection of Fig. 8.21.4 may be used for new installations, as the motor overload relay can he selected, a t the time of purchase, on the basis of the reduced line current due to the capacitors. This connection also has the advantage that the short-circuit current is less hecause of the impedance of the overload relay. Figure 8.21B may be preferred for existing installations as no change i n the overload relay is required because the current through the overload relay is the motor current. (Reduhon of current due to capacitor stems from the electrical location hack through the power syskm; therefore the current through the overload relay is not reduced in this case.) The arrangement shown in Fig. 8.21C is used when capacitors are permanently connected to the system. Its main advantage is the elimination of a separate switching device for the capacitors. LIMITATION OF CAPACITOR AND MOTOR WHEN SWITCHED AS A UNIT

Capacitors have been applied to induction motors and switched with the motor as a unit with good results except in a few cases. Experience has shown that when difficulties are encountered it is because too large a capacitor bank is used or the capacitors have been applied on jogging or quick reversing service. The two factors which limit the value of capacitors to be switched with a motor are (1) overvoltage due to selfexcitation and (2) transient torques. These limitations apply when the capacitor is connecled to the load side of the motor starter, as shown i n Fig. 8.21A and B , and the capacitor and motor are switched as a unit. These limitations also apply to that type of induction motor which has an auxiliary winding in the stator (transformer action) for connection to an external capacitor. Overvoltage Due to Self-excitation. A capacitor can supply part or all of the motor magnetizing requirements. Thus, when the motor line switch is opened &nd the motor disconnected from its power source, the capacitor furnishes the motor magnetizing current and the motor will “self-excite,” i.e., act as a voltage generator. The magnitude of the

470

POWER-FACTOR IMPROVEMEM

generated voltage will depend upon the value of the capacitor current, motor speed, and type of motor. However, as a bench mark, the resulting overvoltage with capacitor values to improve the full-load motor power factor to unity will range from 35 to 75 per cent. Self-excitation can he measured by connecting a voltmeter across the motor terminals and reading the voltage after the motor-starter switch is disconnected from the power source. In the usual motor application, the motor slows down rapidly after the switch is opened; so the voltage rapidly decreases. A 15 to 20 per cent reduction in speed eliminates self-excitation sufficientlyto cause the voltage to collapse in a few seconds. However, in a few cases with highinertia loads the voltage of self-excitation has been sustained for several minutes. Transient Torques. Transient electrical torques of twenty times fnllload motor torque have been obtained in tests when too large a capacitor w&s switched with a motor. These transient torques may occur when a motor is transferred to the line from the reduced-voltage tap of an autotransformer-type starter or when a running motor is temporarily disconnected from its line and the starting switch is reclosed while the motor is still running and maintaining voltage a t its terminals by self-excitation due t o capacitors. To produce transient electrical torques, the motor voltage developed POWER SOURCE

CONTACTOR

RELAY CAPACITOR MOTOR

(A)

(8)

(C)

FIG. 8.21 Electrical location ot capacitors when used with induction motors for powerfactor improvement.


479

by self-excitation must be of appreciable magnitude and out of phase with the line voltage when the switch is closed. This torque is similar to that caused by connecting two synchronous generators which are out of phase.

SELECTION OF CAPACITOR RATING FOR INDUCTION MOTORS

Tables 8.4 and 8.5 list the suggested capacitor ratings for the two most popular types of polyphase induction motors when the capacitors and motor are switched as a unit and connected as shown in Fig. 8.21A or R. The capacitor values are based on actual tests for transient torques and overvoltages by one large motor manufacturer. In addition, electrical characteristics of motors of other manufacturers as obtained from puhlished data have heen correlated and a representative average used. The capacitor ratings are conservative and prevent overvoltages due to selfexcitation and limit transient torques to normal values. The data are also applicable to wound-rotor-type motors. The operating power factor of motors with the suggested capacitor ratings mill generally range from 95 to 98 per cent a t full load and 95 to 100 per cent a t partial loads. There will be exceptions, of course, because of the variation in motor electrical characteristics and the fact that capacitors are built in only certain ratings. The greatest effect will be a partial-load operation, especially if the size of capacitor available is less than the value listed in Tables 8.4 and 8.5; even so, the power-factor characteristic with load is relatively flat from 25 to 100 per cent load. For example, a 20-hp 1200-rpm motor with a capacitor rated 25 per cent less than listed in the table will have a power factor of 92 per cent a t half load. For motor ratings not listed in these tables the following conservative rule may he used. The capacitor current should not exceed the motor no-load magnetizing current. This is approximately the motor no-load current, which can usually be conveniently measured with a hook-on ammeter. The tabular data may he extrapolated for estimating purposes, provided that allowance is made for the fact that the per cent magnetizing current and therefore the capacitor current, decreases as the motor horsepower rating increases. All the capacitor values listed in these tables do not necessarily correspond to standard S E M A ratings. However, the trend is toward additional ratings beloy 10 kvar, and some manufacturers have a line of these ratings to match closely the requirements of motors in the 5- to 25-hp range. When capacitors are not available for the values listed, it is suggested that the closest lower capacitor rating he used.

POWER-PACTOR IMPROVEMEN1

480

TABLE 8.4

Suggested Maximum Capacitor Rating When a n Induction Motor and Capacitor Are Switched as a Unit

For Normal Starting Torque, Normal Starting Current, and NEMA Classification Design B Motors* Nominal motor speed in rpm and number of poles

Induction m0tW

horsepower rating

"0

-

- -

.-

'b K"W r, IR ,R% - -

1. 2 2. 3 4 5

14 I2 I1 10 9 9

6 7 9 I2 I4 17

9 8 8 8 8

100 125 150 200

22 27

250 300 350 400 450 500

50

3 5 7t 10

15 20 25 30

40 50 60 75

32, 10

57 55

ro 75

77

-

-

~

% 1R

-

~

K"W

% AR

-

~

2 3 4 5 6.5 7.5

27 25 22 21 18 16

thr

2.5 4 5.5 6.5 8 9

20 I17 5

7.5 9

I5

I1 I1 10 10 10

18

10

8 8 8 8

25

9 9 9 9

27 31.5 37.5 47.5

10 10 10 10

32.5 40 47.5

7

52.5

7 7 6 6 6

$0

8 8 8 8 8 8

57.5 65 75 85 92.5 97.5

9 9 9 9 9 9

70 80 87.5 95 100 107.5

3

I

4 5

0 0

6 7 9

8

14 16

0 9 9 9 8 8

8 8 8 8

21 26 30 37.5

8 8 8 8 8 8

45 52.5 60 65 67.5 72.5

II 13

30 35

12.5

57.5

rs

30 82.5

14 13

I2

- -

9 10 12 15 18 21

I5

11

14 13 12

12 15 19 22 26

11

10

-

% 4R

- -

I .5

5 3 2

- -

- - - _. :"or

600 I2

720 10

8

2 3 3.5 5 6.5

1.5 2 2.5

11

900

1200 6

1800 4

3600 2

60

35 32 30 27 23 21

~K"a,

3.5 4.5

6 7.5 9.5

I2

% 4R

41 37 34 31 27 25

14

27 32.5

23 22 20 I9 I9 18

13 13 12 12

40 47.5 52.5 65

17 16 15 14

I1 I1 10

77.5 87.5 95 I05

13 12

20 18 16 15 15

lo 9 9

14 16 20 24

110 115

11 I1 11 10

-- --

NOTE:For 50-cyclp-operation the following representative data may be used: 1. For standard 60-cyelc motors opwating a t 50 cycles: kvsr = 1.4 to 1.7 of the kvar values listed yo A I l = 1.35 to 1.8 OF the Yo A R values listed 2. For standard 50-cycle motors operating s t 50 cycles: kvnr = 1 . 1 to 1.4 of the kvar values listed % .4R = 1.05 to 1.4 of the '3 A R values listed . (The larger multipliers apply to motors having the higher speeds.) Far standard OO-cycle wound-rotor-type motors operating a t 60 cycles, the following representative data may be used: kvar = 1.1 of the kvar vslurs listed % A R = 1.05 of the % AIi values listed For explanation of *, t, and i,see footnotes to Table 8.5.


TABLE 8.5

481

Suggested Maximum Capacitor Rating When an Induction Motor and Capacitor Are Switched as a Unit

For High Starting Torque, l o w Starting Current, and N E M A Classification Design C Motors*

I

Nominol motor speed in rpm and number of poles

Induction-

1800 4

~

mOlOI

horsepower

%

W

3 5 7% 10

15 20 25 30 40 50 60 75 100

125 I50 200

.... ..

K"0,

2 2.5 3.5 4 5 5 6.5

15

7.5

I2 12 11 11

16

11 11

19 24

11 11 11 11

32.5 40 45 55

28 35 42.5 50

13

13 12 12

8 9 I2

600 12

~

10

-

~

2 2.5 3.5 5 6

13

720

1200 6

% AR

K"or

%

K"0,

AR

,"k

~

Kvor

- ~-

-

~

26 21 16 18

3 4 5 8 9.5

31 29 27 25 21 19

12 12 I1 11 11 11

I1 13 16 20 23 27.5

17 17 16 16 15 14

14 16 20 25 27.5 35

11

35 42.5 50 60

13 13 13 13

42.5 50 55

6

13 I4

II

I1 I1

-

% AR

4 5 6.5 8 10

12

40 36 34 32 28 25

1

4.5 6.5 8 9.5 13 17

,': ;: 22 30 19 19 19

55

51 46 41 39 36 35

33 32 30 29 27 25 24

* These data apply when a capacitor and motor are electrically connected RS shown in Fig. 8.21A or B. These data are representative for three-phase 60-eyele general-purpose open- or splashproof-type motors of 220-, 440-, 55&, or 2300-volt rating. The operating power factor, for capacitor ratings as listed, will range from 95 to 98 per cent a t full load and 95 to 100 per cent a t partial loads. t Kvar is the rating of the capacitors in kilovars. This v d u c is approximately equal to the motor no-load magnetizing kilovars. i % A R is the per cent rcduetion in line current due to espaeitars and is helpful for selecting the proper motor-overload relay. If a capacitor of lower kilovar rating is used, the actual percentage reduction in the line current ( % AR) will be approximately proportional to Actual capacitor rating kvsr value in tables -

The relay selection should be based o n t h e motor full-load name-plate current reduced by the % A R value.

482


THERMAL-OVERLOAD PROTECTION O F MOTORS

When capacitors are connected to the motor terminals, they supply some of the motor magnetizing requirements so that the actual line current is less than it would be without capacitors. If the capacitors are connected to the motor starter on the load side of the thermal-overload device, as shown in Fig. 8.21A or B , the overload relay will not provide proper protection to the motor if it is selected for the uncorreeted motor full.load current. The relay should be selected for a smaller current rating commensurate with the reduced line current due to the effect of the capacitors. The percentage of this current reduction is shown in Tables 8.4 and 8.5. For the low-speed motors this reduction in line current is rather Iarge, particnlarly for the motors of lower horsepower rating. If the motor rating is not included in these tables, the reduction in line current due to capacitors may be obtained by measuring the line current witb and without capacitors a t full motor load or by ealculation. PRECAUTIONS IN APPLYING BUS CAPACITORS FOR A GROUP OF MOTORS

Wben capacitors are connected to a bus serving a group of motors, their effect can be similar to that of switching a capacitor and motor as a unit since overvoltages and excessive transient torques can result. However, if the power factor of the bus is improved to about 95 per cent under fullload conditions, there will seldom be harmfnl overvoltages or transient torques. Overvoltages and excessive transient torques will seldom occur if Tables 8.4 and 8.5 are followed in applying capacitors to single motors or groups of motors where a11 motors remain connected to the same bus as the capacitors. However, special precautions should be taken when applying large banks of capacitors in systems with fast reclosing high inertia loads, or where a major portion of Lhe load may be switched off leaving a small group of induction motors on tbe same bus with a large capacitor bank. INDUCTION MOTORS AND CAPACITORS VERSUS SYNCHRONOUS MOTORS

Where power factor has been an importmt consideration in the past, the synchronons motor has often been selected on the basis that it will give “free” power-factor improvement. Actnally, it is often more economical to purchase an induction motor plus capacitors than asynchronous motor. Factors Affecting Selection of Synchronous Motors or Induction Motors and Capacitors. First cost is one factor, and often the deciding

factor, in the selection between a synchronous motor and an induction


483

motor with capacitors. In some cases the type of drive or inherent characteristics of the motors dictate the selection. Synchronous motors have the advantage that they are capable of supplying smoothly varying values of kilovars, and the number can be readily changed by adjusting the motor field rheostat. The net kilovar output increases as the load decreases. However, there is little difference in the net kilovar output for load changes between a synchronous motor with fixed and rated field excitation and an induction motor with capacitors selected for the same kilovar output as the synchronous motor at full load. From the standpoint of losses, the two are about equal. Synchronous motors are generally more efficient than induction motors, but when the exciter losses are included with the synchronous motor losses, the totals are about the same. The synchronous motor has the disadvantage that it must be in operation to produce its kilovars; in the induction motor and capacitor comhination, the capacitors do not have to be connected and switched with the motor but can be permanently connected to the power system. This can be important from the loss standpoint if the synchronous motor must be operated just to produce kilovars. Maintenance is apt to be higher on the synchronous motor with its exciter and more complicated control than on the induction motorcapacitor combination. To obtain the same power-factor improvement from the induction motor and capacitor arrangement that can be obtained from an 0.8- or unity-power-factor synchronous motor, more capacitors must be used than can safely be switched with the motor. The cost of an extra switch a t low voltage is small, but above 600 volts it becomes appreciable. Price Comparison of Induction M o t o r and Capacitors vs. Synchronous Motors. Initial cost is perhaps the most important guide in making the

selection between the induction motor with capacitors and the synchronous motor. Surely, it is the easiest to evaluate. Tables 8.6 and 8.7 show the motor ratings where the cost of an induction motor plus capacitors (including a separate switching device for the capacitors) and motor starter is less than an 0.8- or unity-power-factor synchronous motor, exciter, and starter. For these comparisons, the capacitor rating was selected on the basis that the induction motor and capacitor Combination will furnish the same amount of power-factor improvement a t full load as an 0.8- or unitypower-factor synchronous motor of equal horsepower rating. For 2300- or 4006volt service the synchronous-motor equipment costs less than the induction-motor equipment over the entire horsepower and speed range if a power circuit breaker is used for switching the capacitors. However, if the capacitors can he located on the 480- or 600-volt systems,

484


and that is where they mill give the maximum benefit, then the inductionmotor arrangement becomes attractive. In that case the motor ratings will he approximately the same as listed in Tables 8.6 and 8.7. There are many cases where the induction motor-capacitor method is economical for much higher motor-horsepower ratings than those indicated in Tables 8.6 and 8.7. Often it is necessary only to improve the power factor of the motor to around 95 per cent, as that is generally accepted as a high operating power factor. In such cases the capacitor kilovars required are considerably less than those required to equal the full-load kilovar output of a unity-polver-factor synchronous motor. The kilovars required to improve a load whose power factor is already high increase much faster than the improvement in power factor. For example, it requires just as many ckvar to improve a load power factor from 95 to 100 per cent as it does to improve a load power factor from 80 to 92 per cent or 85 to 96 per cent. The full-load operating power factor of induction motors for the capacitor values suggested in Tables 8.4 and 8.5 ranges from 95 to 98 per cent, and these values can be switched directly with the motors. Therefore, where an operating power factor in the order of 95 per cent is all that is required, the induction motor with capacitors will he au economical selection for motor horsepower ratings much higher than those listed in Tables 8.6 and 8.7. TABLE 8.6

Horsepower Ratings Where an Induction Motor and Capacitors Cost Less Than a 0.80-power-factor Synchronous Motor 440- and 550-volt 60-cycle Equipment (Full-voltage motor starter1 Motor Rating, Hp 1800 250 and less 1200 I 5 0 and less 900 150 m d less M)O 200 and loss

Motor Speed, Rpm

TABLE 8.7 Horsepower Ratings Where an Induction Motor and Capacitors Cost Less Than a Unity-power-factor Synchronous Motor 440- and 550-volt 60-cycle Equipment Motor rating, hp Motor speed.

rpm

Wilh full-

Wilh reduced.tarter

voltage ,toner

I I800 t 200 900 600

dtage

I 750 and less 350 and leu 300 and less 300 ond less

400 350 300 300

and and and and

lers

INS leu leis


485

SYNCHRONOUS MOTORS

Synchronous motors are available in a wide range of horsepower and speed for various types of drives and for the common utilization voltages. General-purpose synchronous motors are available in standard ratings a t 1.0 and 0.8 power factor. For driving a given mechanical load, the unity power-factor motor is the lower in price, the more efficient, and draws no lagging or leading current. The 0.8 power-factor motors operate a t leading power factor and are used t o improve the power factor of the power line t o which they are connected, as well as for delivering rated horsepower output. If full-load excitation is maintained while the motor operates a t part load, the motor-will operate a t a more leading power factor, giving more than rated power-factor improvement. Figure 8.7 shows the approximate amount of corrective leading kvar delivered by the synchronous motor a t any load, provided the excitation is maintained a t its rated fullload value. SELECTION OF CAPACITOR SWITCHING DEVICE AND CABLE SIZE I n low-voltage circuits, i.e., 600 volts and below, air circuit breakers or fused safety switches are used for manual switching and air circuit breakers or contactors for electrical switching of capacitors. In medium-voltage circuits, 2400 t o 13,800 volts, power circuit breakers are almost always used. AMPERE RATING

Circuit breakers, contactors, disconnecting devices (except safety switches), and all other current-carrying parts should have a current rating of a t least 135 per cent of the rated capacitor current; safety switches, either fusible or nonfusible, should have a current rating of not less than 165 per cent of the rated capacitor current according to NEMA Standards for Shunt Capacitors. When contactors are used and they are housed in an enclosure, the rating should be taken as 90 per cent of the open rating. The standard ampere rating for molded-case air circuit breakers is based on 25 C ambient. Tables 8.8 and 8.9 include the appropriate derating factor for 40 C ambient and for installation of the breaker in an enclosure. Table 8.8 lists the recommended minimum ampere rating of safety switches, contactors, and air circuit breakers for standard ratings of lowvoltage capacitor equipments.

equipment ,oti"g.

45 60

50 100 150 200

90

300

180 270 450 630

600

15

30

Molded-

Mognetic-

I000 I600 3000

100 150 300 300 600 900 1350 2500 2500

100 200

70 150

200

225

400 400

300 450

800

For 4MI-mlt Copocitors ~

~

~

30

50

60

I00

90

150 200 300 600

300

100 200 200

300

400

600

400 800

540

I000

900

1600

1260

3000

1350 2500 2500

120 180

360

100 150

900

70 150 225 300 450

~

For 575-volt Capacitors

__ 30 60 90 I20

I80 360 540 900 1260

I

50

90 I25 I75 250

500 800 I200 2000

300

300 600 900 . .. 2500 2500

1I

I

200 440 600

70 125 175 225 350

Table 8.9 lists the appropriate capacitor kilovar multiplier for selecting the sivitrhing device for nonstandard hank ratings or various comhinations of capacitor units.


487

I n high-voltage circuits, 2400 t o 13,800 volts, the breaker continuous ampere rating is rarely the determining factor in the breaker selection because at these voltages the line current per kilovar is low arid for the usual capacitor ratings the current is well below the breaker ampere rating. TABLE 8.9

Approximate Capacitor Kvor Multipliers to Obtain Ampere Rating of Switching Device Three-phase Service [Enclosed Rating end 40 C

1

1 FI Ambientl copocimr "oltoge

Switching device

230

460

575

3.38 3.76 4.14 4.68

1.69 1.88 2.07 2.34

1.35 1.5 1.65 1.87

~~

.......... .......................... ..................... ...........

Magnetic-type circuit breoken. contoctor.. Safety switches.. Molded-case ~ i r ~ ubreakers. it

INTERRUPTING RATING

The switching device should also he selected for the short-circuit duty of the system on which it is to operate. Since a circuit breaker has a relatively high interrupting ability, it is well suited for such service. Most contactors have a limited interrupting ability; so short-circuit protection should be provided externally. REPETITIVE DUTY OF SWITCHING DEVICES

Repetitive duty of the switching device is seldom a factor unless switcbing is automatically controlled. Even then the number of switching operations which occur in practice is relatively small and rarely exceeds five to ten per day in industrial service. Contactors are designed for an exceedingly large number of operations before requiring maintenance of mechanical parts. The number of operations for which low-voltage air circuit breakers and oilless power circuit breakers are designed is not so great as for contactors hut is entirely satisfactory for this type of switching duty. SELECTION OF CABLE -SIZE FOR CAPACITORS

In selecting cables for capacitor applications, allowance must be made for the 35 per cent factor plus any additional factors, such as the decreased rating of a cable if it is operated at higher than rated 30 C (86 F) ambient.

488


Sinre most capacitors for industrial service are designed for use in an ambient of 40 C (104 F) maximum, the cables should also be selected for that ambient operation unless it is definitely known that the ambient temperatures are less. More care should be given in selecting the cable size for capacitors than for usual distribution feeder circuits because the load factor of a capacitor when energized is 100 per cent. H A R M O N I C S A N D RESONANCE AS AFFECTED BY CAPACITORS There are probably more imaginary difficulties associated with capacitors than with any other electrical equipment. Almost invariably these imagined difficulties concern harmonics and resonance. These fears are often based on the fact that theoretical calculations indicate such possibilities; anu circuit having inductance and capacitance has a resonance frequency, but in power systems such resonant combinations rarely occur. So many articles have been written on the theoretical aspects of the subject that they have contributed to fear, doubt, and misunderstanding of capacitors. Allied is capacitor fuse blowing, which invariably is blamed on harmonics and in many cases without any basis of fact. From a practical standpoint the subject can be dismissed; over 20 million kvar of capacitors are now installed in this country, and there have been relatively few cases of difficulty. If difficulties do arise, they can be overcome with knovn practical remedies. Capacitors do not generate harmonics, but they may reduce or increase harmonics, depending upon the particular circumstances. The major sources of harmonics are transformer-magnetizing current, rectifiers, arc furnaces, and generators. Harmonic voltages and current exist on all systems hut generally go unnoticed because of their small effect on operations. On rare occasions they may affect telephone communication service or cause capacitor fuses to blow. The reactance of a capacitor decreases as the frequency (harmonic) increases ( X , = 1O6/24' ohms). This means that for a given harmonic voltage Eh the harmonic current Ihincreases. However, in practical power applications the harmonic voltage decreases with increase in harmonic frequency; so from an over-all standpoint the current is small. Although any combination of capacitance and reactance has a resonant frequency, that fact is not important unless a harmonic voltage of that frequency and of sufficient magnitude is also present. A good practical rule for industrial applications is as follows: the kvar rating of the capacitor hank should not be greater than two-thirds of the transformer kva rating; if more capacitors are required, they should he automatically switched.


489

Capacitors have a large margin for harmonic currents and voltages. They are designed in accordance with KEMA requirements to carry 135 per cent of rated kilovars, including that of the fundamental and harmonics. This means that they can carry considerably more than 135 per cent of current, depending upon the magnitude of harmonic voltage. I t was recognized long ago that high harmonic currents were unusual; therefore, the NEMA standards pertaining t o the rating of switching devices, wire sizes, and other thermally rated devices prescribe their selection on the hasis of 135 per cent current rather than kilovars. Unless such a practical approach were taken, it would he necessary that each installation be checked for harmonic content before the switching device and wire sizes could he selected. In summary, it may he stated: i f the voltage i s approximately normal, i t i s practically impossible to overload a capacitor by harmonics.

AUTOMATIC SWITCHING OF CAPACITORS Automatic switching of shunt capacitors is seldom necessary in industrial plants, but when used is generally for one of the following reasons: 1. To reduce plant voltage or losses during light-load conditions 2. To control current or circuit loading 3. To meet the requirements of a rate clause 4. To comply with utility requirements 5. To prevent instability of generators during light-load conditions TYPES AND SELECTION OF CONTROL

Although many types of automatic controls are available, the ones most suitable for industrial spplications are Time clock Current, single-step Voltage, generally single-step Kilovar, generally multistep I n general, the master element should be responsive to the quantity to be controlled or regulated. For example, if capacitors are to be switched off a t light load, then a load-responsive master element would normally he selected. One exception to the above rule is the power-factor control, which is not recommended even though capacitors are used for power-factor improvement. The reason is that for constant power factor the load kilovars vary directly with the kilowatts, and such a control would have to be made inoperative when the load kilovars are equal t o capacitor kilovars of the largest step to avoid pumping action of the control equipment.


490

Table 8.10 iiidicates the preferred types of controls t o use. I n general, i t is desirable t h a t the master element obtain its signal from a point of measurement which includes the effect of the capacitor. One exception is the current control where the uncorrected load measurement must be used. The multistep voltage- and kilovar-responsive controls should always obtaiii their signal from the corrected load measurement. Table 8.10 and Fig. 8.22 indicate the point of signal measurement for various types of controls. SWITCHING FOR LIGHT-LOAD CONDITIONS

Most attention in automatic switching is directed t o the problem of overvoltage during light-load conditions in a plant. However, before purchasing automatic controls for this purpose, one should be certain t h a t the overvoltage is due t o capacitors. Ofteii most of the voltage variation is in the incoming supply source and is not due t o the voltage variation or regulation within the plant. Automatic switching of capacitors in such a case would be of little help. If there is any question regardiiig the magnitude of voltage rise due t o capacitors, t h a t can be easily checked by direct measurement or by calculation using the simple formulas (8.21) or ( 8 . 2 ) . If automatic switching is required, the simpler controls are preferred. The time-clock control is especialiy applicable where plant working hours are regular, e.g., capacitors are switched off at the end of the working day. In most industrial applications involving single-step capacitar switching, a current-responsive control usually results in better over-a11 operation than a voltage-responsive control because the operation of a current-respousive control is practically independent of system voltage

y

va (BUS)

SUPPLY .

"

.

r POINT OF SIGNAL

:

FEEDER CIRCUIT

ulu

T

Va

n

BUS

LOAO

CAPACITOR (0)

FIG. 8.22

ib)

Electrical location of capaciton ond point of signo1 meaiurement.


uuu 3 3 3

$ 2 5

" U " U 3

.: .: u= .. .. ; u . .uu3

U 3

U 3

,.

L

O

U

uu3

U

3

u;:

O

U

.. . ...

.. . ...

uu3

.. .. ..

uu

: U

492


fluctuations: also, the capacitors are switched off in accordance with the plant load, so a t light load the capacitors are off, which is desirable in reducing the voltage. A current-responsive control is also better than a voltage control when capacitors are located on adjacent feeders. In that rase, switching capacitors on one feeder affects the voltage level on the other feeder, thus requiring the O N and OFF settings of a voltage-responsive master element to he much wider than desirable. If a voltage-responsive control is used, the O N and O FF settings must be wide enough to avoid unnecessary operation due to voltage fluctuations in the supply source. Also, a voltage-responsive control may switch the capacitors independently of the actual load requirements and thus switch them off when they are needed most. Furthermore, in some plants, most of the voltage variation is in the incoming supply source and is not due to variation or regulation xvithin the plant. If there is any question regarding the magnitude of voltage rise due to capacitors, that can be easily checked by direct measurement or by the calculations. Single-step current and kilovar-responsive controls have the same general application. However, the current control is generally used because it costs less and requires only a current transformer for signal indication.

SERIES CAPACITORS

It is the purpose of this section t o familiarize the reader briefly with the series capacitor and its functions. Complete information on its applications, limitations, and operating benefits is available elsewhere. * The main applications of series capacitors have been to individual resistance welders, the main power supply for welder services, graphitizing furnaces, high-frequency generators, and to the main power system to reduce voltage variation for fluctuating or flicker-producing loads. WHAT IS A SERIES CAPACITOR?

The term series capacitor is an ahhreviation and is the engineer’s terminology for a series-connected capacitor. Similarly, a shunt capacitor is a shunt-connected capacitor. It is the type of connection and not the type of capacitor which is implied. These are shown in Fig. 8.23. Although the difference between the shunt and series capacitor is the manner in which they are connected in the circuit, they perform different functions. The shunt capacitor is simply a capacitive reactance in shunt

* Ser Chap. 13 in “Capacitors for Industry,” hy W. C. Bloamquist and R. C. Wilson, copyright General Electric Company, John W h y & Sons, Ine., New York. 1950.

(0)

LINE

SHUNT CAPACITORS

Ec

I

(Lower) Vector diagram

(b) SERIES CAPACITORS

9 ORIGINAL POWER FACTOR ANGLE 8' POWER FACTOR ANGLE WITH CAPACITORS

LOAD

FIG. 8.23 (Upper] Typical shunt and rerier connections of capacitors. showing how power-factor improvement is obtained.

LINE

I'

= I, = I

U'OAD

c w 0

7

5F

z

2

2

i

2?

B

494


with the load or system and is fundamentally for power-factor improvement. The benefits of improved voltage level, released system capacity, reduced system losses, and the reduction in the power bill all stem from the improvement in power factor. Unswitched shunt capacitors do not improve voltage regulation caused by load changes, but they do increase the normal voltage level. A series capacitor may be considered a nrgative (capacitive) reactance in series with the line. The voltage rise across the caparitor, a function of the circuit current, is automatic and practically instantaneous; so the series capacitor may be thought of as a voltage regulator. However, it differs from an induction or step voltage regulator in one very important respect-it cannot compensate for voltage variations originating in the supply source. A series capacitor at rated load provides power-factor improvement to the same degree as do the same kilovars of shunt capacitors. It provides power-factor improvement by an out-of-phase component of voltage, in contrast to the shunt capacitor, which accomplishes this by an out-ofphase component of current; these relationships are shown in Fig. 8.23. I n the usual application for power service, the series-capacitor kilovar rating is too low to improve the power factor significantly. WHAT THE SERIES CAPACITOR DOES

If a series capacitor is thought of as a negative reactance to neutralize the system reactance, rather than a device producing a voltage rise 180 degrees out of phase with the system reactive drop, its function in the system performance may he more readily understood. Principle of Operation. The voltage across a series capacitor is a function of its reactance and current, or

Ec

=

IXc

(8.26)

However, the effect of the series capacitor on the circuit voltage depends upon the power factor of the load current and is ec =

(8.27)

I X , sin 0

The basic voltage-drop formula, Eq. (8.16),for three-phase service may be rewritten to inrlude series capacitance as follows: e

=

I R cos 0

+ Z(X,

-

X , ) sin 0

where P = circuit roltage drop or change, volts ec = circuit voltage rise due to series capacitor E , = voltage across series capacitor only

(8.28)


I

=

495

total current, amps

0 = power-factor angle

X c = capacitive reactance, ohms X , = inductive reactance, ohms ( X , is positive and X C is negative in accordance with accepted terminology. Convention also defines the effect of a lagging power-factor current through a n inductive reactance as voltage drop.) Equation (8.28) indicates that the voltage drop and, therefore, the voltage regulation are reduced by reducing the reactive drop I X , which is usually the largest portiori of the voltage drop on industrial power systems. The effect of the negative reactance of a series capacitor on voltage regulation is illustrated hy the vector diagram of Fig. 8.24, using the voltage at the receiving end or load, Ex,as the reference. The voltage regulation of a system without a series capacitor is large, as indicated by the relative lengths of the voltage vectors En and E,. However, with a series capacitor of X , = X,, the capacitor completely neutralizes the system reactance; so the regulation is reduced considerahly as indicat,ed by vector E S I . This may be improved even further by overcompensation as indicated by E,,, which is the case for zero regulation. As far as the power-system operation is concerned, the less the voltage regulation, the “stiffer” the system is electrically, i.c., it is equivalent t o one with a higher iriterrupting duty. A system with zero voltage regulation is equivalent t o an infinitely large power system. I t is practical t o compensate for all the react,ancc i n many iirdustrial systems; so the voltage regulation is only that due t o the resistance component, vtrirh is usually only a few per cent. Thus, a series rapacit,or stiffens the system, which is especially beneficial for starting large motors from ail otherwise weak power system, for reducing light flicker caused by large fluctuating loads, ctc. Even though a series capacit,or stiffens the system, it docs not increase the interrupting duty herause of the control scheme associated ivith the series-capacitor protect,ive equipment. W H E N TO USE THE SERIES CAPACITOR

h good test for the application of series capacitors for line-rractance compensation is: “Would the problem be satisfactorily sol\wl by an induction or step regulator that could automatically and almost instantaneously correct the voltage drop for the new load rendition:"' If the answer is “yes,” theii it is very probable that a series rapacitor is applicable. (Voltage regulators arc riot applicahle for flui,tuatiiig loads siiice the voltage drop itself is used to initiate the voltage correction.) Because of the instantaneous aud automatic response of the serics


496

capacitor, it is admirably suited to compensate for voltage drop or fluctuation assoriated with intermittent, fluctuating, or suddenly applied loads, such as n-elders, arc furnaces, and motor-starting conditions; or fluctuating motor loads, such as sawmills, rubber mills, elevators, and shovels. The rlne to the applicahility of a series capacitor lies in an examination of the voltage-drop formula F =

I R cos 0

+ I ( X L - X , ) sin 8

and the relative values of the resistance I R and reactance [ X companents. I n most industrial power systems, the reactance is much higher than the resistance, with the ratio of X r / R ranging from 3 to 10. If the power fart,oris low (sin 0 is large), as is true for motor starting, then the I X drop will be the largest portion, and a series capacitor mill be helpful in reducing the voltage drop.

\

\

ES

\

1 ESI

Esz

I

9

I,

III

ES XsONLY

r;, X C ' X S ESp FIG. 8.24

xc ) xs

Diagram showing now o series copocitor reducer voltoge regulation.


FIG. 8.25

Voltage charts on

(I

system with a highly fluctuating rowmill motor lood.

497

Chart

B shows the effectiveness of the series capacitor for improving the voltage.

The series capacitor is effective when 1. The reactance is greater than the resistance. 2. The load power factor is low. The ideal application is that in which the capacitive reactance completely neutralizes the inductive reactance, leaving only the small resistance component. Thus, fluctuating loads, regardless of changes in the load power factor, will have very little effect on the voltage drop. The resistance and reactance of cables, lines, transformers, etc., given in Chap. 1, will he useful for preparing data for predicting the performance of a system with capacitors. Figure 8.25 shows actual voltage charts for a highly fluctuating motor load. Note how flat the voltage is with the series capacitor. The capacitive reactance can be selected greater than the inductive reactance and will thus compensate for part or all of the resistance drop a t a given power factor, as indicated in Fig. 8.24. However, if the power factor varies appreciably, overcompensation may cause a voltage rise a t


498

some load and power-factor conditions. Applications involving overcompensation should be carefully scrutinized to make certain that the over-all voltage pattern will be satisfactory. It is important to know something about the motor characteristics to predict the effect of a fluctuating motor load on the voltage drop, especially if the series capacitor does not compensate completely for the system reactance. The power factor of an induction motor a t no load ranges from 10 to 40 per cent and a t full load from 80 to 90 per cent, depending primarily upon the motor speed. A typical characteristic is shown in Fig. 8.20. However, the power factor changes very little in the motor overload region; e.g., for a 100 per cent overload the power factor may change only 5 points. Therefore, a motor load fluctuating from light to full load will have a much greater effect on voltage drop than a load changing from 100 to 200 per cent of rated load. SERIES-CAPACITOR LIMITATIONS

There are several limitations of series capacitors relating to circuit performance; so a word of caution is mentioned here. The prediction of circuit performance during transient conditions, i.e., loads suddenly applied or removed, such as motor starting and welder operation, is not so simple as for steady-state conditions as expressed in Eq. (8.28). The difficulty arises when the circuit constants become nonlinear. Although the performance can be predicted, a great many data are necessary, and even then prediction is usually difficult without the aid of a network or differential analyzer. Therefore, the practical solution for most applications of this type is based on experience. Some of the difficulties which may occur are self-excitation of induction and synchronous motors during starting, causing motors to lock into step a t suhsynchronous speed, vibrate excessively, or produce large current pulsations; hunting of synchronous motors during normal operation; and ferroresonance in transformers. Even though such abnormal operations can be eliminated, it is well that each series-capacitor application be thoroughly checked. CAPACITOR RATINGS AND OPERATING CHARACTERISTICS KILOVAR R A T I N G S

Standard ratings of individual shunt capacitor units are listed in Table 8.11 and of capacitor equipments in Table 8.12. Although the standards list only a limited number of ratings for low-voltage service, some manufacturers have additional ratings, usually in unit sizes from 1 t o 10 kvar,

499


to cover applications of small motors. There are no S E M A standards for two-phase caparitors, although manufacturers are often able to supply low-yoltage units in standard kilovar raliirgs. TABLE 8.11

NEMA Standard Ratings for 60-cycle Shunt Capacitor Units*

1

I

Kilovor ratings

Indoor enclosed ""it.

230 460 575 2,400 4,160 4,800 7,200 7.960 12,470 13,800

Phase

lndaor nonenclosed ""it.

.............. .............. .............. .............. ..............

.............. ..............

units

I

I

0.5. I , 2.5.5, 7.5 1, 2, 5, 10. 15 1. 2, 5, 10, 15

Outdoor

5, 7.5 10.15 10.15 15. 25 15.25 15.25 15, 25 IS, 25 15 15

I 2.5, 5, 7.5 5, 10, 15 5, 10, 15 15, 25 15, 25 15.25 15. 25 15, 25 15 15

Single, three Single, three Single, three Single, three Single, three Sngle,threet Single, threet Single

Single Single

* From KE41.4 Standard CA1-1949.

t Three-phase ratings outdoor only.

It will he noted from Table 8.11 that the standard capacitor voltages generally correspond t o the system operating voltages. The capacitor voltage ratings mere selected that way for very sound reasous. For example, take a nominal 480-volt system; motors for this service are rated 440 volts, supply transformers are usually rated 480 volts secondary a t no load, and generators are rated 480 volts. The actual plant operating voltage is less than the supply voltage by the voltage drop from the supply point to the load. Generally, industrial plant voltages under normal load conditions are about 460 volts, which correspond to the capacitor voltage rating for that service. The capacitor then operates a t or near rated voltage and output which contributes to long life. VOLTAGE

Because shunt capacitors are designed for operation a t comparatively high electrical stresses and continuous full load, definite overvoltage limits must be established. Most other electrical apparatus has an allowance for a duty cycle or load factor, taking advantage of partial-load or noncontinuous operation to allow for short-time overload or overvoltages. Capacitors are suitable for operation a t a terminal-to-terminal voltage

500

TABLE 8.12


Standard Kilovar Ratings of Indoor and Outdoor Types of Capacitor Equiprnents for 60-cycle Operation

1 46iLld 1

Copocitor d t a g e rating

230

2,400 and 4,160

4.800-13.800

~

Kilovor rating of units ~

7%

15

15 ~

15

30

30 45 60

60 90 120

90 180 270 450

180 360 540 900 1.260

-

~

25 ~

15

25

~

45'

630

90'

....

PO*

135. 180.

....

180'

.... .... .... ....

600 900

1,200 1.800 2,700 3,600

.... .... .... ....

900 1,800

2,700 3,600 4,500 5,400

* Outdoor equipments may be pole or base mounted. (including harmonics) a t a maximum of 110 per cent of rated voltage for either continuous or short-time operation, but exclusive of transients.

KlLOVAR MARGIN

From a practical standpoint it is necessary to allow some kilovar margin in capacitors for increased output due to operation above rated voltage, harmonic currents if present in the power system, and the plus kilovar tolerance in manufacture. This kilovar margin, an industry standard, is 35 per cent; i t is a rare application where this margin is not ample.

TIME OF DISCHARGE

When a capacitor is disconnected from its power source, it will retain its charge (voltage) and therefore be a hazard to personnel unless some


501

means is provided to discharge it. All modern capacitors have built-in discharge resistors, which usually discharge a unit in less time than is required by the National Electrical Code. The NEC requires capacitors to he discharged to a residual voltage of 50 volts or less in 1 min for capacitors rated 600 volts or less and in 5 min for capacitors rated more than 600 volts.

OPERATING CHARACTERISTICS

Voltage. While the operating voltage should he vithin the voltage limits discussed previously, it may not he the same as the name-plate voltage and the actual kilovar output will be different from the rated value as follows:

Actual ckvar = rated ckvar X

)'

operating vo~tagc rated voltage

(8.29)

Frequency. Capacitor kilovar output is directly proportional to the frequency of the applied voltage. Thus (for a given voltage)

Actual ckvar = rated ckvar X

operating frequency rated frequency

USEFUL DATA FOR CAPACITOR APPLICATIONS

Nomenclature

C Xc f kvar E kv I

kva

= =

= = =

= = =

capacity in microfarads reactance in ohms frequency in cycles per second kilovars, reactive kilovolt-amperes line-to-line voltage line-to-line voltage in kilovolts amperes kilovolt-amperes

Formulas

Capacitor connected in parallel:

c = c1+

c2

+ C3 +

(8.30)

502


Capacitor connected in series:

x, = 2653 -

at 60 cycles C (1 ,if = 2653 ohms) 1000(kv)2 kvar

xc =

106

c=-

(2d)xc C' = 1000 kvar 2rf( kv) * 2lrfC(kv) * kvar = 1000 1000 (kv)* kvar = kvar

=

=

kvar

X, G E X I (three phase) 1000

4X

(kv) X I

13I

= -(single =

1000 (kv) I

phase)

Capacitor Constants

Single-phase capacitors: Capacitor volts

Microfarads per k v d

230 460 575 2,400 4,160 4,800 7,200 7,960 12,470

50.15 12.54 8.025 0.4606 0.1534 0.1151 0.05118 0.041 87 0.01 706 0.01393

13.800

52.9 211.6 330.6 5.760 17.300 23,040 51,840 63.360 155.500 190,400

* To find microfarads for other kvar values m n l l i p l y by t h c number of kvar. t To find ohms for other kvar valucs divide hy t h e number of kvar.

POWER FACTOR IMPROVEMENT

503

Three-phase, Y-ronnerted caparitors: Line-to-neutral microfarads per kvar and line-to-neutral ohms for 1 kvar are same as single-phase values. Three-phase, delta-roiruerted capacitors: Line-to-line microfarads are one-third of the single-phase values; line-to-line ohms are three times the siugle-phase values.

Chapter 9

by Francis P. Brightman

Svstem Overcurrent Protection Fault-current (also designated overcurrent or short-circuit-current) protective devices-fuses, circuit breakers, and relays-are the watchmen of a power system, whose job it is to detect trouble and get rid of it as expeditiously aspossible. Thedesign of surh a protective system iuvolves two separate although iriterrelated steps: 1. Selection of the correct devices to do the job 2. Choice of correct current and time settings for the adjustable devices that \rill enable them to function selectively with other adjustable and nonadjustable devices to disconnect that portion of the system in trouble with as little disturbance to the rest of it as possible The two steps are interrelated in that the devices selected for a given system must be capahle of the required range of current and time settings needed, or inherently have the desired characteristics incorporated in their design. The objectives of this chapter are twofold. First, to describe the various types, rharacteristics, and principal uses of the fault-current protective devices commonly used ou industrial plant electric power systems as a guide to the reader in selecting suitable protective devices for his system. Second, to evplain how the time-current operating characteristics of the various devices should be selected initially i n the case of nonadjustable devices, or subsequently set on the adjustable ones, to obtain the selective operation essential to good system performance. BASIC TYPES OF SHORT-CIRCUIT DETECTION DEVICES There are three fundamental types of devices designed to detect overcurrents due to short circuits somewhere on the system. These basic devices are: 1 . Relays 2. Direct-acting trips on circuit breakers 3. Fuses 504

S Y S T u l OVERCURRENT PROTECTION

505

RELAYS

Relays are devices installed on the system to detect trouble and complete a circuit to electrically trip their associated circuit breakers, or eontactors, when necessary to isolate the trouble spot. Relays may be simple nvercurrent devices responsive to current magnitude only, or they may have a combination of current and voltage, or current and current coils, to detect the direction of current flow, current balance, differences in the current a t two ends of a circuit, distance, etr. The majority of relays in modern power systems operate from the secondaries of current and potential transformers rather than from series current coils or line voltage. Relays provide the best protection. They can he built to a much higher degree of accuracy than fuses and direct-acting trips, and they are adjustable both as to time and current. Also they can be designed to operate on only one direction of power flow to the point of fault or t o locate the fault by measuring the line impedance (distance) from the relay to the trouble spot. DIRECT-ACTING CIRCUIT-BREAKER TRIPS

Direct-acting trips are mounted directly on the circuit breaker they are associated with and trip it by direct mechanical action in response to current magnitude in the circuit. The direct-acting trips on low-voltage (600 volts and below) air circuit breakers are almost always actuated by the current in the circuit, but direct-acting trips on high-voltage (2300 volts and above) circuit breakers are usually energized from the secondaries of current transformers. Most time-delay direct-acting trips on low-voltage air circuit breakers are adjustable in the field, but some of them and many of the instantaneous trips are preset a t the. factory to operate a t a given multiple of the trip-coil rating of the breaker. Although direct-acting trip coils are much less accurate than relays, they are good enough for most low-voltage power system applications and for the small medium-voltage systems where t.hey are occasionally used. The justification of their use is strictly economic. FUSES

Fuses are thermally operated devices that combine the functions of fault detection and circuit clearing in one device. They are used on both high- and low-voltage systems. Refer to Chap. 3 for further discussion of the economics and other factors involved in their application. Fuses are subject to the disadvantages of being nonadjustable and

506

SYSTEM OVERCURRENT PROTECTION

quite slow in operation on moderate values of short-circuit rurrent. They are less accuratc than relays, but comparable n i t h direct-acting lomvoltage cirruit-breaker trips on high-current and superior to them on lowcurrent short circuits. Fuses also have the disadvantage that only one may be hlown, thereby leaving the circuit operating on single phase, which may cause trouble if the circuit is not properly protected with a thermal or other type of relay that will operate quickly enough on the current drawn during single-phase operation to protect the motor or other equipment on the circuit involved.

TYPES, OPERATING CHARACTERISTICS, A N D USES OF PROTECTIVE DEVICES In order to use any tool or device correctly, it is essential to know how it works and what it can do, or, putting it another way, what can be done with it, Therefore, the first step in learning t o apply and coordinate short-cirruit protection relays and other devices is to find out what types are available, how they operate, and what their time-current characteristics are. All relays and other short-circuit protective devices except fuses and the thermal trips on some low-voltage breakers work on one or the other of two fundamental operating principles: 1. Electromagnetic attraction 2. Electromagnetic induction

ELECTROMAGNETIC-ATTRACTION-TYPE RELAYS AND OTHER DEVICES

Devices of the electromagnetic-attraction class are operated by means of a magnetic plunger drawn into a solenoid or a hinged magnetic armature attracted t o the poles of an electromagnet. I n some cases the operating coil has taps to permit adjustment of the pickup-current setting. Pickup current is that value of current at which the plunger or armature will just start to move. Changes in pickup setting of plunger-type devices are accomplished by varying the position of the plunger in the coil. The pickup of hinged-armature devices may likewise be vaTied by changing the air gap or by maintaining a fixed air-gap setting and varying the spring tension. The farther it has to travel, or the greater the spring tension holding it back, the greater the amount of current or combination of current and potential required to operate the device. Hinged-armature-type construction is used in the direct-acting trips of air circuit breakers (600 volts and below) and also in some relays.


507

Plunger type construction is used for direct-acting trips on power circuit hreakers (2400 volts and above) and some relays. In modern practice most of t,he plunger-type relays are instantaneous units. Those used i n industrial plants are usually intended to provide fast tripping on high-magnitude short-cirruit currents. In such cases accuracy of the pickup setting is not so important as it is with time-delay relays, which may he required to operate accurately on relatively low currents. However, plunger-type relays with a bellows and adjustable air valve are available for use on systems where time-delay operation is required, but the more accurate and expensive induction-type relays cannot be justified. Since electromagnetic-attraction devices work about equally well on either direct riirrent or alternating current of the frequencies ordinarily used, all instantaneous plunger and hinged-armature-type relays and other devices are affected by the d-c component of asymmetrical shortcircuit current. Consequently, the offset (asymmetrical) factor must be taken int,o account when determining the performance of such devices. ELECTROMAGNETIC-INDUCTION RELAYS

The electromagnetic-induction principle is used in the design of many relays but not for direct-acting trip mechanisms. Such relays are essentially induction motors. The “stator” has current, or current and potential coils, and the fluxes created by the flow of current in them induce corresponding currents in a disk or cup. Interaction of the induced currents and fluxes creates torque to drive the rotor and thereby close or open the relay contacts. Surh relays are commonly referred to as “induction” relays. Figure 9.1 shows an induction-type overcurrent relay removed from its case. Electromagnetic-induction relays do not operate on direct current and consequently are not affected by the d-c component of an asymmetrical short-circuit current, as plunger-type relays are. Actually, the rate of change of the d-c component has some effect, but i t is of no practical significance. The rotor of the relay, which usually carries the moving contact, works against a restraining spring which returns it to the normal position when the relay is deenergized. It rotates a very small fraction of a turn, in the fast-operating nonadjustable time-setting relays, or almost a full revolution on the maximum time-dial setting of the adjustable time-current characteristic relays. Variations in time are accomplished by moving the time dial, or lever, t o a specified setting previously determined from a family of time-current curves supplied by the manufacturer for that

SYSTEM OVRCURRENT PROTECTION

i

FIG. 9.1 Induction avMulrent relay wHhDut cam.

particular type of relay. Moat adjustable induction-type overcurrent relays have 10 or 11 timedial positions, whose identifying numbers are arbitrarily assigned without regard to the actual operating time for the particular setting. The relay contacts are closed a t zero setting, and the contact gap opens progressively aa the timedial settings are increased. All short-circuit protection relays have silver contacts capable of cloaidg breaker tripping circuits up to 30 amp without injury to themselv& They cannot open them, however, without being damaged by the resulti arc. Consequently, they are sealed in by a seal-in unit, and an auxilia switch on the breaker mechanism is connected in the circuit to open it when the breaker opens. If the tripping current exceeds 30 amp, auxiliary tripping relay must be used. Occasionally, induction relays have to operate on quite small values f Short-circuit current. When doing so, there is relatively little torque available to hold the relay contacts tightly closed while the auxiliary switch on the breaker ia opening the tripping circuit. Therefore, in order to prevent the possibility of damaging the relay contacts by arcing which may occur with light contact pressure and also to ensure positive tripping,

ti

c

d


SOP

provision is made t o bypass the main relay contacts with a seal-in circuit or alternatively t o hold them closed magnetically until the tripping circuit is interrupted by an auxiliary switch on the circuit breaker being tripped. The seal-in, or holding-coil, circuit is completed when the main relay contact closes. The holding coil energizes an electromagnet which attracts a soft-iron armature mounted on the rotating element of the relay, thus holding the relay contacts firmly in the closed position. The seal-in coil closes a seal-in contact which completes a circuit t o bypass the main relay contact. I n some cases the contacts are both held closed and bypassed, in which case one coil accomplishes both functions. The usual practice is t o have the holding, or seal-in contact, coils simultaneously release the operation-indication target also. I n some designs, targets are positively actuated, and in others they are gravity operated when a latch is released. They are always manually reset by means of a button on the outside of the relay case. Sometimes a separate auxiliary seal-in relay is provided inside the main relay case to bypass the relay contacts. Its coil is connected in series with the trip coil and the main relay contacts so that it is energized and closes its contacts as soon as the tripping circuit is completed. The seal-in relay does double duty by completing its own coil circuit and sealing itself in, as well as bypassing the main relay contacts in the tripping circuit. I n such cases, the operation target is actuated by the same coil. Both seal-in and holding coils are in series with the main relay contacts and the breaker trip coil, or auxiliary relay coil, and therefore must be capable of picking up on the current drawn by them when the main relay contacts close. These coils are usually supplied with 0.2-amp and 1.0-or 2.0-amp taps. The 0.2-amp tap is for use with trip coils and auxiliary relays that take 0.2 t o 3.0 amp and can safely carry tripping currents as high as 5 amp. The 1- or 2-amp coils should he used when the protective relay contacts trip a circuit breaker directly and the tripping current is not more than 30 amp. In some cases, i t is necessary for the fault protective relay t o energize several tripping circuits simultaneously, e.g., common practice is to trip the generator main and field breakers and the turbine throttle valve, and possibly also sound an alarm and operate an annunciator when the generator differential relay operates. I n such cases an auxiliary tripping relay having multiple contacts, each capable of energizing a trip-coil circuit, is provided. The opefating coil of the auxiliary relay is connected in series with the main relay contacts. When the main relay contacts close, they energize the target and hold-in or seal-in contact coils, as well as energizing the auxiliary relay coil. The generator-differential auxiliary relay is hand reset, i.e., its tripping and alarm contacts stay in the closed position even though its coil circuit is opened. This makes i t possible for the auxiliary

510


relay t o open its own coil circuit without having any risk of a “race” between the breaker-tripping contacts and the self-deenergization operation of the auxiliary relay. Sometimes auxiliary relays are used simply as a means of increasing the number of trip circuits without including the lockout (hand-reset) feature. OPERATING-TIME CLASSIFICATION OF PROTECTIVE DEVICES

All short-circuit-current protective relays and other devices can be classified under one of these headings: (I) instantaneous, (2) highspeed, (3) time-delay, (4) combination instantaneous or high-speed and time-delay. By ASA (American Standards Association) definition, instantaneous relays are those which have no intentional time delay. Some of them operate in less than one-half cycle, while others may take as much as 0.1 sec (six cycles). Those which operate in three cycles or less are also classified as high-speed relays. Time-delay relays may he induction, hinged-armature, or solenoid type. Usually the time delay is adjustable. Most of them are induction type with an inverse characteristic, i.e., the relay speeds up progressively as the actuating quantity (rurrent alone or the product of current and voltage, etr.) inrreases. However, a few time-delay relays operate a t a constant speed predetermined by adjustment and are independent of current magnitude as long as the current is sufficient to operate the relay. These are known as definite-time relays. The dirert-arting trip mechanisms on rircuit breakers may be instantaneous, or time-delay, or a combination of the two. R’ormal fuses are instantaneous or time-delay in their operation, depending on the magnitude of the short-circuit current. Some fuses, however, are designed t o give evtra time delay on moderate values of overcurrent in order to ride through permissible high overloads. INDIVIDUAL RELAY CHARACTERISTICS

There are many types of short-circuit-detecting relays. Some work on current magnitude only, and some operate on rurrent only, but take into account the direction of current flow. Others work on a differential principle, and still others on the basis of measuring the impedance in the circuit as represented by the current and voltage to determine when the relays shall work. Following is a hrief description of the relays in the different rategories and their general operating chararteristirs.


51 1

Relays Operating on Current Only (Nondirectional). Probably the most commonly used short-circuit protection relays in industrial plant power systems are the instantaneous and time-delay relays responsive to current magnitude alone without regard to the direction of current flow. Most instantaneous overcurrent relays are plunger or hinged-armature type. They may he supplied as an instantaneous element mounted inside the ease of a time-delay induction relay or as single or multiple instantaneous elements alone mounted in a case. Even though such relays are classified as instantaneous, a finite length of time is required for them to operate at different current magnitudes, as shown in the lower left corner of Fig. 9.3. Instantaneous relays are available in a wide range of current coil ratings from 1.5 to 80 amp or even more if required. Most time-delay overcurrent relays are inverse-induction type. An “inverse” time-current characteristic means that the relay operating time decreases as its operating current inrreases. Such relays are classified as “inversetime,” “very inverse-time,” and “extremely inverse-time.” Current pickup of the relay is selected by means of the taps in the operating coil, and time adjustments are made by means of a time dial or a time lever. For convenience in making adjustments, the total dial or lever-movement scale is arbitrarily divided into 10 or 11 divisions. Each design of relay has a family of time-current operating curves corresponding to the numhered divisions on the scale. Figure 9.2 shows such a family of curves for a relay with an inverse-time characteristic. The same family of curves is applicable for all current ranges of a given model of relay. There are different models of relays with varying degrees of inverseness. Figure 9.3 shows the difference in the time-current characteristics of inverse-time, very inverse-time, and extremely inverse-time types of relays on minimum and maximum time-dial settings. It also shows the time-current charact,eristic of the instantaneous element when it is provided in any of these three types of relays. The time-current curves of the three relays are quite different, making it difficult to obtatn satisfactory coordination when relays with different characteristics operate in series (see Fig. 9.17). The inverse-time relay is widely used for general application. It is better than either of the others on systems where there are wide variations in short-circuit current levels because of changes in the number of power sources in use. Its relatively flat time-current curve permits the relay to give reasonrtbly fast operation over a much wider range of shortcircuit current than the others can. This feature enables the inverse-time relay to afford a satisfactory degree of fault protection with one or all of the power sources in operation. The very inversetime relay has a steeper curve, which makes it slower on lox values of current and faster on the higher magnitudes of fault cur-


512

rent. I t is not so good for systems with variable generating capacity as the inverse-time relay, but it is better on systems supplied from large power company systems where the short-circuit-current level at a given point is more or leas fixed by the impedance of the system up to that point. The latt,er limitation restricts the current range over which the relay has to provide fast, performance so the curve ran he steepened. The extremely inverse-time relay was designed primarily for use on power company distribution system feeders, where it is necessary t o have a relay which will ride through the high initial-load current encountered when reenergizing a feeder after an outage and yet provide fast operation when needed for short-circuit protect'ion. At first glanre, one might think that the extremely inverse-time relay was partirularly well suited for coordination with fuses, since the shape of its time-current curve is nearer that of fuse melting-time curves than I

420

360

300

60

0 MULTIPLES OF MINIMUM CLOSING CURRENT (TAPVALUE)

FIG. 9.2

Time-current curyes of inverse-time induction overcurrent relay.


513

are the other relay curves. Practically, however, this relay is undesirable immediately ahead of a large fuse operating 011 the same magnitude of fault current, because it is so fast on high values of current that it is quite likely to tripits breaker unnecessarily when the fusealoneshould bloivfora fault on its load side. The behavior of this relay when installed on the

FIG. 9.3 Curves of time-current characteristics of inverse-time (A), very inverse-time (El, extremely inverse-time (C) induction relays, and instantaneous element ( D ) .

514


line side of large fuses is discussed further in the section on coordination and is illustrated in Fig. 9.21. However, the extremely inverse-time relay is satisfactory when the fuses on its load side are relatively small, or when the circuit impedance between relay and fuse results in sufficient current differential to permit selectivity, and also when used to trip a circuit breaker with a fuse ahead of it. I t will be noted that the time-current curves in Figs. 9.2 and 9.3 are plotted in terms of multiples of minimum pickup. Pickup current of a relay is the rating of the current tap in use. Induction relays generally have several taps brought out of their operating coils to permit selection of the desired pickup current. For euample, the tap range of a 4- to 16-amp relay coil is 4, 5, 6, 8, 10, 12, 16 amps. To illustrate the meaning of multiples-of-pickup current, a relay operating with a 6-amp tap setting connected to the secondary of a 200/5-amp current transformer (40/1 ratio) would see 2400/(40 X 6) = 10 multiples of its pickup setting with 2400 amp flowing in the primary of the current transformer. Since standard relays are built on a production-line basis, manufacturers are allowed certain tolerances in operating characteristics. Therefore, an induction overcurrent relay meets standard specifications if i t operates Ivithin 2 to 7 per cent of the standard characteristic time-current curves for that type of relay. I n general, the accuracy will be highest a t high values of operating current, i.e., high multiples of pickup. However, after a given relay has been adjusted t o have a specific time-current curve by putting current through it and accurately checking its operating time for varying values of current, it should operate consistently within approximately 2 per cent of that time-current curve. Induction-type overcurrent relays are available as standard devices for 25-, 50-, and 60-cycle service and in a variety of current ranges such as 0.5 to 2, 1.5 to 6 and 4 to 16 amp. The low-current coils rated 0.5 t o 2 and 1.5 to 6 amp may, of course, be used wherever low-current pickup is desired, but they are primarily intended for residual connection in the neutral (Fig. 9.4) of the phase-short-circuit-relay current transformers t o detect ground-fault currents. It should be noted that the low-current relays impose heavier burdens on their current transformers than the others do. All time-delay overcurrent induction relays will start t o move and will eventually close their contacts on current equal t o their current-tap (pickup) setting, assuming that they are in good operating condition and free from dust, etc. However, because of the manufacturing tolerances allowed and the low operating torque available from such small currents, i t is desirable to select a current-tap setting such that the relay will not be expect,ed to give accurate time-current performance below approximately 1.5 multiples of minimum pickup.

iYSTEM OVERCURRENT PROTECTION

515

PHASE

J 1

I _ FIG. 9.4

Schematic diagram of residual connection for a ground reloy.

The reset time of induction overcurrent relays, i.e., the time for their contacts t o returii t o their completely open position when the relay is deenergized, varies with the time-dia1 setting and the type of relay. On the 10 time-dia1 setting, the approximate reset time is O sec for the inverse-time and GO sec for the very inverse-time and extremely inversetime relays. For loiver time-dia1 settings, the reset time is reduced approximately in proportion t o the setting. The relay contacts will also reset eventually, if t,hey are not sealed closed, mhen the current decreases t o less than 90 per cent of t h e pickup setting. Although the complete reset time of a relay for a given time-dia1 position may he quite long, the contacts wili have separated in six cycles (0.1 sec) with normal adjustment or “ivipe.” This permits using the relay in instantaneous reclosing schemes; i.e., a circuit breaker can be automatically reclosed following an outage without having t o mait for the disk t o reset completely. The relay contacts will not be in their normal position, however, and therefore, if the circuit is still short-circuited, the relay will operate t o trip its hreaker in much less than the normal time for the time-current setting of t h a t particular relay. Basically, different manufacturers’ versions of a given type of relay will conform t o the general patterns discussed. Nevertheless, there are sufficient variations in time-current curves t o necessitate procurement of the manufacturers’ d a t a for the particular device involved, if reasonably accurate settings and performanee are t o be ohtained. Generator Overcurrent Relay with Voltage Restraint. An overcurrent relay with voltage restraint is an induction-disk unit with a voltage-

516


restraining circuit which restrains or bucks the actiou of
8 6

5 4

3 2 v)

0

z

0w 0 ' 0.8 0.6

0.5 0.4

0.3 0.2 ---I1

pI -1

RELAY

0. I

0. I

FIG. 9.5 restraint.

0.2 Time-current characteristic curves of generator overcurrent relay with voltage

I


9.5 shows the zero- and 115-volt-restraint time-current curves.

517

The amount of voltage restraint will depend on the type, location, and severity of the fault. The curves for intermediate values of voltage restraint, which will fall inside these two extremes, can he obtained from the manufacturer. The relay has current taps, to permit adjustment of the current pickup setting, and time-dial adjustment. to control the operating time for a given combination of current and voltage restraint. Voltage-controlled Overcurrent Relay. The voltage-controlled overcurrent relay has an induction-disk time-delay overcurrent element and a solenoid-operated undervoltage element. The two elements are interconnected so that the voltage element must close its rontacts before the overcurrent element can start t o move. The overcurrent element can have a n inverse, inverse-definite-time, or very-inverse time-current characteristic as desired. It has current t,aps and a time-dial adjustment. The relay is applicable where it is desired that an overcurrent element be set t o operate on less than full-load current when the voltage falls helow a predetermined value. A typical application is overcurrent backup protection for generators. Directional Relays-General. A directional relay operates when the current in the rircuit floivs in a given direction and ignores current flowing in the opposite direction regardless of its magnitude. It ran he designed t o work on either useful power (kw) or short-circuit current (most,ly reactive). This discussion will be confined to the latter type. Directional short-circuit-detecting relays may he of the directionalovercurrent type or directional-product type. The current-measuring element of the overcurrent t,ype is a simple inst,ant,aneous and/or timedelay overcurrent relay. In the product-type relay, which is iised only for directional ground-fault protection, a single operating element,artuated by the product of two currents, or a current and a potential, indicates the direction of current flow as well as its magnitude. Some of the uses of directional relays are the following: 1. T o permit tripping a circuit breaker for one valne of time and cnrrent when the short-circuit current flows in one direction in the cirruit and a different time and current when the current is flowing in the opposit,e direction. 2. T o obtain selective tripping between circuit breakers at the receiving ends of parallel lines when a fault occurs on one of the lines causing the same current to flow thi-uugh the relays on the good line and the one in trouble. 3. T o obtain selective tripping of a grounded-neutral generator or transformer circuit breaker -when a ground fault occurs in the protected unit and there are other sources supplying ground-fault current.

518


Phase-fault directional relays will also operate on ground faults provided that the ground current is large enough t o operate the relay, i.e., is not unduly restricted by external impedance in the neutral of the transformer, or generator, supplying power t o the system. However, directional ground relays, either overcurrent or product type, are usually supplied in addition t o the directional phase relays on both solid and resist,ance-grounded-neutral circuits. Directional-overcurrent Relays. One design of directional-overcurrent relays has a low-energy instantaneous directional-control element whose contacts prevent operation of both the instantaneous and time-delay overcurrent elements unless the short-circuit current is flowing in the direction for which it is desired t o have the associated circuit breaker tripped. The dirertional element operates on the same hasic principle as the wattmeter, but is designed t o respond t o out-of-phase short-circuit current rather than in-phase load current. This element responds t o the direction of current flow without regard t o its magnitude. The overcurrent element, \yhich measures the magnitude of the current, may be instantaneous or time-delay, or the relay may have both elements. The time-delay element is essentially the same as that in the ordinary inversetime or very inverse-time overcurrent relays. It has current taps t o permit adjustment, so that the relay will operate on the desired magnitude of short-circuit current. The relays also have a time dial, or lever, with which t o adjust the distance which the induction disk has t o travel, therchy controlling the time required for the relay t o close its contacts. Their dial, or lever, wales are arbitrarily divided into approximately 11 divisions as in the case of ordinary overcurrent relays, and their timecurrent curves are the same. In other designs of directional-overcurrent relays the time-delay element is (.ont,rolled hy the directional element, hut the instant,aneous element picks up independently of it. Directional operation when the instant,aneous element operates is ohtained hy having the contacts of the dirertional and instautaneous elements in series, so that, even though the instantaneous clement does pick up, nothing happens unless the contacts of thc directional clemciit are also closed. Provision is made to prevent possihle false operation that might occur because of the fact that for a given fault loration t,he current might, he flowing in the wrong direction for tripping and yet the nondirectional instantaneous element would pick up, therehy (.omplet,ing the tripping circuit falsely, if the sensitive, fastoperating ilirertional element should suddenly reverse and close its contarts before the instantaneous element had time t o drop out during the surge folloiving interruption of the short circuit by some other breaker. Iiirectional-o~erciirretit relays for phase-phase and three-phase faults arc single-phase units, and three are normally used for a three-phase cir-


519

cuit. Both the directional and overcurrent-element current coils are connected t o current transformers in the line, and the potential coils are usually connected to t,wo open-delta or three Y-Y potential transformers in a quadrature or 90" arrangement (i.e., the current at unity power factor leads the potential 90'). Directional-overcurrent ground relays are similar in construction t o the directional phase-fault protection relays. I n order t o simplify the application and also reduce the number of varieties to be carried in stock, some of the designs of direcbional-overcurrent ground relays have provision for dual polarization, i.e., they have both current and potential polarizing coils. The use of both coils is often advantageous in that it assures polarization of the relay whether or not the grounded-neutral generator or transformer at a given location is in service. The operating-current coil of a directional-overcurrent ground relay is connected in the neutral of the line-current transformers (Fig. 9.4). Normal load currents in t,he current-transformer secondaries even though unbalanced and phase-to-phase or three-phase short-circuit currents do not cause current to flow in the neutral. Therefore, only ground-fault current flows in the neutral connection except for error currents which will be discussed later under the general subject of coordination. The polarizing-current coil is connected to a current transformer in the neutral ground connection of a power transformer or generator, where the current flow will always he in the same direction (Fig. 9.6.4). The potential polarizing coil of t,hese relays is connected across the open corner of the broken-delta secondary of Y-delta-connected potential transformers, as shown in Fig. 9.6B. Under normal operating conditions the three voltages are equal and no voltage appears across the relay coil, but as soon as a ground fault occurs the relay potential coil sees a voltage, whose phase angle corresponds to that of the grounded primary phase and whose magnitude is three times the zero-phase-sequence voltage. On highvoltage systems, polarizing voltage can also be obtained from hushing potential devices and coupling capacitor potential devices. The latter are seldom required on industrial plant power systems because the power system voltages are relatively low. The potential coils of the phase-fault directional relays can be connected to the same potential transformers by using a 60" connection, i.e., the current at 1.0 power factor leads the voltage 60'. An alternative arrangement would consist of Y-Y-connected potential transformers for the phase-fault relays with Y-broken-delta auxiliary transformers for the . polarizing coil. Product-type Directional Ground Relays. Product-type directional ground relays indicate the direction, as well as the magnitude, of the


520

ETC r.

t

\/

b

I F

-\

t m D

.,

e .

1 MAIN

POLARIZING

GROUNDED Y-CONNECTED TRANSFORMER OR GENERATOR

-

t

TIE LINE

(A)

/.+

PHASERELAYS ETC

.\

4.d

.I

J-

t

TIE LINE

(8)

Schemotic diagrams of directional ground relay polarized with current ( A ) or potential (8) for operation on ground-fault current flowing in direction of owow in tie line.

FIG. 9.6


521

short-circuit current flowing in the tripping direction by means of a single element operating on torque obtained from the product of ground-fault current and a polarizing current or potential. The main coil is connected in the neutral of the three Y-connected line current transformers (Fig. 9.4). The polarizing effect required for directional operation is obtained by means of a current or potential coil. The connections of both operating and polarizing coils would be the same as for the directional-overcurrent ground relays. Either the overcurrent or product type of dirertional ground relay will afford essentially the same degree of protection to industrial plant power systems, which are relatively compact and usually have the system neutral grounded a t only one station. The product-type relays are somewhat more sensitive and permit greater selectivity between relays under certain conditions, which factors may be advantageous on a complex power company system but are of minor, if any, advantage in industrial plants. Directional-overcurrent relays are simpler to apply and adjust than the product type, because only the current magnitude and time-dial setting affect their operating time, whereas the operation of a product-type relay is affected by the line current and the polarizing voltage or current, or both, and the phase angle between them. Directional-overcurrent Relays with Voltage Restraint. These relays are the same as the ordinary directional-overcurrent relays except that there is an additional voltage-restraint circuit in the directional unit. This feature enables the relay to distinguish between short-circuit and heavy-load currents, which may not be very differentin magnitude under some conditions. Under normal overload conditions the system voltage is essentially normal and the voltage-restraining coil prevents operation of the directional element of the relay. As soon as a short circuit occurs, however, the voltage decreases and the restraining effect is reduced or entirely removed, whereupon the contacts of the directional element close and permit operation of the overcurrent element. Since these relays are intended for use on circuits with low magnitudes of shortcircuit current, they are available with 2- to 8-amp coils, as well as the usual 4- to 12-amp or 4- t o 16-amp rating. Differential Relays-General. Differential relays depend for their operation on the fact that when conditions are normal the current flowing into one end of a generator winding, one side of a transformer, or one end of a circuit is balanced by an equivalent current flowing out the other end, i.e., what goes in has t o come out, if everything is in order. This makes it possible to build relays that “watch” the ingoing and outgoing currents and operate when idifference between them indicates that something is wrong inside the protected equipment or circuit. Plain overcurrent relays can be used as differential relays. However,

522


since they work on simple current differential without the help of restraining windings, they must he set quite high in current t o avoid false operation due to current-transformer inaccuracies. Therefore, specially designed differential relays have replaced them almost entirely. Generator Differential Relays. As shown in the diagram of Fig. 9.7, two current transformers of equal capacity and similar characteristics are installed in opposite ends of each generator phase winding, and their secondaries are connected in series with the restraining coils (RC) of the differential relay. Under normal operating conditions the same current flows through the two current-transformer primaries, and corresponding secondary currents circulate through the restraining coils of the relay in the direction of the arrows. When a short circuit develops inside the generator, the current in the two current transformers is no longer the same and the difference in current will circulate through the operating coil (OC) of the relay. When this difference-current flowing through the operating coil exceeds the current in the restraining coil by a certain percentage, the relay operates instantly to trip the generator line and field circuit breakers through an auxiliary relay. These differential relays are of two types. One works on a constant percentage difference in current in the two current transformers (Fig. 9.7) and the other works on a percentage difference that increases rapidly as the short-circuit current increases (Fig. 9.8). They are commonly referred to as constant-slope and increasing-slope relays, because of the shape of their operating characteristic curves. Both types work on the same basic principle of checking the balance of current in the ingoing and outgoing current transformers. The important difference between them is the fact that the increasing-slope type, while costing slightly more, requires less arcuracy in the performance of its current transformers than the constant-slope type does. This means that less time need be spent in calculating the performance of the current transformers, less accurate and consequently less expensive current transformers might possibly be used, or other relays or meters, etc., could be connected in the same circuit with the differential relays, without running the risk of false relay operation due to unequal current-transformer behavior. The purpose of the slope in the two relays is t o prevent false operation due to current-transformer-error currents that might flow in the differential-relay circuit during a severe short circuit outside the differentially protected zone. Error currents ocrur because no two current transformers will perform exactly alike even though made to the same specifications and from the same lot of material. Because the current transformers are not absolutely alike, they saturate unequally when high currents flow through them during external short circuits and t,heir ratio breaks down unequally. If this happens, the unbalanced current flows in the differential-relay circuit, and the relay has no way of knowing

SYSTEM OVERCURPENT PROTECTION

l l / V X A A / I / ~ / W l

I

4 2 0 2 12 AMPERES CURRENT FLOWING FROM BUS T O GEN 6

I

I

I

1

1

I

1

I

I

1

8 10 12 12 AMPERES CURRENT FLOWING FROM GEN TO BUS 4

6

523

I

1 14

~

!

I

16

FIG. 9.8 Schematic connection diagram and operating choraderirticr of increaring-slope generator differential relay.

514


whether the current it sees indicates a fault in the generator or a “mistake” on the part of the current transformers, which the relay should ignore. The constant-percentage differential relay works on a 10 per cent slope, as shown in Fig. 9.7. The V-shaped shaded area represents the plus or minus 10 per cent margin allowed for current-transformer errors due to unequal characteristics and saturation. The current transformers to be used with such a relay should be selected so that the difference in secondary current output of I , and I , current transformers will not exceed 5 per cent under maximum fault conditions, which leaves a safety factor of 5 per cent without exceeding the 10 per cent margin built into the relay. The increasing-slope differential relay works on the same principle of watching the difference in current output of the current transformers in opposite ends of the generator winding. The essential difference is that the relay is designed so that the margin allowed for current-transformer errors increases rapidly as the short-circuikcurrent magnitude increases. This permits the relay to operate on 10 per cent current differential on low-magnitude faults when there is no danger of current-transformer errors and still not operate incorrectly during severe through faults (external to the generator) even though one current transformer falls down completely. In order for generator differential relays to operate when current is flowing from the generator to the bus, two requirements must be met, namely, (1) the per cent difference between I I and 12 must be greater than the per cent slope shown in the relay curves for the current magnitude involved in that particular fault, and (2) the differential current must equal or exceed the minimum pickup current of 0.1 or 0.2 amp. The relays will always operate on internal faults in a generator when there is feedback current from other power sources, provided the differential current exceeds their minimum pickup current. For all practical purposes both the constant- and increasing-slope differential relays are instantaneous in operation, but the latter is slightly faster. The limitation of ground-fault current, which is usually done on industrial plant power systems for reasons discussed elsewhere in t,his book, reduces the sensitivity of generator differential-relay protection in varying degrees, depending on the relative magnitude of the ground-fault and rated-load current of the generator and the characteristics of the current transformers. The generatar neutral relay can be set to compensate for this deficiency as far as sensitivity is concerned, but i t is only an ordinary overcurrent relay and hence may not operate selectively ip case of an internal fault in one of several individually grounded generators in parallel. Consequentk, a sensitive differentially connected directional ground


525

relay should be installed to supplement the main generator differential relay unless the nature of the load is such as to permit shutting down all generators in the plant, if an internal fault in one generator happens to be close enough to its neutral to limit the ground-fault current to such a low value that the main differential relays cannot operate. The unprotected area varies from about 11 to 19 per cent when the neutral current is limited to generator full-load current, and increases if the ground-fault current is further reduced. Transformer Differential Relays. Transformer differential relays are of the constant-percentage-differential type. They work on the same basic principle as generator differential relays, i.e., they watrh to see that when current enters one winding a corresponding amount leaves the other winding, or windings in the case of multiple-winding units. The connections of a transformer differential relay are the same as for a generator differential relay except that the current transformers in the different circuits to transformers have to be of different ratios to compensate for the fart that there is a different voltage in each circuit, and ronsequently the currents are not the same. One type of differential relay used with large transformers has a harmonic restraining feature \\ hich prevents false operation of the relay on the magnetizing current when the transformer is first energized. Since this magnetizing current may be as much as twelve to fifteen times normal, a differentially connected relay responsive only to current magnitude has to be set high enough in current and time to ride through the magnetizing period, which means corresponding reduction in the protection it can provide. The harmonir-restraint feature works on the basis of the fact that the magnetizing inrush current has a distorted wave form containing mostly second- and third-harmonic components, which are not present in short-circuit current. These harmonic components are filtered out and used to restrain the relay and prevent its operation on the inrush of magnetizing current. Provision is made to keep the harmonic-restraint circuit from blocking desired relay operation on internal transformer faults severe enough to create substantial amounts of third-harmonic current due t o a-c saturation of the current transformers. Another form of differential relay used with large transformers uses a supplementary tripping suppressor to prevent false operation on inrush current. The suppressor introduces the necessary time delay to ride through the magnetizing-current inrush under normal conditions and also makes provision for immediate tripping in the event that there is a fault inside the protected zone during the magnetizing period. The non-harmonic-restrained relays used with relatively small transformers depend on a slight time delay to enable them to ride through the magnetizing inrush. I n some cases, however, the inrush current is so

526


high that this time delay is not sufficient t o prevent false operation, and i t becomes necessary t o supply supplementary desensitizing equipment, or the suppressor just mentioned. The desensitizing equipment automatically makes the relay less sensitive until the magnetizing current inrush has disappeared. Transformer differential relays require more slope, i.e., a greater per cent difference in the output of t.heir current transformers than generator differential relays do, to allow for the unbalances in current caused by transformer-tap changing in addition t o the differences due to currenttransformer saturation. Without the extra slope it would be necessary to readjust the relay taps whenever the tap ratio of the transformer was changed, e.g., a 5 per cent change in transformer taps causes a corresponding change in the current in that winding, whereas the c.nrrent in the ot.her winding remains essentially t,he same for a given kva load. Some transformer differential relays have a single per cent slope characteristic, while others are adjustable for different slopes. The higher percentage slopes are for use with transformers having a wide tap range, as in load-ratiocontrol equipment. The currents in the different connections to transformers differ depending on the voltage ratio, and consequently different current-transformer ratios are required. I t is necessary for economical reasons to use standard current transformers, however; so it is seldom possible to obtain a cornhination of current transformers that will produce exactly equal current in their secondaries. Therefore, some transformer differential relays are provided with several taps in their windings to permit balancing the ampere-turns in the relay elements connected to the different main-transformer circuits. Other designs of relays depend on external tapped autotransformers to do this. The tap range is sufficiently broad to take care of the requirement that the current-transformer secondaries must be connected in the reverse order of the main-transformer windings, e.g., the current transformers for a Y-delta-connected transformer would he connected delta on the Y-connected side of the transformer and Y on the delta side. This means that the relay coil connected to the delta-connected current transformers would see 1.73 times the current in its individual current-transformer sesondaries, i.e., there would he 8.7 amp in the relay coil when a 5-amp secondary current transformer had full load in its primary. Differential relays are available for hoth two- and three-winding transformers. Those designed for use with three-winding transformers work on the same principle as the others. Bus Differential Relays. There are three types of relays available for differential protection of buses:


527

1. Ordinary time-delay overcurrent relays 2. Current-actuated bus differential relays with restraining coils 3. Voltage-actuated relays either differential-voltage or linear-coupler type Ordinary overcurrent relays can he differentially connected, so t h a t they will operate on the difference between the summation of the currents entering and leaving the bus. They are not particularly well adapted for the purpose, however, because the only vay to prevent their false operation when the ratio of a current transformer breaks down, because of d-c and a-c saturation during a severe external fault on a feeder, is to set them high in both current and time. When so set, they are too slow and insensitive to he very effective. The current-actuated differential relays have restraining coils and are much better than the ordinary overcurrent relays, hut they are still not so effective as the voltage-actuated relays. The latter are able to discriminate instantly between faults in the bus and external short circuits on feeders outside the differentially protected zone, even though one or more of the current transformers in the group become completely saturated. One form of these voltage-actuated relays is known as a differentialvoltage relay. It is designed for use with hushing-type or the through or window-type current transformers used in metal-clad switchgear. The relay is connected across the paralleled secondaries of current transformers i n each of the incoming and outgoing circuits. When conditions are normal, the vertor sum of these secondary currents is zero, and no voltage appears across the relay coil. But as soon as either a bus (internal) or external fault occurs, the vector sum is no longer zero and the flow of unbalanced current creates a voltage drop across the relay. Because the voltage appearing across the relay for any magnitude of bus fault is greater than that during an external fault, the relay is able to distinguish between an external and internal fault. The other form of voltage-actuated bus differential relay is known as a linear-coupler relay. These relays are connected in series with air-core mutual react,ors (linear rouplers), which look like conventional bushing current transformers. T h e couplers are installed in all the circuit,s connected to the hns and generate volt,ages in proportion to the flow of current in their primary circuits. Under normal load and external fault conditions, the voltages generated in the couplers in the incoming power circuits' rancel those in the outgoing circuits and no current circulates through the relay. V h e n a bus fault occurs, however, the voltages do not cancel each other, and the resulting current flow causes the relay to operate.

528


A-C Wire-pilot Relays (Tie-cable and Relatively Short Transmissionline Protection). 4 - C wire-pilot-relay protection is a form of currentdifferential relaying modified so that a very small portion of the current appearing in the secondary of the current transformers at opposite ends of t.he circuit, actually f l o ~ through s the pilot wires. I t is designed to provide fast phase-phase, three-phase, and ground-fault protection of tie cables or relatively short transmission lines. In an ordinary differential-relay circuit, the secondary current from the current transformers circulates continuously through the relay and current transformers over a set of four control wirea. Since the four wires must be large enough to limit the impedance sufficiently to avoid saturaliou trouble by overburdening the current transformers, such a system is impractical unless the latter are within a few hundred feet of each other and the relay. This obstacle is overcome in the a-c mire-pilot protection system, because the relays and auxiliary devices are designed to take just a “sample” of the currents flowing in the current transformers at each end of the cable or transmission lines and then compare these samples over a pair of relatively small pilot wires. Under normal load conditions, or even when short-circuit current flows through the line to a fault outside the protected zone (through short circuit), the samples match each other and the relays do not operate. As soon as a fault occurs in the line between the relays, however, the samples no longer match each other, indicating that there is trouble in the protected zone, and the relays operate instantly t o trip their respective circuit breakers. There are two types of a-c wire-pilot relay systems. One operates on the opposed-voltage principle and the other on the circulating-current principle. In an opposed-voltage system, the current samples taken at each end of the line create voltages that are equal and opposite to each other, with the result that no current f l o w in the pilot mires under normal conditions. When a fault occurs, the voltages are thrown out of balance, current flows through the pilot wires, and the relays operate. In the circulating-current system, the small sample current circulates continuously through the pilot wires and directional relays a t each end. When a fault occurs inside the protected zone, the pattern of current flow changes and the relays operate to trip the breakers. Either system will provide adequate protection, although each has some minor advantages that might make it bett,er adapted for a specific application. These systems work best on two-terminal lines, i.e., single circuits with no t,aps or branches, but they can be modified for use with some combinations of lines with branch circuits or taps. . Both systems incorporate the necessary restraint features t o prevent


529

false operation due to current-transformer errors during severe short circuits outside the protected zone. These systems will work with pilot wires having a total loop resistance of 2000 ohms and a capacitance of 1.5 pf, which is equivalent to approximately 23 miles of single-conductor No. 19-Awg telephone wire (11.5 power-line-circuit miles). With certain special modifications, the systems will work with higher resistance pilot v.ires, permitting smaller wires or a longer pilot circuit. Industrial plant tie cables are generally of moderate length installed inside the plant, and privately owned multipleconductor control cables are used for pilot wires. It is common practice, however, to rent pilot wires from the local telephone company for the longer circuits involved in transmission-line protection. Pilot-wire Supervisory Equipment. Supplementary eouiprnent can be provided, if desired, which will continuously check on the condition of pilot wires to give warning in case the wires become short-circuited or open-circuited. Current-balance Relays. Current-balance relays are available in induction-disk and induction-cup-type construction for the protection of parallel transmission lines that feed a common terminal and have no branch lines. They operate on the principle that, under normal load or through-fault conditions, the current will be balanced between the two lines. If the currents become unbalanced by a predetermined amount because of a fault in one line, the relays operate instantly to trip the circuit breaker on the line in trouble. Current-halance relays provide phase-phase and three-phase fault protection. Current-halance protection of lines is subject to the handicap that the fast-operating current-halance relays must he disconnected from the tripping circuit when one of the circuit breakers is opened, either manually or by relay operation. Otherwise, the load current which was formerly divided between the parallel lines will unbalance and, therefore, operate the relay when i t is all thrown onto the remaining line or lines. The transfer from current-balance to plain overcurreut relay protection is accomplished automatically by means of auxiliary switches on the circuit breakers. Because of this disadvantage, i.e., reversion to ordinary overcurrent protection on occasion, current-balance protection of transmission lines is losing popularity in favor of distance relays, which provide highspeed protection to the lines whether they are operating individually or in parallel. Distance Relays. Distance relays obtained their family name from the fact that they operate on the basis of the balance between voltage and current, the ratio of which can he expressed in terms of impedance. Since impedance, in turn, is an electrical measure of distance on a trans-

530


mission line, which these relays are rommonly used to protect, it seemed quite appropriate to call them “distance” relays. They are a very versatile and useful family of relays operating on the hasis of the location of the fault without regard t o the magnitude of the short-circuit current. Consequently, they can give much faster tripping than relays depending on current magnitude and time settings for selectivity. They are seldom used on industrial power systems, however, because the latter are too compact and the distances too short in most cases to permit proper application of these relays. Consequently, it seemed best not to attempt t o discuss them in detail in this book. CURRENT-LIMITING AND STANDARD FUSES

I

The two basic types of fuses are the current-limiting and non-currentlirnking or what might he termed standard fuses. Practically all fuses will melt in considerably less than one-half cycle on a 60-cycle basis when subjected t o high values of fault current. However, the arc is a conductor and enables the current to reach its mavimum crest value unless provision is made to put the arc out before the current can reach its crest. When such provision is made, the fuses are classified as current-limiting. Most fuses are self-protecting, that is, they are capable of extinguishing the arc for any value of current within their interrupting-capacity rating limit. Current-limiting fuses for motor-starting service are purposely designed t o carry low values of current for considerable periods of time t o permit repeated starting or jogging of motors. Such fuses must he used in conjunction with a thermally controlled contactor, or other circuitinterrupting device, capable of interrupting currents that would require more than 10 sec to melt the fuse link. Otherwise the entire fuse unit might he overheated to such an extent that. it would not be able to successfully interrupt the current when the link finally melted. Some current-limiting fuses used in the primary of load-center transformers, where it is desired to have selective operation between the fuse and a transformer secondary breaker, are designed to give somewhat longer melting time than standard fuses on high values of fault current in order t o clear the instantaneous trip in the rircuit breaker. Most fuses have a “smooth,” that is, a continuous melting-time curve, but certain types are purposely designed to give more time on moderate overloads of two or three times fuse rating. Such fuses have a jog in their melting-time curves a t the point of transition from slow t o st,andard melting-time characteristic. Fuse time-current operating characteristics are given in terms of the


531

532


melting time for a given value of current, but unfortunately there is no accepted industry-wide standard as t o the method of showing them, i.e., whether they should be plotted on t.he basis of short time, or minimum melting, maximum melting, or total clearing time. The various timecurrent characteristics of a fuse can be calculated from the given characteristic by adding or subtracting allowances, as shown in Fig. 9.9. If, for example, the characteristic of a power fuse is given in terms of minimum melting time, the maximum melting time can be determined by plotting another curve 20 per cent higher in current for each value of time t o allow for variations in manufacture of the fuse wire. Distribution-type fuse links have a smaller manufacturing tolerance. After the fuse melts, some time is required for the arc t o go out, and this must be added to the maximum melting time t o obtain the total clearing time curve. Another factor that must be considered in determining the over-all time-current characteristic of a fuse for coordination purposes is the damage tolerance (see Fig. 9.9). This is an allowance that must be made if the fuse is used ahead (on the power-source side) of some other shortcircuit protective device. If the fault is on the load side of one of these other devices, the latter must operate and clear the fault in less than the time shown by the damage-tolerance curve, in order t o avoid any possibility of overheating the fuse link sufficiently t o weaken i t and thereby eventually cause false operation. I n some cases the damage tolerance is minus 25 per cent in time for any given current, i.e., the protective device on the load side of the fuse should open and clear the circuit for any given value of current before that time is reached in order t o avoid overheating the fuse detrimentally. I n other cases, the damage tolerance is given in per cent current. Specific curves for the particular fuse involved should be obtained from the manufacturer. The short-time curves sometimes provided include the damage tolerance. The suffix “ E ” on fuse current ratings simply means that the fuse or fuse link is made to meet the 1944 EEI-NEMA Committee Standards for temperature rise a t certain loads, and the suffix“ N ” means that the fuse or fuse link conforms to the corresponding 1936 NEMA Standards. Both the N and E fuses will carry their rated current continuously. DIRECT-ACTING-TRIP DEVICES-GENERAL

A direct-acting-trip device is one that trips its circuit breaker by direct mechanical action when the flow of short-circuit current reaches a predetermined value, whereas a relay works through the medium of an a-c or d-c potential trip coil, or an undervoltage device, t o trip the breaker. Direct-acting trips may be operated by (1) an armature attracted by the


533

electromagnetic force crrated by the short-circuit current flowing through a series trip coil or (2) a bimetallic strip or the equivalent actuated by the heat generated by the fault current. The bimetallic strip is usually in series with the circuit on small breakers. On large breakers it may be heated by induction. The majority of direct-acting trips are used on low-voltage (600 volts and below) air rirruit breakers. However, they are also used on power (above 600 volts) circuit breakers in small installations as the only form of short-circuit protection and in conjunction with relays to avoid the use of a tripping battery. All direct-acting-trip devices are less accurate ill their time-current characteristics than relays. It might he possible to build such devices capable of developing sufficient force to trip their breakers directly and still have as high a degree of accuracy as is afforded by overcurrent relays, but the cost would not be justifiable on systems where this type of device is used. Inasmuch as their construction varies, the discussion mill he continued under the heading of the types of breakers with which they are used. Direct-acting-trip Devices on Power Circuit Breakers (above 600 volts). Direct-acting trips are commonly used on relatively low-interrupting-capacity power circuit breakers and are available when required on some of the larger breakers. They are electromagnetic, plunger operated, and may he either instantaneous or time delay. They are usually operated from the secondary of current transformers rather than being connected in series with the primary circuit. Instantaneous-trip direct-acting-trip devices may he used alone to trip their respective circuit breakers, but, for the most part, they are used in conjunction with time-delay overcurrent relays to trip the circuit breakers when there is no tripping batt,ery available. Time-delay direct-acting-trip devices have an oil dashpot with provision for adjustment of the time delay. Adjustment of the current pickup of either the time-delay or inst,antaneous-trip devices is accomplished by varying the position of the plunger in the solenoid. Direct-acting-trip Devices on Air Circuit Breakers (600 Volts a n d below). The direct-acting-trip devices on low-voltage circuit breakers (600 volts and helow) are electromagnetic. They are usually of hingedarmature construction actuated by t,he flow of current through coils in series with the circuit. Three tripping devices are provided on each three-phase breaker. They can he built with instantaneous and long- or short-time-delay trips. These different trips-can he used in a variety of combinations: (1) instantaneous, ( 2 ) long and short time-delay, (3) instantaneous and long or

534


short time-delay, (4) instantaneous and both long and short time-delay. A “family” of long- and short-time-delay tripping-device time-current characteristic curves for 15-, 25-, and 50,000-amp interruptingcapacity low-voltage air circuit breakers is shown in Fig. 9.10. The shape of different manufacturers’ direct-acting-trip curves differs somewhat, but the curves have the same basic characteristics as those shown here. I n the design of a tripping device whose curves are shown in Fig. 9.10, the current pickup is adjustable, but the operating time for different current values is not adjustable. I n other designs, both the current trip and the time delay may be adjustable. Tripping-device operation coils are available in a wide range of current ratings up to and including the maximum continuous-current-carrying capacity of the circuit breakers. This is 225 amp for the 15,000-, 600 amp for the 25,000-, and 1600 amp for the 50,000-amp interruptingcapacity (IC) breakers, respectively. The current rating of a circuit breaker is determined by its trip-coil rating rather than its maximum continuous-current-carrying capacity, or so-called ‘‘ frame size.” The current setting of long-time-delay trip devices is adjustable in the field, and pickup-setting calibration markings are a t 80, 100, 120, 140, and 160 per cent of the trip-coil rating. They can also be set a t any intermediate value between these calibration points. Some short-time-delay trip devices are also adjustable, in which case they usually have three current pickup settings marked on the calibration scale. The NEMA (National Electrical Manufacturers’ Association) standard calibration markings of 5, 756, 10 times the trip-coil rating are supplied unless another combination is requested. They can he had with ealihrations anywhere in the range of 2 to 10 times the trip-coil rating, as long as the desired maximum setting is not more than 255 times the minimum setting. Short-time-delay trips are sometimes supplied without provision for adjustment of the setting, because only one of a possible choice of pickup settings will coordinate properly with the characteristics of the tripping devices on the other hreakers in series with it in a factorydesigned selective-tripping system. The curves in Fig. 9.10 are plotted in “multiples-of-the-pickup setting” of the trip coil, e.g., if a 400-amp trip is set a t 80 per cent pickup, 1 multiple of pickup would be 320 amp, 2 multiples would he 640 amp, etc. Present NEMA Standards refer only to upper and lower boundaries of air-circuit-breaker tripping-device curves, and some published curves are plotted in that manner. The area hetween these NEXA limits is shown divided into it,s two components in the curves of Fig. 9.10, so that their significance could be explained in order to give the reader a better understanding of the operation of the device.


535

The solid-line curves represent the boundaries of the tripping-time zones of the respective devices. Points on the lower of these lines represent the minimum time required for the device to trip on a given value of current. Similarly, points on the upper curves show the maximum limit of the tripping time of the device plus the arc-clearing time of the breaker. These maximum and minimum tripping-time curves represent the manu-

FIG. 9.10 Typical time-current choracterirtic curves of long clnd short time-delay trips on6W-volt aircircuit b r e o k e n ( l 6 0 0 a m p and below).

536


facturing tolerances adopted by the industry to define the limits of acceptable performance of hreaker time-delay tripping devices. These tolerances include allowance for variations in magnet and armature structure performance due to machining and slight differences in magnetic material, plus oil-viscosity changes due to temperature. These two curves represent the extremes of acceptable performance, and the operating-time tolerance hand of a given tripping device will be much narrower. All mechanically operated devices have a certain amount of overtravel, i.e., they continue to move in the original direction of motion because of inertia after removal of the driving force. Also, less current is required to creat,c the same pull in an electromagnetic circuit when the air gap is reduced. The dash-dot (reset-time) curves in Fig. 9.10, which correspond to the lower boundary of the operating zones of the tripping device as defined by S E M A , make allowance for these factors in order to avoid false tripping of circuit breakers on the line side of another hreaker, as explained below. As soon as t,he current in the cirSERIES TRIP COIL ON ARMATURE POSITION cuit exceeds the pickup setting of BREAKER AT RESET TIME either a long- or short-time-delay AIR trip device, the armature starts to at a rate proportional to the move ARMATURE PoSITIoN NORMAL series-connected electromagnetic force trip exerted coil. At by the

w/

f

/ FIG. 9.1 1 Schematic diagram rhowing operation of 6Mlvolt air-circuit-breaker tripping mechanirm.

time indicated by the intersection of the dash-dot curves in Fig. 9.10 and a current-magnitude ordinate, the armature will have reached a part.ially picked-up position designated position at resettime”

in Fig. 9.11. When the armature of the tripping device reaches this position, the current in the circuit must he reduced immediately to 80 per cent or less of the pickup setting of a long-t,ime-delay (LTD) device, or 20 per cent or less of the pickup setting of a shortbtime-delay device. Otherwise the armature will continue to move and eventually trip the breaker. This continuation of movement in the closing direction from the reset-time position is caused by a combination of overtravel and the increased effectiveness of current in the series coil. The position of the reset-time curve shifts somewhat with variations in the magnitude of the current left flowing in the circuit after the major reduction in current is effected, hut it is impractical to try to make allowance for such-variations. The permissible 80 per cent current limit on the long-time-delay trips enables the hreaker t o carry some normal load current after the high overcurrent period has passed. The same is true of the 20 per cent limit


537

on the short-time-delay trip device. The latter, incidentally, is more generous than appears a t first glance, since minimum pickup on the commonly used short-time-delay devices is five times normal. To illustrate the meaning of the reset-time curves, assume that a 200amp 1C long-time-delay tripping device, as shown in Fig. 9.10, has its pickup set a t 100 per cent of the coil rating and that the current in the circuit is five times normal. I n order t o prevent false operation, the current must be reduced to 80 per cent of the pickup setting by the opening of some ot,her device, or by a reduction of the start,ing current in the case of motor-starting duty, a t not later than 5.7 sec; otherwise the device will continue to move and eventually trip the breaker, even though its minimum normal tripping time would be 11 sec with five times normal current flowing. When plotting coordination curves of time-delay air-circuit-breaker tripping devices, the area hetween the reset-time and the maximum-time curves (NEMA’S lower and upper boundaries, respectively) should he reserved for the tripping device of each circuit breaker. The long-time-delay and instant,aneous trips are usually combined in a common device. When this is done, the long-time-delay curves “blend” smoothly into the instantaneous-setting curves, as shown by the broken lines on the trip-coil curves in Fig. 9.10 and the typical combination of long-time-delay and instantaneous-trip curves shown in Fig. 9.12. The instantaneous curves are plotted in multiples of rated current and, therefore, are applicable t o any rating of trip coil available in the 15-, 25-, and 50,000-amp interrupting-capacity circuit breakers. It should he borne in mind, however, that the multiples of rated current times the current rating of the particular trip coil must not exceed the interrupting capacity of the breaker. The short-time-delay trip is usually a separate device and does not blend with the others. Consequently, when it is used in conjunction with a long-time or instantaneous trip, or both, there are sharp corners where its time-current curves meet the others, as in the typical combination of curves shown in Fig. 9.13. Instantaneous trips, when supplied alone, are generally adjustable in the field. Some of the instantaneous devices furnished in combination with time-delay trips are also adjustable, whereas others have a fixed setting a t perhaps 4, 8, 12, or 15 times normal, whichever is deemed best suited for the intended use of that particular breaker. DIRECT-ACTING THERMAL-MAGNETIC TRIPS ON MOLDED-CASE BREAKERS

A combination of thermal and instantaneous magnetic trip is commonly used on the so-called molded-cape low-voltage circuit breakers t o provide time-delay operation on moderate overcurrents and instantaneous opera-

538


CURRENT I N TIMES BQEDUER RATING

FIG. 9.12

Time-current curve of long time-delay and instantaneous trips on MX)-volt air circuit breakerr (1600 amp and below).


539

FIG. 9.13 Time-current curves of long and short time-delay and instantaneous trips on 600-volt air circuit breakers (1603 amp and below).

540



541

tion on high-magnitude short-circuit currents. I n some cases the thermal element only is used on small circuit breakers. The thermal characteristic is usually nonadjustable after installation, hut the iustantaneons trip is available in adjustable or nonadjustable eonstrnction. Figure 9.14 shows the a-c time-current tripping characteristic curves of the nonadjustable thermal trips in combination with adjustable instantaneous trips for the various ratings available in a 600-amp frame-sizemolded-case breaker. It will he noted that the minimum tripping-time curve of the timedelay thermal device, as shown in Fig. 9.14, is the same for all ratings, whereas the maximum tripping-time curve varies with the current rating of the trip unit. Arc-clearing time is included in both the time-delay and instantaneous-trip curves. Breaker trip units are usually supplied with the instantaneous trip set at the maximum position. Adjustment to other trip positions is made by means of a lever, or knob, provided for this purpose. Maximum and minimum tripping current positions are stamped “ H i ” and “Lo,” respectively. The maximum, or “Hi,” setting is the same for all ratings, usually ten times normal. The minimum, or “Lo,” setting varies with the breaker (trip-unit) rating as shown by the short heavy lines a t the left side of the magnetic-trip-adjustment range in Fig. 9.14. The two lighter lines on either side of the heavy ones show the manufacturing tolerance in the low-current-adjustment pickup. The instantaneous trip on the W a m p frame-size breaker of Fig. 9.14 is adjustable from 2.6 t o 10 times normal in seven uniformly spaced steps, including the “ H i ” and “Lo” positions. The number of steps varies with different breaker designs and frame sizes. The manufacturer’s operating-tolerance bandwidth decreases from plus or minus 25 per tent a t the “Lo” setting, approximately proportional t o the pickup setting, until it reaches the plus or minus 10 per cent allowable tolerance a t the maximum current (“Hi”) setting. Figure 9.14 shows how the “Lo” setting of a 600-amp breaker and the “Hi” setting of all ratings blend into the time-delay thermal-trip curves. Intermediate instantaneous-trip settings would blend in a similar manner a t whatever point they intersected the thermal-trip curves for that tripunit rating. A molded-case breaker having a nonadjustable instantaneous as well as thermal trip would have a time-current characteristic curve similar t o one of the combinations that might be made up from the curves of Fig. 9.14.

542


FIG. 9.15 Schematic diagram of typical industrid power system showing possible choice of fault proteclive devices.

DeRnition of Device Numbers and Abbreviations

50 51 51 * 51N 51G 67 67N 87 87G ACB LTD STD Instant

MCB

Inrtmtmeous overcurrent relay [phase protection) Time-delay overcurrent relay (phore protection) Time-delay generator overcurrent relay with voltage restraint Time-delay overcurrent residually connected ground relay Time-deloy overcurrent ground relay in neutrol circuit Directional overcurrent relay (phase) Directional overcurrent relay (ground) Differential relay Generotor differential ground relay Air circuit breaker Long time-delay direct-acting trip Short time-deloy direct-acting trip Instantaneous direct-acting trip Molded-care breaker with nonadjurtoble thermal and instantaneous trip


543

TYPICAL INDUSTRIAL POWER SYSTEM FAULT-CURREN1 PROTECTIVE-DEVICE LAYOUT

The schematic diagram of a typical industrial plant electric power supply and distribution system shown in Fig. 9.15 with the fault-current protective devices that might he used on it illustrates the use of the various relays, fuses, and direct-act,ing trips described in this chapter. In accordance with usual practice, the protective devices are identified on the diagram by means of standardiaed device function numbers. For convenience, the numbers that are used on the diagram are identified in the tabulation helow it. A complete list of the industry standard ( M A ) device function numbers is given in the Appendix. This plant has some generators of its own in addition to an incoming circuit from the local power company. This necessitates the use of directional relays for selective fault detection on the incoming line and also introduces another prohlem, namely, how t o disconnect the local plant’s system from the power company line in the event that something goes wrong on the power company’s system resulting in the two systems being split apart, with some of the power company’s load being fed from the industrial plant generators. This condition, as well as the case where t.he local generation is inadequate to carry its own plant load in case the incoming line is disconnected, can cause serious trouble by overloading the local plant generators to such an extent that the whole system collapses and shuts down the plant. The possihility of such a collapse can be prevented by the use of underfrequency relays and other devices whose characteristics and use are outside the scope of this book. COORDINATION OF PROTECTIVE DEVICES The final step in the creation of an adequate a-c short-circuit protective system is to be sure that the various devices are selective in their operation, i.e., coordinated nith each other. WHAT IS COORDINATION?

On all but the simplest systems there will be two or more circuit breakers, or other circuit interrupting devices, hetween a fault and the source of power. In order to localize the disturbance as much as possible, these devices should be selective in operation, so that the one nearest the fault on its power-source side nil1 have the first chance to operate. If for any reason this protective device fails t o function on schedule, the next device in the chain must be ready to take over the task of opening the circuit, and so, in successive steps t o the power-source circuit breaker

544


itself, if necessary, to clear the fault. T o accomplish this objective, the faulecurrent protective devices must have been selected, as in the case of fuses and instantaneous trips, and so forth, or be capable of adjustment to operate on the minimum current that will permit them to distinguish between fault current and perniissible load-current peaks and to function in the minimum possible time and still be selective with others in series with them. When these two requirements are met, the damage to equipment, or the interference with production due to loss of power during a short circuit, or both, will also be a t a minimum. On very simple systems with but one fault-current protective device between the power source and the load, there is no coordination problem, but i t is still necessary to choose a current setting compatible with the load characteristics. All the adjustable devices must be set in the field to achieve the desired coordination. Following is a detailed discussion of some of the basic fundamental procedure involved in making the necessary study of an a-c power system to determine what the current and time settings of the adjustable devices should be, assuming that they and the nonadjustable devices have been selected correctly for the application. DATA REQUIRED FOR A RELAY-SETTING STUDY

'

An overcurrent protective system is simply a multiplicity of coordinated individual devices, i.e., fuses, direct-acting trips, and plunger or induction-t,ype relays. Therefore, an understanding of their individual behavior, in so far as it affects their coordination with other devices, is a logical starting point for a coordination study; frequently termed a relay study. The necessary data on their individual characteristics were given earlier in this chapter under the heading Types, Operat,ing Characteristics, and Uses of Protective Devices. The next problem is to secure the necessary data from which specific relay or other overcurrent-proteetive-devieesettings may be determined. Following is a list of the basic information needed for this purpose: 1. A one-line diagram of the power system involved, showing the type and rating of the protective devices and their associated instrument transformers and t,he impedances of all transformers, rotating machines, and feeder circuits. 2. Maximum and minimum values of short-circuit current that are expected to flow through each protect,ive device whose performance is to be studied under varying operating conditions. These data can be obtained from a short-circuit study based on the information contained in the complete detailed diagram of item 1.


545

3. Starting current requirements of motors, and so forth, and their maximum peak-load current. 4. Manufacturers’ characteristic performance curves of the relays, trip coils, and fuses to be coordinated. 5. Manufarturers’ performance curves of current transformers, especially bushing types. 6. Any special overcurrent protertive requirements such as those stipulated by the National Electrical Code or dictated by the load characteristics. 7. Any special overrurrent-protective-device setting reqnirements stipulated by the public-utility company with which the industrial plant may be interconnected. 8. Decrement curves showing the rate of decay of the fault current supplied by generators. SHORT-CIRCUIT-CURRENT CALCULATIONS FOR A RELAY STUDY

The basic data and procedure required for making short-circuit calculations as a preliminary t o a relay study are the same as for a circuitbreaker-duty study, but the combinations studied will be somewhat different. In a relay study the problem is t o determine the characteristics of devices that will (1) be sure to operate on the minimum values of fault current expected at certain fixed times following the instant of short circuit, and (2) be selective in their sequence of operation over the range between minimum and maximum values of short-circuit current, so that the relay or other device nearest t o the fault will be the first to operate. The magnitude of the short-circuit currents, which will determine the settings of the overcurrent protective devices, should be calculated on the basis of the fault current from any power-company ties, plus that contributed by all rotating machines directly connected to the local power system, that is, the user’s own power generation and distribution system. It is, of course, unlikely that every one of the local machines will he in operation simultaneously under normal conditions, but it could be the case during transfer periods when load is being switched from one machine to another; hence the relays, and so forth, must be capable of giving the desired sequential operation on the maximum possible magnitude of current. Usually the fault-current contribution from the power company can be calculated on the Basis of a single reactance value because the relatively high reactance of the tie circuit supplying the industrial plant is sufficient to limit the fault current so much that the effect of differences between maximum and minimum generating capacity and subtransient X y and transient X : reactances are unimportant. Also the decrement effect will

546


he eliminated for all practical purposes, see curve 4, Fig. 9.16. If, however, the tie circuit hetween the power company and industrial plant is of large kilovolt-ampere capacity and relatively low reactance, variations in power-rompany system reactance will affect the magnitude of fault current on the industrial system sufficiently to require taking them into account. I t would be advisable in all cases t o check the effect of any variations in power-company reactance until enough experience has been accumulated t o warrant deciding by inspection whether or not their effert would be significant. EFFECT O F FAULT-CURRENT DECAY DUE T O GENERATOR-CURRENT DECREMENT

O N RELAY PERFORMANCE

As shown by the upper curve of Fig. 9.19 and curve 1 of Fig. 9.16, the fault. current supplied by generators to a local bus, where the total reactance is relatively low, decreases very rapidly and reaches steady-state short-circuit current in 0.5 to 1.0 see for the usual industrial-plant-size generators. The operating time of all time-delay overcurrent protective devices is affected by the decay of fault current due to this current decrement of the rotating machines supplying the short-circuit current. The decrement can be neglected when selecting settings for feeder relays, or time-delay trips, if the steady-state short-circuit current will be sufficient t o operat,e them and the number of relays in series is small so that the accumulated operating-time error will not he objectionable (see discussion relative to Fig. 9.20). I t must he taken into account, however, when (1) setting a generator overcurrent relay with voltage restraint (Fig. 9.20) and (2) setting relays or circuit-breaker trip coils on small systems with very limited generating capacity. In case 1 an actual decrement curve is needed to determine the settings of the generator and feeder relays. I n case 2 i t may not be necessary actually to construct a decrement rurve or curves, hut their general behavior and effect on the magnitude of fault current must he visualized mentally and taken cognizance of t o he sure that the protective devices will operate on the current available a t the time they are supposed t o function. For example, the decrement factor must be given consideration in those cases where a relatively small generator unit supplies power to a large system over week ends. On such a system it is quite possible that the available fault current will he insufficient t o operate the relays on large feeders, in which case a practical solution would be t o let the generator relay trip the generator circuit breaker and shut down the whole system in case of trouble on such feeders. It is particularly important to keep the current decrement in mind when


547

selecting the devices, as well as their settings, for a system supplied from one or two engine-driven or geared-turbine-driven salient-pole generators, because even their initial fault current is quite low herause of their high inherent reactance. It will he noted from the curves in Figs. 9.16 and 9.19 that the presence of a voltage regulator and the addition of evternal circuit impedance have 40

30 25

KV BUS AT POWERHOUSE NO 2 FAULT AT SUB 3- LINES I 8 2 IN PARALLE NO 3 FAULT AT S U E 6 - 2 LINES PARALLELED NO 4 FAULT AT SUE 9 - I L I N E ONLY IN USE

20 15 10 9 E 7 6 5 4

3 2

m 0

!

1.0 1.5 2.0 SECONDS T I M E AFTER SHORT CIRCUIT 0 20 40 60 60 00 120 140 CYCLES FIG. 9.16 Short-circuit decrement curves on 33-kv system supplied by two 10.000- and four Moo-kw hrrbine-generator-h~nrformerunits.

0.5

548


a pronounced effect on the magnitude of fault-current contributed by generators. The amount of external impedance in the circuit between the central station generators and the industrial plant is usually sufficient virtually t o wipe out the decrement on short-circuit current from that source. Therefore, the current will be maintained a t the magnitude determined by the system reactance.

SHORT-CIRCUIT CURRENTS NEEDED TO PREDICT THE OPERATION OF VARIOUS OVERCURRENT PROTECTIVE DEVICES

Instantaneous Induction or Plunger-type Relays and Circuit-breaker Trip Coils. All these devices are responsive to direct current as well as

alternating current and are fast enough t o operate on the first half-cycle of fault current. Therefore, their operating current will he the initial (instantaneous) asymmetrical fault current contributed by all rotating equipment (including induction motors and generators) calculated on the basis of its suhtransient reactance ( X y ) . Use a 1.6 multiplier t o obtain the magnitude of the asymmetrical or offset current, except on systems of 5 kv and below when there are transformers or appreciable line resistance between the point of fault and the source, in which case use a 1.5 multiplier. High-speed Induction-type Relays. Since these relays operate in three cycles or less, the initial X y symmetrical fault current may also be used as their operating current. Actually, the fault-current magnitude will drop during the three cycles because of the generator-current decrement, but, the influence of this reduction can generally be neglected in industrial-system relay studies. Time-delay Relays and Direct-acting Trips. Generally speaking, the devices of this class are too slow t o be appreciably affected by the subtransient reactance Xl: values of fault current. Therefore, their maximum operating current will be the initial value of symmetrical transienb reactance ( X i ) current supplied by the power company and the local plant’s synchronous equipment (generators, condensers, and motors). Current-time settings of such devices that permit operation in approximately 0.1 sec (six cycles) or less on high-current magnitudes are theoretically an exception to this rule, hut practically, such relays usually are the first in a series of relays (farthest from the power source), so that any reduction in their operating time simply increases the time margin between that relay and the one ahead of it. Unless very accurate relay operating time is needed, the reduction in fault current acting on it due to generator decrement can be neglected. The procedure involved and the effect on relay operating time, when


549

allowarce is made for the decay in generator current, are discussed in detail later in this chapter under the heading How to Make Allowance for Current Decrement on Operation of Feeder and Generator Overcurrent Relays. Fuses. The performance of fuses is an inherent characteristic of the particular unit involved and must be taken into account when they are selected for the job. Nevertheless, the following discussion of expected fuse performance seems desirable a t this point to afford a better understanding of the problem of coordinating them with other devices. Fuses of all types depend on the beat created by the passage of the fault current through them to melt their fusible elements. Since the heat is due solely to the P R loss, d-c as well as a-c components of current will affect them; so the initial asymmetrical subtransient current will affect fuse operation. However, the effect of the d-c offset current decreases in importance very rapidly as the expected fuse melting time increases. The melting time of fuses varies from less than 0.01 sec to several hundred seconds, depending on the fault-current magnitude. For example, the lOOE current-limiting fuse in Fig. 9.9 has a melting time of 0.01 sec (0.6 cycle) a t 2700 amp, 0.1 sec a t 780 amp, 10 sec at 320 amp, and 300 sec a t 200 amp. For good performance on a system which has most of its power locally generated, the fuse minimum melting-time curves should show that the initial magnitude of unrestricted (no intentional limitation) fault current under minimum power-supply conditions is sufficient to blow the particular fuse within 1 or 2 cycles (0.033 sec). If it cannot do that, the use of fuses should be carefully scrutinized because the rapid generator-current decrement may reduce the fault current so much that the fuses either might not blow a t all, or else might be objectionably slow. Also there would be no margin t o allow for any current limitation due to fault impedance. An exception would be the case of a fuse a t a location where the fault current was reduced substantially by high circuit impedance, as shown in curve 4 of Fig. 9.16. Under such a condition the difference between the initial symmetrical X y current as represented by the curve and the initial X : value of current as represented by the “dot” is negligible, and so is the decrement effect. Therefore, if the X i current is sufficient to blow the fuse in a few cycles, the application is satisfactory. Modern power systems are often designed to have quite low values of line-to-ground fault current. In such cases, i t will often be found that the available ground-fault current will be so low as to make large fuses either extremely slow or inoperative. When such a condition is encountered, additional means such as relays and circuit breakers must be provided for opening the circuit under line-to-ground fault conditions. There are certain types of low-voltage fuses that are purposely designed

550


t o be slow on low magnitudes of fault current in order to ride through high normal-load peak currents. Such fuses, however, have normal fastmelting characteristics on high-magnitude fault currents. SETTING OF RELAYS AS AFFECTED BY THEIR FUNCTION

When the short-circuit currents are all calculated, the next step is to start at the load end of each circuit and work back toward the power source. The minimum current setting of the relays must be high enough to carry normal load swings and yet low enough t o be sure that they will operate positively on the minimum expected short-circuit current. The latter is the magnitude of current ralculated with minimum power supply available and assuming a solid short circuit with no allowance for any current limitation due to impedance in the fault itself. Normally, there will be plenty of margin between the minimum shortcircuit-current level and the relay setting dictated by the maximum permissible load. Occasionally, however, that will not be true, such as in the case of high-reactance generators whose total output is transmitted over a single line. Such possibilities must be recognized and provided for initially when designing the protective system by specifying devices that do not depend on current magnitude alone for operation and selectivity. The following general rules will serve as guides in determining the minimum relay current settings that will not trip on permissible load currents. Relays on Transformer Feeders. The National Electrical Code states that time-delay overcurrent relays on a feeder for a single transformer should he set not more than 2.5 times the full-load rating of the transformer for units without secondary circuit breakers and up to six times transformer full-load current if there is a secondary circuit breaker and the transformer reactance is around 0.06 per unit. The fact that the Code permits it, however, does not mean that the relays always should be set at these upper limits. If the load on the transformer is diversified with no relatively large motor units whose starting current might be the limiting item, a relay current setting of 1.5 times transformer rating is often sufficient to enable time-delay relays t o carry normal load-current swings. Ordinary time-delay overcurrent relays are too slow t o be operated by the transformer magnetizing current inrush. If there are several transformers on the feeder without individual primary-side fault protection, a relay current setting corresponding t o 1.5 times the total full-load rating of the several transformers should be


55 I

ample, except in those cases having large individual motors. However, the setting selected should not he more than the six times full-load current of the smallest transformer as required by the National Electrical Code. If there are several transformers on a main feeder, each of which has its own primary protection, this Code rule does not apply and the main feeder relays should he set t o coordinate with the transformer primary protective devices. An instantaneous relay in the primary circuit of a single transformer with an instantaneously tripped secondary circuit breaker should he set a little above the asymmetrical value of the fault current in the primary during a short circuit close t o the transformer secondary terminals. This setting is usually high enough to clear the magnetizing current inrush of the transformer, although the latter may theoretically be as much as 15 times normal for a few cycles. If the primary feeder circuit breaker trips when the transformer is energized, the relay current setting should be increased slightly until the circuit breaker will stay closed. The minimum setting of the instantaneous relay in a high-voltage feeder with several transformers on it will quite likely be dictated hy the need for riding over their total magnetizing current inrush. As stated, this will be in the neighborhood of 10 to 15 times normal, which should he high enough to clear any probable condition such as a short circuit on the secondary of one or two units and normal load current on the others. Relays on Single-motor Feeders. The minimum setting of a timedelay relay intended t o provide short-circuit protection for a motor feeder has to he above the maximum motor-starting current so that the relay will not operate on the latter. This is necessary because any time-delay overcurrent relays that are expected to afford worthwhile fault protection are much too fast to ride through motor starting current. The longtime-delay induction relays that can be set lower than this in current and still he able to ride through the starting period afford a measure of normal overload (overheating) protection hut are much too slow to provide adequate short-circuit protection. Instantaneous relays, usually plunger type, with a current range of 10 t o 40 or 20 to 80 amp, normally are used to provide fast short-circuit protection in case of a fault in the motor or its leads. These relays should not operate on the short-circuit current which their particular motor will cootribute to a fault elsewhere on the system. Therefore, since they are fast enough to operate on the first half-cycle of short-circuit. current and also are affected by the d-c component,, it is necessary t o set them slightly above the initial asymmetrical fault current which their motor can supply. This current, which-may he slightly higher than t,he full-voltage lockedrotor starting current, can be determined by dividing the full-load current

552


by per unit subtransient reactance ( X y ) of the motor (either synchronous or induction) and then multiplying by the factor of 1.6 to account for the d-c component. Relays on Incoming lines and Feeders with Miscellaneous load. The minimum setting of a time-delay overcurrent relay on an incoming line or feeder with miscellaneous load, which may include one or more motors, should be just above the expected peak load on the circuit. This usually itill be the total of the starting current of the biggest motor plus full load on the other circuits. This will hold unless it is intended to start more than one motor a t a time normally or the motor controls are so designed that the motors would restart automatically when the circuit is reenergized following an outage. Instantaneous relays may not be usable on such circuits because of inability to coordinate them with other instantaneous relays in the branch circuits, as previously discussed. If they can be used, they should be set a little above the total of initial asymmetrical short-circuit current which all the motors could contribute to a short circuit elsewhere on the system, plus full-load current on t,he nonmotor load. It is unlikely that any normal load snrges would exceed this value. Residually Connected Ground-fault Relays. Residual ground-fault relays are connected in the Y of three current transformer secondaries where they see only the unbalanced residual current flowing during ground faults (Fig. 9.4). Under normal balanced three-phase load conditions, no current will flow in the residual relay. Therefore, it can be set at low current levels to give sensitive operation on ground fanlts, which incidentally are the most common variety. Such relays are normally rated 0.5 to 2.0 amp or 1.5 to 6.0 amp. These relays will operate satisfactorily over their entire tap range when connected to high-accuracy current transformers, but they may not on the 0.5-amp tap when connected to the low turn ratio taps of some bushing current transformers. This is due to the fact that the bushing-current-transformer ratio breaks down because of the high volt-ampere burden imposed by the relay on the low-current tap. If a check of the bushing-current-transformer performance curves with the expected volt-ampere burden indicates that the relay may not operate, it will be possible usually to obtain satisfactory operation by using a higher current tap on the relay or bushing current transformer, or both. Ground relay current and time settings are handled in the same manner as phase relays. However, ground-fault current will be of different magnitude than the phase-fault current, and it may not appear in all parts of the system since a transformer in the circuit blocks its flow. Also its source may be different; for example, a grounding transfomer might be used to ground a system supplied by a delta-connected generator.


553

Frequently ground-fault relays can be set to operate faster than the phase relays a t the same location because the latter must coordinate with other phase overcurrent protective devices on the secondary of transformers, for example, whereas a residual ground relay on the primary will not see ground-fault current in the secondary system. Residually connected ground-fault relays (Fig. 9.4) are subject to possible false operation because of the current-transformer error currents during a phase fault. This possibility is caused by the fact that duplicate current transformers are not exactly alike, even though made from the same batch of material. Although the variations are small, they are enough t o cause appreciable amounts of unbalanced current to flow in the Y of residually connected current transformers during severe phase-phase or three-phase short circuits. Such unnecessary operation of the ground relay is of minor importance if the ground and phase relays involved control the same circuit breaker. I t might be quite troublesome, however, if the ground relay tripped a main-line circuit breaker, thereby shutting down several branch feeders when only one of the latter was in trouble. One possible solution t o the error-current tripping problem is to design the system t o have sufficient ground-fault current to ensure positive tripping with the ground relays set above the expected maximum currenttransformer error current. Unfortunately, however, available data regarding current-transformer saturation indicate that the possible magnitude of error currents due t o differences in ratio breakdown caused by unequal saturation because of differences in their burdens, combined with d-e saturation and the differences permissible under manufacturing tolerances, may he as much as 30 to 40 per cent or even more. Consequently, this solution would require increasing the ground-fault current to several thousand amperes, resulting in far greater damage during a ground fault. The other and better way is t o keep ground-fault current and relaycurrent settings a t a minimum consistent with positive low-current relay operation and then prevent false operation by timing the ground relay to be selective with any phase relays on its load side. This, of course, raises the question as to which causes more damage, high ground-fault current and less sensitive relay pickup with slightly shorter relay operating time or minimum ground-fault current and sensitive relay pickup but maybe one step (0.4 sec) slower operating time. It is a difficult qnestion t o answer, hut probably the odds are in favor of the latter combination. If there are single-phase loads connected from line to neutral, any expected unbalance in current must be taken into account when setting ground relays, whether they are connected residually (Fig. 9.4) to feeder current transformers or t o the secondary of a current transformer in the neutral of a generator or transformer (Fig. 9.15).

554


Ground-fault Relays in Series with Generator or Transformer Neutral.

It is general practice to connect a final backup ground-fault relay to a current transformer in the generator or transformer neutral earth connection. A low-current relay and a current transformer whose current rating is 25 or 30 per cent of that of the neutral grounding resistor for that unit is the combination generally used. This combination, which affords maximum sensitivity and protection, is permissible since current flows in the neutral of a balanced threephase three-wire circuit only during a ground fault. The relays usually furnished for this purpose are capable of being set slow enough to permit the flow of ground-fault current up to the thermal limit of the neutral grounding resistor, if there is one, and many users set the neutral relays todo that. Sinceitisfairlygeneralpresent-day practice t o ust 10-sec rated resistors, the relay would have t o be set for 7 t o 10 sec in such cases. Such a slow setting is probably a carry-over from former days when it was standard practice t o use a 60-sec or longer rated resistor. The accompanying relay system would (1) sound an alarm immediately in the hope that the operator might be able to clear the fault soon enough t o avoid system shutdown by manually tripping a circuit breaker which had failed to open, and (2) if necessary trip the main generator or transformer circuit breaker in time to avoid dangerous overheating of the resistor. I n actual practice i t is doubtful that the operator was able to do much in the 60 see allowed, and he could accomplish even less in a 10-sec resistor-heating-time allowance. Another method of setting the neutral relay is t o time it to coordinate with the other system ground relays, that is, one time step slower. This seems preferable, since experience indicates that if a relay or circuitbreaker trip coil is going t o work a t all i t will do so in its normal time, which means that any further delay simply means more damage caused by the ground fault. COORDINATION OF RELAYS IN SERIES

Time-delay Relays. As mentioned previously, the time-current curves of direct-acting time-delay trips, fuses, and time-delay thermal devices include the necessary allowance for overtravel, manufacturing tolerances, and so forth. The time-current characteristics of relays, on the other hand, are represented by families of single-line curves (Fig. 9.2) to which tolerance bands must be added. Most relay time-current curves begin a t 1.5 multiples of minimum closing current or pickup setting because their performance cannot be predicted too accurately below that value.


555

However, curves showing expected time-current performance down t o 1.1 times pickup usually can be obtained if required. This time margin or tolerance band is based on the fact that the second relay in a “chain” of relays continues t o see fault current until the circuit breaker associated with the first relay has opened and the arc has been extinguished. This is nominally eight cycles for the circuit breakers commonly used in industrial systems, although actually the opening time will be four to five cycles. Then after the first circuit breaker has opened the circuit and deenergized the second relay, the latter’s contacts will continue to close for approximately 0.1 sec because of the inertia in the induction disk to which the movable contact is attached (Fig. 9.1). A minimum total time margin of 0.4 see with maximum fault current flowing should be sufficient to afford satisfactory selectivity between inverse-time relays. This margin allows for the 0.13-sec circuit-breaker opening time (eight cycles), 0.1-sec overtravel, and a safety factor of 0.17 see t o cover manufacturing variations and inaccuracies in positioning of the time dial or lever when setting the relay. The minimum time margin between very inverse-time and extremely inverse-time relays can be reduced t o about 0.33 see because their overtravel is less and a shorter safety factor is needed since the difference in relay operating time caused by changing from one timedial setting t o another is also shorter on these higher speed relays, so that the positioning error is reduced. If the total accumulation of even these short time intervals makes the operating time of the last relay (one nearest the source) undesirably long, the total time can be reduced a little by very careful checking of relay operating time. Generally speaking, however, the fraction of a second saving that might be made in the time margin between these values and the minimum possible settings that would not cause false operation is not of sufficient importance on industrial power systems to warrant the extra trouble involved in making the settings. When selecting current-tap and time-dial settings for induction relays, it should be borne in mind that, when two relays in series are set t o coordinate properly under the maximum value of fault current expected, they will always be satisfactorily selective on lower values of current if they have the same shape time-current curves and the current setting of the slower relay is equal to, or higher than, that of the faster relay. If the current setting is lower, the operating-time curves of the two relays will cross each other a t some low value of fault current, and the slow relay will beat the fast one for all currents below that value (curves 2 and 3, Fig. 9.17). Another factor t o be considered, when choosing between two combi-

556


nations of current-tap and time-dial settings, either of which will give a desired operating time with maximum fault current flowing, is that the combination with the lower current and higher time-dial setting is usually preferable. The reason is that such a setting will be more sensitive and faster on low values of short-circuit current (curves 1 and 2, Fig. 9.17). For example, suppose an operating time of 0.5 sec is desired with a relay connected to 1000/5-amp current transformers in a circuit with maximum symmetrical fault current of 20,000 amp. Relays with 6-amp tap and 2.1 time-dial or 10-amp tap and 1.7 time-dial settings will both give the desired time. But in case of a fault involving only 3000 amp, the relay with the 6-amp setting would operate in 1.25 sec compared with 2 sec for the 10-amp combination. If the current is still further reduced to 2000 amp, the first relay will still operate in 2.1 sec, but the second one will be very, very slow, since the current is only 1.0 times relay pickup, at which point operation is uncertain. 901 TD = R E L A Y T I M E DIAL SETTING

250 MVA AVAILA0LE

'0

1500 2000

3000

5000

10,000

20,000

40,000 6Q

AMPERES AT 2400 VOLTS

FIG. 9.17

Examples of incorrect coordination of induction relays.

0


557

Instantaneous Relays. When two instantaneous relays are used in series, their selectivity is dependent solely on their current settings. Therefore, they must be set so that the second relay will not pick up on the maximum asymmetrical current which the first relay (nearest the source side of the fault) can see. Such being the case, there must be sufficient impedance in the circuit between instantaneous relays in series t o create the necessary current differential. If this ohjective cannot be attained, the second relay must be rendered inoperative by short-circuiting its coil or opening its contacts. Usually the impedance of a transformer is sufficient to permit coordinating an instantaneous relay on the high-voltage feeder panel with the instantaneous trip coil of the low-voltage secondary circuit breaker. Also, the reactance of open transmission lines is frequently sufficient to provide the necessary differential in short-circuit-cnrrent magnitude t o permit the use of instantaneous relays a t both ends. Instantaneous relays cannot be coordinated on ordinary-length cable systems (2400 volts and above) because the circuit impedance is too low t o cause the necessary current differential. On such circuits it is preferable t o use the instantaneous relays on the branch feeders a t the receiving end of the main feeder cable (Fig. 9.15). This will afford fast selective tripping of the branch circuit in trouble, whereas using the instantaneous relay on the main feeder circuit hrea,ker a t the bus would shut down all the branch circuits regardless of fault location. Ordinary instantaneous overcurrent relays should not be used on low-impedance ties between two buses with power sources connected to each. Incidentally, the fact that a relay-setting study reveals that some of the instant,aneous relays must be made inoperative should not be interpreted as a sign of a poorly designed protective system. This is so, because it is quite common practice to include instantaneous attachments on all timedelay overcurrent relays on switchgear equipment so that the relays will be interchangeable. GRAPHICAL PROOF OF DEVICE COORDINATION

I t is always advisable t o plot the curves of relays and other devices that are t o operate selectively in series, because rarely will all the fault protective devices involved have the same shape time-current curves, and it is difficult to visualize mentally the relationship of the many different shapes of curves. They should be plotted on a single sheet of graph paper, using a common current scale. It seems t o work out best to use a scale corresponding t o the currents expected a t the lowest voltage level; for example, for faulhcurrent protective devices on both sides of a 2400-480volt transformer plot everything on a 480-volt current scale. To plot

558


2400-volt device time-current curves on the 480-volt scale, first determine the desired time and current settings in the usual manner on the basis of current expected in the 2400-volt circuit. Then plot time directly since that scale is unchanged, but multiply the 2400-volt currents by 5 (ratio of voltages) before plotting on the 480-volt scale. Preferably the curves should be plotted progressively as each circuit is studied, starting with the device a t the end of the chain (farthest from the source). This procedure will show whether or not the proposed timecurrent characteristic of each successive device coordinates with the one on its load side. Figure 9.17 illustrates the importance of doing this by revealing two conflicts that otherwise might have escaped notice. I n the case illustrated, it was assumed that the power supply was capable of maintaining a 250,000-kva short-circuit level without appreciable decrement and that there was no pump-hack short-circuit-current contribution from synchronous equipment on the 2400-volt system. On this basis, the maximum symmetrical fault current would he 20,000 amp on the 2400-volt system and 10,250 amp on the 13.8-kv system (60,000 amp on a 2400-volt base). I t was also assumed that the end relay in the chain ( D ) must be set a t a minimum of 0.5 see. The three sets of 2400-volt relays were coordinated by selecting timecurrent settings that would make their operating times 0.4 sec apart at the maximum current of 20,000 amp. Then, the single set of relays on the 13.8-kv system was coordinated with those on the 2400-volt system on the same value of fault current (3480 amp a t 13.8 kv), because that is all that the relays at A would see during a fault on the 2400-volt side of the transformer. This was accomplished by selecting time-current settings that would give 0.4 see between relays A and B for a 20,000-amp fault on the 2400-volt system. This procedure would give satisfactory results without plotting the curves, if the basic rules for coordinating relays were observed: (1) use relays with the same shape curves in series with each other, and (2) make sure that the relays farthest from the source of power always have current settings equal to, or below, that of the relays ahead of them. Unfortunately, however, it is easy to overlook one, or both, of these basic rules, and that is when the effort required t o plot the curves proves worthwhile. As shown in Fig. 9.17, time-current settings on relay D that would give either curve 1 or curve 2 would satisfy the requirement that it take a minimum of 0.5 sec a t 20,000 amp. Curve 2 setting, however, is slower and less sensitive than that represented by curve 1 throughout most of its range. Furthermore, curve 2 crosses curve 3, representing the desired setting for relay C, thereby compelling desensitization of C to make i t selective with D. Thus curve 2 setting on relay D would mean that much poorer short-circuit protection could be provided by either C or D.

SYSTEM OVERCURRENl PROTECTION

559

Curves 4, 5 , 6, and 7 illustrate what would happen if relay B had a very inverse-time characteristic instead of an inverse-time curve, as the others do. Curve 4 meets the requirement that i t be 0.4 sec slower than curve 3, representing relay C, when both ere operating on 20,000 amp. Also curve 6 satisfies the requirement that relay A be 0.4 sec slower than B when A is operating on the equivalent of the 20,000-amp 2400-volt system short-circuit current. If the curves had not been plotted, there would be reason to believe that the contemplated settings for A and B as represented by curves 6 and 4 would be satisfactory. Actually, however, the very inverse-time characteristic of relay B causes its curve to cross that of A a t a high level of fault current, which would mean that the tripping sequence of the breakers would be reversed. For this particular circuit, that would not be too serious, since tripping either breaker would shut down the whole circuit, but it would still nullify the effectiveness of the relays in giving indication as t o where the trouble was. If it was necessary to retain the very inverse-time relay a t B , the setting of relay A would have to be desensitized and increased in time, as shown by curve 7, t o he selective with B. This would presumably result in greater damage during a short circuit; so it would be better to substitute an inverse-time relay for the very inverse-time relay a t B , thereby making it possible to se: B to give performance as shown by curve 5. This would mean that A and B could both be more sensitive and faster and, consequently, could afford better protection for the system. I t should also be noted that, if the very inverse-time relay was used a t B , the backup protection that relay B could afford t o C would beverypoor indeed, because of the big “gap” in the pickup currents (sensitivity) of relays B and C , as shown by curves 3 and 4. EFFECT O F CURRENT-TRANSFORMER SATURATION ON RELAY BEHAVIOR

Industrial plants have a considerable number of relatively small feeders operating at the main-bus voltage, which may be subjected to high levels of fault current, especially when there are “local” generators. Consequently, some of the current transformers in such plants may be subjected to symmetrical short-circuit currents as high as 200 times their rating, and multiplies of 50 t o 100 times current-transformer rating will be quite common. U hen subjected to such magnitudes of fault current, most curient transformers will saturate, and then their accuracy breaks down badly. The extent of this breakdown in current-transformer ratio depends on the burden imposed by the impedance of the relays, instruments, and leads connected to their secondaries and the primary current imposed on them. The available data are not conclusive, hut it apppars that the ratio can break down so badly that only 50 per cent, or even

560


less, of the theoretically correct secondary current will be available (Fig. 9.18). Such a breakdown in current-transformer ratio means that relays connected t o them will not operate so fast as their time-current curves indicate they should, assuming perfect performance of the transformers. This factor should be taken into account when coordinating relays in series that are subjected to the same fault current, but connected to different size current transformers. For a given fault condition the smaller transformers would be subjected t o a greater ndmber of multiples of their rating, and if their burden was the same as that on the larger current transformers, they would probably show a greater breakdown in transformation ratio. The relays connected t o them would, therefore, see fewer multiples of their tap setting and be slower in operation than they would have been if the current-transformer ratio had been maintained. Meanwhile the relays connected t o the larger current transformers would be seeing more current in proportion and consequently would be operating nearer to their theoretical time-current curves, thereby reducing the time margin needed for selectivity between the relays. The possibility of nonselective operation from this cause should he investigated when it is important to have selective operation of relays connected to widely different current transformers or having substantially different burdens on their respective current transformers.

FIG. 9.18 Approximate performance of 4@3/>8W/5-amp wound-type current transformers with varying secondary burdens and primary overcurrentr.


561

This problem is less likely to he encountered on power company systems, hecause a t the generating stations where the short-circuit-current levels are high, the voltages are higher than in industrial plants, meaning less current for the same short-circuit kva. Also, larger blocks of power are being dealt with on the average utility system; so the current transformers are of higher ampere rating. These two factors comhined mean fewer multiples of their rating when current transformers are subjected to short-circuit current. On those portions of utility systems where the blocks of power are smaller, requiring smaller current transformer, the fault-current magnitudes are also greatly reduced by system impedance, so that the current transformers are not subjected to such high multiples of their rating. However, power company st,ation auxiliary power systems, when supplied by “house” generators rather than transformers, will often encounter the same problem. In many cases it will be possible to ignore the effect of current-transformer saturation, but it should be given consideration when trying t o make closely coordinated relay settings on large systems. I t will be necessary t o obtain the required current-transformer-performance curves from the manufacturer for the specific case involved, because a t the present time the published ratio-correction-factor (RCF) curves stop a t 20 times current-transformer rating. H O W T O M A K E A L L O W A N C E FOR CURRENT DECREMENT ON OPERATION FEEDER A N D GENERATOR OVERCURRENT RELAYS

OF

As indicated earlier in this chapter, it is usually permissible on indnstrial plant systems to ignore the slowing-down effect of the fault-current decrement when selecting time-hrrent settings for devices that are expected to operate in not over approvimately 0.6 sec after the short circuit occurs. I n such rases the settings are made on the basis that. the fault current is maintained at the initial X : or X y level. However, those devices having loager time settings, such as the generator-overcurrent and main-feeder relays, are affected sufficiently by the reduction in operating current t o warrant making allowanre for it as evplained below. The problem in setting a generator external-fault backup relay is to select time-current settings that are selective with the feeder relays and will also enable the relay to distinguish between fault currents and legitimate high-overload currents. The recommended current setting is between 200 and 250 per cent of generator full-load rating if there is a generator voltage regulator and 150 to 200 per cent if there is no regulator. The reason for the difference in settings is that, as shown in Fig. 9.19, the short-circuit-current output of a generator with a voltage regulator decays t o a steady-state value of 2 to 3.5 times normal, depending on

562

SYSTEM OVERCURRWT PROTECTION

generator characteristics and exciter ceiling voltage, but without a regulator the steady-state fault current would he less than twice normal. The shape of the generator-relay time-current characteristic curves changes radically with variations in the amount of voltage restraint, as shown in Fig. 9.5. During normal operation with 115 volts restraint, the relay curves are even more inverse than those of an extremely inversetime induction-overcurreut relay. With zero-volts restraint their shape approximates that of a very inverse-time relay, and the relay picks up at about 25 per cent of the 115-volt-restraint setting. This change in the shape of the operating curves makes it possible to set the relay so that i t will permit the generator t o carry any permissible overload likely to he encountered and yet be capable of operating quickly on the same, or lower, current levels when there is a fault. In the following discussion i t is assumed that the generator and feeders are connected directly to the hiis, as is the usual practice in industrial plants. Consequently, the worst conditions for coordination with the feeder relays is for a threephase fault close to the bus under which condition there would be zerovolts restraint on the relay so that it operates a t maximum speed for a given current. If the system connections were such that a three-phase fault on or close to the bus did not produce zero-volts restraint on the relay, it would be necessary t o use the appropriate time-current curves for voltage restraints between zero and 115 volts. To illustrate the procedure involved in selecting the settings for a

30

60

90

120 CYCLES

Short-circuit decrement wrvei of 9375-kw 13,800-volt 3600-rpm generator for faults on o m bur and on 4160-roll bur fed through 2Mx) h a 0.055 per-unit reactance transformer.

FIG. 9.19

SYSTEM OVERCURRENT PROTECllON

563

generator external-fault backup relay, consider a portion of a power system supplied by a 9375-kva 13,800-volt 3600-rpm turbine-generator with direct-connccted exciter controlled by a voltage regulator (see Fig. 9.20). Full-load rurrent of the generator is 390 amp, and 200 per cent of that would be 780 amp, which is greater than the 720-amp current-transformer primary rurrent corresponding t o the nearest current tap on the relay (6 amp). Consequently, it mill be neressary to use the 8-amp tap, which is the equivalent of 960 amp primary current, or 246 per cent of generator rating. The next step is to select a time-dial setting, which will make the generator relay A in Fig. 9.20 selective with relay B, which in turn must be selective with relay C farthest from the power source. I t was assumed that relay C had to be set on the 6-amp tap and 2 time dial (TD) to give 0.6 sec on the maximum symmetrical X i fault current of 3320 amp for a fault a t I on the 4160-volt circuit, in order to be selective with other devices on its load side. I n making this time-dial selection the effect of generator fault-current decay was neglected, because for practical purposes the 4160-volt-system decrement curve in Fig. 9.19 is so flattened out because of the transformer impedance that the effect of variations on relay performance is unimportant. If extreme accuracy is desired, allowance could be made for it as described later. Since relay R has to be selective with C, it must be 0.4 see slower when the two are operating on the 3320-amp fault current on the 4160-volt side of the transformer (see points D and D’ on curves C and B of Fig. 9.20). I t was necessary to set B on its 6-amp tap in order to ride through the starting current of a large 4160-volt motor on another feeder. With this tap setting, relay B would see 3320 200 (CT Pri)

4160 13,800

x

(CT 6 (tap)

- 4.17 multiples of

its tap setting

for a 3320-amp fault on the 4160-volt circuit. Since B is an inverse-time relay, it will require a 2.5 time-dial setting to give the required 1.0 sec with this much current, as determined from the family of time-current curves for the relay (Fig. 9.2). Generator relay A must be made selective with B for faults a t H on the 13.8-kv side of the transformer. Since the short-circuit-current magnitude there is much bigher than a t I, B’s time-current curve should be extended to the right of point D’where it coordinates with C. Then if the user was willing to sacrifice some of the benefit from the generator relay by acrepting slow operation, it could be assumed that the fault current on the 13.8-kv system would also be sustained a t the transient reactance X i level. If this were done, generator relay A could easily be

564


coordinated with feeder relay B by selecting a zero-restraint time-dial setting from Fig. 9.5 that mould make il 0.4 sec slower than B at this maximum value of current, as shown by points F and F’ on curves B and A’ of Fig. 9.20.

N

FIG. 9.20

CURRENT IN AMPERES ON 4160 VOLT BASE

Coordinated time-current C U W ~ Sof feeder and generotor overcurrent relays.


565

If greater accuracy is desired i n order t o obtain maximum value from the money invested in the relays, it will be necessary to secure a generator decrement curve (Fig. 9.19) and calculat,e the performance of both feeder relay I?arid generat,or relay 4 on the decaying current. The first, step would be to determine how long it would take for relay B t o close its contacts on the decaying current for a three-phase fault at N when set a t 6 amp and 2.5 time dial as required t o coordinate xit,h C. Then select a zero restraint time-dial setting for A , whirh is set on its 8 amp tap, t o make it 0.4 see slower than B , as shown by points E arid 12’ on curves R and 4 of Fig. 9.20. The difference in time between curves .4 and .4‘ represents the worthwhile improvement in performance of the relays made possible hy simply taking the trouble to determine the effect of the generator-current, decrement,. The behavior of relays A and B on the decaying generator current is a cut-and-try operation. First determine the arithmetical average of the symmetrical fault current represented by the 13.8-kv-circuit decrement curve in Fig. 9.19 for the estimated period of time required by the relay t o dose its coritacts when operating on the decaying rurreut. I n determining the average current. the decrement-curve ordinates that are averaged should be uniformly spaced and close enough together that the segment of the curve between adjacent ordinates is essentially a straight line. Incidentally, it should be noted that the subtransient reactance ( X y ) value of fault current is included when using decrement curves, whereas it, is neglected when determiuiiig the performance of relays assumed t o he operating on fault current sustained at the initial value. Then convert this average value of short-circuit current t o multiples-ofthe-tap setting of relay R , and find the operating-time rate of the relay on that much current from its time-current curves (Fig. 9.2). If the rate of speed thus determined is insufficient t o complete the closing operation in the estimated time, continue the same procedure until the contacts will close in the total time allowed. For this illustration, it was first .assumed that 36 cycles mould be long enough for B t o close its contacts for a fault at H . The average current under the decrement curve for 36 cycles is 5.47 times generator rating, or 8.85 times B’s & pickup setting. The relay curves in Fig. 9.2 show that an inverse-time induction overcurrent relay would travel at a rate that would take it 0.75 see t o close its contacts with this much current when-set on 2.5 time dial. Thus, for the first 0.6 see, relay B would have traveled 0.6/0.75 = 0.80 of the required distance t o close its contacts. I t appears, therefore, that the first estimate was too short as the relay still has t o go 0.20 of its t.otal distance; so it is necessary

566


to estimate how many more cycles are needed for it to close, when operating a t the slower rate of speed which the lower magnitude of fault current available during an additional period will produce. Since the decrement curve has flattened out considerably by this time, it is relatively easy to pick an average, convert it to multiple@-of-relaypickup, and find the rate of travel from the relay curves. I n t,his case, 11 more cycles with an average current of 3.51 times generator rating (5.70 times the relay pickup setting) gives a rate of travel that would close the contacts in 0.90 see. The disk will, therefore, travel 11/(60 X 0.90) = 0.204 of its total distance in t,he estimated 11 cycles. Adding this to the 0.8 of the distance already traveled makes a total of 1.004, which is close enough. The total time therefore is 36 11 cycles, or 0.783 see. This procedure gives the same results as the so-called step-by-step method. The latter involves determination of the percentage of its total travel that the disk would move when subjected to the arithmetical averege for each portion (increment) of the decrement curve that follows a straight line, then summing up the total distance traveled for the successive increments. Either method applies only when a relay is working on the saturated portion of its curve, which condition prevails in less than twice pickup current in the case of inverse-time relays and only a little later in the case of the generator and very inverse-time relays. The next step is to select a time-dial setting for generator relay A that will make it selective (0.4 see slower than relay B ) a t point E . This setting should be selected on the basis of zero-volts restraint corresponding to a solid three-phase fault close t o the bus, which would be the worst condition in the average industrial plant. The generator-relay curves (Fig. 9.5) show that a 4.5 time-dial setting would meet the requirement of 1.2 sec. If A is selective with B for a maximum fault with zero-volts restraint (curves A and B ) , it will be selective a t any lesser fault condition, when the system voltage is higher, because any voltage restraint slows down the generator relay. Curve A" shows how the 115-volt restraint imposed under normal load conditions pulls relay A ' s operating curve out of the way of any permissible short-time overloads. It appears from an inspection of the curves of Fig. 9.20 that neglecting the effect of generator fault-current decay would slow down the generator relay about 0.5 sec. While this may seem too short a space of time to be conrerned about, it is a relatively long time electrically, and could be the major factor in the amount of damage done by the short circuit. On low-voltage systems (600 volts and below) the circuit and fault impedance have a very pronounced effect on both the magnitude of fault current and the voltage a t the generator. These factors should be taken into account when selecting coordination settings for either the generator

+

SYSTEM OVERCURRB(1 PROTECTION

567

overcurrent relay with voltage restraint or the voltagecontrolled overcurrent relay discussed in the following section.

SELECTING SETTING FOR VOLTAGE-CONTROLLED OVERCURRENT RELAY FOR GENERATOR EXTERNAL-FAULT PROTECTION

The voltage-controlled overcurrent relay, when used for generator external-fault backup protection should he set so that it will not operate on any maximum permissible overload but will operate on short-circuit current selectively with feeder relays. The instantaneous voltage element controls operation of the induction overcurrent element and should be set to operate a t a voltage below the minimum value expected under the worst conditions likely to be encountered in normal plant operation. The current element has essentially the same time-current characteristics as a conventional induction overcurrent relay and is provided with current-tap and time-dial adjustments. The desired time-current setting is obtained from a family of curves similar t o those in Fig. 9.2. The shape of the time-current curves depends on which of three possible time-current-characteristic elements was supplied by the manufacturer for that particular relay. In order t o obtain maximum benefit from them, the effect of fault-current decay due t o generator-current decrement should be taken into account, as described in the preceding paragraphs, when selecting the time-current settings for coordinating these relays.

SETTING DIFFERENTIAL, CURRENT-BALANCE, AND WIRE-PILOT RELAYS

Generator and transformer-differential, parallel-line current-balance, and wire-pilot relays operate only on faults within the zones they protect and, therefore, do not require time-current-characteristic coordination with other devices. Generator-differential relays require no adjustments of any kind. Transformer-differential relays have current taps to compensate for differences between current-transformer ratios and the currents on opposite sides of a transformer, as well as the effect of Y-delta phase shift. Detailed instructions for selecting the proper taps are supplied by the manufacturer. Wire-pilot relays have provision for adjustment to (1) compensate for variations in length-(resistance) of the pilot wire, (2) vary the phase and ground-fault sensitivity, (3) permit the use of current transformers of different ratios. The manufacturer’s instruction book contains the

560

SYSTW OVERCURRENT PROTECTION

necessary information for making or changing any of these adjustments in the field. EFFECT OF Y-DELTA AND DELTA-Y CONNECTED TRANSFORMERS ON OVERCURRENT-RELAY COORDINATION

Y-delta and delta-Y connected transformers introduce a 30” shift in the phase relationship of fault currents ou opposite sides of them. On threephase fanlts this does not matter as Ear as ordinary overcurrent relays are coucerned, because the currents in the three phases are still halanccd. The phase shift is import.ant, however, when coordinating overcurrent deviccs on opposit,e sides of a transformer for phase-phase faults. In this case, t,he currcrits are no longer balanced, and the phase shift causes current to appear in the three line leads on the unfaulted side in the ratio of 2: 1:1. The magnitude of this “ 2 ” current is the same as that which would flow during a three-phase short circuit on the other side of the transformer, whereas the phase-to-phase short-circuit current on the faulted side of the transformer is only 86.6 per cent of that which would flow in a three-phase fault. This factor should be taken into account when selecting relay settings if complete selectivity is desired. EXAMPLES O F OVERCURRENT-PROTECTIVE-DEVICE COORDINATION

The following two examples of artual overcurrent-protect,ive-device roordination graphically illustrate the application of some of the principles discussed in this chapter. Case 1. The diagram of a portion of the distribution system in a plant and the coordinated time-rurrent curves of the various overcurrent protective devices installed on it are shown in Fig. 9.21. This plant had some local generators, and it was also connected t o a large utility system, so for all practical purposes it was safe to assume that the fault current could be maintained a t the 150,000-kva level in case of a severe short circuit on the 2400-volt system. The entire load on this port,ion of t,he system consisted of electrical heating equipment; so there was no pump-back fault oirrent,. Also, continuity of power supply mas of sufficient importance to require selective operation of all short-circuit protective devices. The calculated mavimum symmetrical fault current was approvimately 6500 amp on the 480-volt buses and 36,100 amp on the 2400-volt bus, as indicated by the heavy “stub” ordinates at the bottom of the graph in Fig. 9.21. Curve 1 shows t,he long and short-time tripping mechanism time-current characteristics of the 225-amp frame size 15,000-amp interrupting-


5bQ

FIG. 9.21 Coordination of time-current curves of relays, fuses, and oir-circuit-breaker hips (Case 1).

570


capacity feeder breakers. It was necessary to use the 2C short-timedelay trip instead of an instantaneous trip for high-current faults in order t o make the feeder breakers selective with the devices in the branch circuits supplied hy them. The short-time-delay 2C trip, being a solenoid-type device, is affected by the d-c component of the fault current, as well as the symmetrical a-c component. Therefore, the right-hand edge of the operating-time zone of the 2C trip is tapered to allow for the effect of the decay of the d-c component on its operating time. Actually, the d-c component follows a curve in its decay, and also it may on occasion disappear completely somewhat sooner than indicated, but for practical purposes the straight line shown is sufficiently accurate. I t was determined by drawing a line from the point on the base line representing the maximum asymmetrical short-circuit current t o the point on the upper boundary of the operating-time zone of the 2C trip where the latter crosses the 6500-amp ordinate (maximum symmetrical current). Curve 2 shows the combination of long- and short-time-delay trips required on the transformer secondary breaker t o make it selective with the feeder breakers. In plotting the high-current end of the 2B shore time trip i t was assumed that the d-c component of the fault current had disappeared and that the current was maintained a t the maximum symmetrical level. The operating-time bands of curves 1 and 2 include the reset-time allowance. Curve 3 represents the expected performance of the 200E (200 amp continuous) fuses protecting the primary of the transformer. B and C are its minimum- and maximum-melting-time curves, and the area between them is the accepted manufacturing tolerance. D represents the time required for the arc t o clear after the fuse melts, and A is the minus 10 per cent current tolerance, which must be allowed to avoid the possibility of the fuse being damaged, if some device on its load side fails to clear as soon as it should. Total area between curves A and D is the time-current operating zone which must be allocated to the fuse when plotting coordination curves. As long as there is clear space on the graph between the areas allocated t o the various circuit-breaker tripping mechanisms and fuses, they will coordinate satisfactorily with each other. Relay curves do not have tolerance hands similar to those used in coordinating fuses and low-voltage circuit-breaker trips, because the necessary allowance for relay overtravel, inaccuracy in adjusting the time dial and safety factor are taken into account when selecting the desired minimum time margin between a relay and the adjacent protective devices operating on the same maximum short-circuit current. The fact that one corner of the short-time trip 2 B overlaps slightly on


571

the damage-tolerance zone of the fuse is theoretically wrong. Practically, however, the risk of having exactly the right combination of faultcurrent magnitude and maximum-time operation of the 2B trip to cause damage to the fuse appears to be so small that it is better to accept it than to (I) use a larger fuse or (2) lower the circuit-breaker trip settings. The latter would probably cause complications by crowding the timecurrent characteristics of the various devices on their load side, and changing the fuse would mean poorer protection of the transformer. In this particular case, the original plan was t o use an extremely inverse-time induction overcurrent relay on the main-line circuit breaker. Subsequent consideration, however, revealed that an inverse-time relay should be used instead, because the time-current characteristics of the extremely inverse-time relay do not coordinate satisfactorily with the fuse. As shown by curve 5 of Fig. 9.21, the extremely inverse-time relay could he set t o give satisfactory performance on moderate values of fault current, but when so set, it would be much too fast on high-current faults to permit satisfactory selectivity with the fuse. Conversely, if the minimum-time (maximum current) end of the curve was raised t o 0.2 see, which is about as close as it is safe to try to coordinate relays with fuses unless extreme care is taken to time them, the relay became too slow on moderate values of fault current as shown by curve 7. Curve 4 shows how an inverse-time relay would be coordinated with the fuse to provide good protection in both the high- and low-current zones of fault current. The performance of a very inverse-time relay, as represented hy curve 6, is slightly better than the inverse-time relay in the high-current zone, hut not as good in the low-current zone, and, therefore, the latter was substituted for the extremely inverse-time relay originally considered. I t will he noted that curves 3 and 5 appear to show ample time margin between the fuse and extremely inverse-time relay curves a t high magnitudes of fault current, although the margin is actually too short for dependable coordination. This deceptive chararteristic of log-log graph paper should be constantly kept in mind when plotting and reading curves, in order to avoid false conclusions. Case 2. The diagram of a small portion of a mill power system and the coordination curves of the protective devices used on it are shown in Fig. 9.22. In this case the calculated maximum symmetrical current was 21,600 amp and was assumed t o he sustained a t that level, since the power came from a large system through the impedance of the step-down transformers. The maximum asymmetrical current was 60,000 amp, including the pump-back current from the large number of induction motots on the system and using a 1.6 d-c component factor.

572


Curve 3 shows the time-current characteristics of the temperature overload relay and current-limiting fuses protecting the 600-hp motor. Since there were no other protective devices on the load side of the fuse, the damage-tolerance curve was omitted. Curve 1 represents the minimum-melting-time curve of the fuses for

CURRENT IN AMPERES

FIG. 9.22

I 0 I

Coordination of time-current curves of relays and fuses (Core 2).


573

the 300-hp motor. The rest of the operating-time band of this fuse and the associated Ghermal relay characteristics were not shown, because the behavior of the fuses and t,hermal relays of the 600-hp motor was the controlling factor in securing coordination with the incoming-line relays. The original intention was to use 300/5-amp current transformers on the incoming feeder in accordance with the usual practice of having the full-load current of a circuit half t o two-thirds of full-scale reading on the ammeter. However, the subsequent coordination study of the circuit revealed that it would be better t o use 600/5-amp current transformers. With the original 300/5-amp current transformers the inverse-time relays, when set on their highest tap (I 6 amp), would have provided good short-circuit protection on high fault-current levels, as shown by curve 2. When so set, however, the time-current curve of the relay cuts across both the upper portion of the fuse curve and the lower portion of the thermal-relay curve. Theoretically, that is not good practice because the relay would not be selective with the other devices, but practically such a setting might be acceptable when viewed from the standpoint of the probabilities that a fault in the motor would persist a t that relatively low value of rurrent for just the right time to blow the fuse or operate the thermal relay, as well as the line relay. The operating time of the inverse-time relay could have been increased sufficiently to make it clear the fuse and thermal relay by setting i t on the 10 time dial (curve 5), but such a setting would have made the relay too slow, even on high fault currents, with consequent poor protection t o the system. Substitution of 600/5-amp current transformers proved a satisfactory solution to the problem. They permitted setting the relay as shown by curve 4 t o afford fast short-circuit protection and still be selective with the fuses and thermal relays of the 600-hp motor. To be sure, the change in current-transformer ratio resulted in a full-load-current reading rather low down on the ammeter scale, but that was of minor importance rompared with the improvement in fault protection. I n general, if one or the other has t o be sacrificed, i t is much better to accept a poor ammeterscale reading than poor short-circuit protection. As indicated by the vertical portions of curves 2 and 4, the instantaneous elements on the line relays could not he used, even when set at their maximum current of 80 amp, because they would not be selective with the fuses, which are essentially instantaneous devices a t high levels of fault current. This is in accordance with the rule that instantaneous devices in series cabnot be coordinated when both are subjected to the same fault current. Curve 6 shows the safe heating limit for the incoming-circuit cable, i.e., how long fault current of a given magnitude can be left on without damaging the cable insulation.

chapter 10

by Donald Beeman

Selection of System Voltages The selection of voltage is one of the most significant factors in the design of a power system for an industrial plant, a commercial building, a hospital, or any other type of building or plant. I t is a major factor in determining over-all system cost, flexibility, and ease of future expansion. This chapter will deal with the selection of voltage only for industrialtype plants. The selection of voltage for commercial-type buildings is included with the over-all discussion of power systems for commercial and office buildings in Chap. 14. FACTORS AFFECTING VOLTAGE SELECTION

There are several factors which affect system voltage selection: 1. Load magnitude 2. Distance power is to be carried 3. Utilization-device availability as a function of voltage ratings and limitations 4. Safety 5. Codes and standards The effect of these factors vanes widely in an industrial plant. Total load magnitude and plant area have little if any hearing on the selection of voltage in the 600-volt class. Here the utilization-equipment voltage ratings and voltage limitations are almost the entire governing factors. However, when selecting the primary voltage, load magnitude and distance are very important. If the plant is compact, 8s most of them are, then magnitude is the major determining factor. However, should the buildings he widely scattered as in some plants, both magnitude and distance are important factors. In some plants there are sizable portions of load, such as a pump house, quite a distance from the main plant. Here the problem is one of transmission of a block of power from one point to another. Magnitude and distance and type of line construction are important factors. 574

SELECTION OF SYSTEM VOLTAGES

575

EFFECT OF UTILIZATION DEVICES

Utilization devices have a profound effect upon voltage selection in industrial plants. Incandescent lamps, small fractional-horsepower motors, hand tools, business machines, appliances, etc., are generally available with 115- or 120-volt single-phase ratings. Where this class of equipment is used, 120 volts single-phase must be available. Polyphase motors are available for voltages from 110 to 13,200 volts. However, the desirable voltage from a motor-design standpoint and from a system-cost standpoint is a function of horsepower. Motors up to 200 hp are generally lower in price when rated 550 volts or less and because of their small size lend themselves better to those lower voltages from a design standpoint. Motors below 40 hp are not listed above 550 volts. From these same standpoints there are lower desirable and economic horsepower limits for other voltage ratings of motors. Thus, horsepower ratings of motors definitely affect system voltage selection in the secondary voltage class of 600 volts and below and in the plant primary voltage class of 2400 t o 13,800 volts where larger motors are used. Lighting-fixture voltages have, in the past, been generally limited t o 120 volts. With the advent of gaseous-discharge lamps the 120-volt limitation no longer exists. In many cases the gaseous-discharge lamp fixtures can he operated directly off the 480-volt power system that supplies motors and other power loads. Many furnaces cannot be designed and built most economically for voltages above 240 volts. Where these are used eutensively, they may have a very decided bearing on the selection of system voltage in the class 600 volts and below. Large arc-furnace transformers are generally limited to mavimum 23or 34.5-kv primary voltage because of the arc-furnace-transformer switching means. Thus, these can affect primary voltage selection for very large plants. The foregoing illustrates how the utilization equipment may affect system voltage selection. These specific effects are treated in the following text as a function of voltage class. I n many cases two or more secondary voltages in the 600-volt-and-below class are required in a given plant. How to obtain the combination of voltages most economically is discussed later. Similarly, more than one voltage may be required in the voltage range 2400 to 13,800 volts. SAFETY

Safety is generally a factor in selecting system voltage only in the area where appliances or portable tools are used. The National Electrical

576

SELECTION

OF

SYSTEM VOLTAGES

Code for safety reasons requires that voltages on some lighting circuits and portahle-appliance units he limited to 150 volts to ground. Where there is great, possibility of contact with energized parts such as in ungrounded frame portable tools, some portable lamps, etc., voltages of the order of 32 volts are seleckd because it has been shown that voltages above 50 volts t o ground can he lethal. On circuits above 120 volts there seems to be little concrete evidence that voltage selection from a safety standpoint is a major consideration. While the chances of electrocution do, to a certain extent, increase with an increase in voltage, any higher voltage can electrocute a person if h e contacts a live part under proper conditions. For a highest order of safety on circuits 50 volts and higher, work on current-carrying part,s should be done only with the circuit deenergizcd. CODES AND STANDARDS

The National Electrical Code places definite limitations on the voltages that can be applied to equipment. As previously mentioned, some lighting circuits arc limited to 150 volts to ground. Primary circuits in buildings are limited to 15,000 volts line-to-line. There are other limitations, hut these are not particularly effective in the area under discussion here. TABLE 10.1 Advantages of 480-volt Systems O v e r 240-volt Systems*

1

........ ........................ .......................... ................. Safely.. ......................

240volb

Avmilobilily of equipment. cost.. Lolei Voltage drop..

1

:

480volh

Same

135% High High

100%

Lox Low No proved difference. Bolh should b e worked on only when deenergized

* While 600-volt systems are less expensive, 550-volt motors, control, &c., 80

m e not

readily available as 440-volt equipmcnt.

SELECTION OF VOLTAGES 600 VOLTS OR LESS In most industrial plants the majority of the loads are integral horsepower polyphase motors, furnaces, and welders which are most suitable for operation on systems 600 volts or less. The choice of system voltage in this class for serving these loads is 208Y/120 or 240 or 480 or 600 volts. The use of a 480-volt system is generally desirahle for the average manufacturing plant because it costs less than 240-volt systems, and 410-volt


577

equipment for use on 480-volt systems is much more readily availahle than 550-volt equipment for use on 600-volt systems. These reasons are discussed in the following text. 480- VS. MXI-VOLT SYSTEMS

While 600-volt load-center systems cost ahout 2 to 7 per cent less than 480-volt load-center systems (Fig. l O . l ) , they have derreased in popularity primarily because of the lack of availahility of standard 550- and 575-volt utilization equipment from manufacturers' and distributors'

STANDARD KVA RATINGS OF LOAD CENTER UNIT SUBSTATIONS

FIG. 10.1

Chort showing the approximate comparative costs of 208Y/120-,

and 600-volt redid lood-center-systems.

240.. 480..

9 s


stocks. This was brought out forcibly during World War 11, during which time the percentage of standard 550-volt induction motors decreased from 11 to 4 per cent of the total standard induction motors made. When ordering machine tools or other utilization equipment with a considerable number of electric control circuits, it is often difficult t o obtain 550-volt equipment, particularly on short shipment. Pumps and other equipment which are stocked by the manufacturers with motors already mounted are generally stocked with 220- or 440-volt ratings and not with 550-volt ratings. Availability of utilization equipment is the major problem when choosing between 600-volt and 480-volt systems. Today, 600-volt systems are limited pimarily to expansion of those plants which already operate a t 600 volts or t o some textile plants where most motors are of special design, thus are not widely available from manufacturer’s stocks. 480- VS. 240-VOLT SYSTEMS

Economically speaking, there is seldom any reason for selecting 240 volts instead of 480 volts. The 240-volt load-center systems cost about 35 per ccnt more than 480-volt load-center systems (Fig. 10.1). As outlined in Table 10.1, 240-volt systems generally have higher losses and higher percentage voltage drop than 480-volt systems. If enough copper is used in t h e 240-volt feeders, the losses and percentage voltage drop can be made comparable to those in 480-volt systems, but in practical systems this is seldom done. The lower voltage systems cost more because there is more current per kva to he carried, thus larger circuit breakers and feeder conductors are required. For example, a 500-kva unit substation costs about $25 per kva a t 240 volts secondary compared with $20 per kva a t 480 volts secoudary. A single cable feeder to carry 200 kva costs about $9 per ft a t 240 volts compared with $3.75 per ft a t 480 volts. Motor control on an average will cost more a t the lower voltage. For example, a combination magnetic motor starter for a 745-hp motor would be size 2 a t 220 volts and size 1 a t 440 volts with comparative approximate costs to the user of about $135 a t 220 volts and about 595 at 4iO volts. Combining all these factors in their proper proportion results in the cost difference shown in Fig. 10.1, Some industries, where there is considerable dampness such as in dairies and slaughterhouses, have often selected 240 volts because i t is felt t o be safer than 480 volts. Operating records show that-the biggest factor in safety in this respect is to ground all non-current-carrying parts properly and securely, as outlined in Chap. 7, so that insulation hreak-


579

downs cannot place dangerous potentials on the uon-current-carrying parts. When working on circuit conductors while energized, there is a greater chance for injury from electric shock with higher potentials to ground or phase-twphase. However, as pointed out previously, any voltage above 50 volts can be lethal; therefore, the only safe way to handle these circuits in damp or other locations is to enclose the current-carrying conductors in securely and properly grounded metal enclosures and to work on current-carrying parts only when deenergized and if necessary with a temporary ground established on the current-carrying parts. In areas having mostly electric furnaces, 240 volts may be most advantageous because of the limitation of voltage which may be applied to these furnaces. In general, however, furnaces are large spot loads and cover only a small portion of the area of the plant. Therefore, these furnaces may be supplied by a separate load-center substation stepping down to 240 volts for the furnace only, the rest of the load being supplied a t 480 volts. For smaller furnaces the power may be stepped down from 480 to 240 volts. Seldom is it economical to use 240 volts for general distrihution in a plant even though a sizable portion of the total kw load may he 230-volt furnaces. WHERE 208Y/lZO-VOLT SYSTEMS ARE APPLICABLE

There are certain areas where 208Y/120-volt systems are more economical than 480-volt systems because the type of utilization equipment involved should or must be operated a t 120 volts. When such utilization equipment and incandescent lamps constitute a major portion of the load (more than about 50 t o 65 per cent of the total load), 208Y/120-volt systems may he more economical than 480-volt systems. Typical of such a load is that of a clothing-manufacturing establishment where practically a11 the power is utilized by motor-operated hand shears. Other areas in which 208Y/120 volts may be desirable are on assembly benches where small components are assembled and where small portable tools such as soldering irons, electric hand drills, electric nut tighteners, etc., are used. A typical case would he a small electronic-equipment assembly line. Again the choice of the lower voltages is primarily based on the desire to limit the voltage on the hand tools to 120 volts. In these assembly areas, most of the power is utilized at this low voltage. THREE-PHASE VS. ONE-PHASE CIRCUITS

I n general, threephase voltages are used exclusively for general distribution of power except for 120-volt loads which are fed from very small

580


transformers. When fed from small transformers ahout 50 kva or less, 120/240-volt single-phase three-wire systems may be more economical than 208Y/120-volt three-phase four-wire systems if there are no polyphase motors t o he fed a t this particular voltage.

LOW-VOLTAGE POWER SYSTEMS SERVING DIFFERENT VOLTAGES

The following discussion shows, for lights and power in most plants where integral horsepower motors are the major load, that combined light and power substation systems are more economical than separate light and power substation systems. Later there is a discussion that shows how power loads, such as electric furnaces, electric kitchens in cafeterias, and other electric heating equipment requiring 240 volts or 208 volts, can he served from a480-volt system.

FIG. 10.2 Two combinations of substation voltage ratings and Irva ratings to supply loodl which must operate at different voltages in the d m s 6CQ volts and less.

SELECTION OF SYSlEh4 VOLTAGES

58 t

Besides the economic consideration given later, there are other important factors of the combined power system with all substations stepping down t o one voltage such as 480 volts. The single-voltage systems are simpler in their concept and often enable a substantially lower cost primary system. For example, as shown in A , Fig. 10.2, a single primary feeder will care for four 1000-kva load-center substations requiring only one primary-feeder breaker and no protection a t each individual load-center unit substation. If, instead of stepping all power down to 480 volts, half were stepped down to 480 volts and half t o some lower voltage such as 210 volts, then two primary feeders may be required as shown in B , Fig. 10.2, if no individual primary protertion is used a t each load center. This method doubles the primary feeder system cost. An alternative method would be t o use fuses ahead of each individual load-center unit substation; this also increases the cost of the primary system compared with the arrangement shown in A , Fig. 10.2. In other cases where there may be large power units and smaller lighting units, fusing of the smaller units may be required, which is still an additional cost to the primary system compared with that shown in A , Fig. 10.2. Arrangement B , Fig. 10.2, requires more installation expense because there are more substations t o put in, it requires more space for the substations, and there is extra maintenance because of the greater number of units. It costs very little more t o maintain a 1000-kva transformer than it does t o maintain a 300-kva transformer. The elimination of fuses is also an advantage from the standpoint of selectivity of overcurrent devices. Greater service continuity can often be obtained by stepping all power down t o one voltage instead of two different voltages. For euample, a plant office building may require 400 kva of power a t 480 volts and 200 kva of lights, etc., a t 120 volts. A nominal arrangement might be a 500kva power substation and a 200-kva lighting substation, Fig. 10.3A. Should either substation be out the building may be useless. However, if two 300-kva 480-volt substations were put in and the 120-volt load served off the 480-volt system, as shown in Fig. 10.3R, then the loss of one substation or primary feeder would leave some lights and some power in the building to permit a t least essential operations t o be carried on. Even a t greater over-all system cost i t may be desirable t o make all substations 480 volts, for example, to get increased flexibility and to save money when future changes are made. By having all substations with the same secondary voltage, then the same voltage is available everywhere in the plant. Therefore, the availability of proper voltage will not be an expensive bottleneck for flexibility in shifting loads.

582


rvyn 500 K V A

480 V

(nl

"7n 200

KVA

208 Y / I20 v

FIG. 10.3 One-line diagram of ( A ) two substations operating a t different recondory voltages and IN two substations in iecondory selective orrongement operating at the some voltage with step-down trclniforrneri to supply 120 or 240 volts.

SELECTION OF LIGHTING VOLTAGE

The most general plant load other than that for motors is generally the lighting load. At the present time, illumination levels of 50 foot-candles are considered good practice, and levels as high as 100 foob-candles are


583

not out of the ordinary. Load densities of 2 t o 4 va per sq ft for lighting and 8 t o 10 va per sq f t for power are about normal. Until the advent of the fluorescent lamp, most distribution of power for lighting loads was a t 120 volts for the most efficient operation of incandescent lamps. Higher voltage incandescent lamps are fragile and generally not satisfactory. Modern lighting practices, however, include more and more use of fluorescent lamps, either alone or in combination with other types. 3'ince the fluorescent lamp has a ballast in series with it, the 120-volt limitation no longer applies. The required voltage a t the instant of starting is always above 120 volts and is supplied by a transformer built into the ballast. Thus the use of fluorescent lamps has opened new possibilities for the use of higher voltage circuits t o reduce the cost of the power system. Three Systems for Obtaining Proper Voltage for Lighting Loads. Because of the 120-volt limitation imposed by incandescent lamps, the 208Y/120-volt system has been commonly used for lighting circuits. However, most industrial plants have' long recognized the benefits of higher voltages for motors, and 480 volts is a common power-load voltage. Thus system A (Fig. 10.4) includes a 208Y/120-volt three-phase fourwire system for lighting and a 480-volt system for the power load. Such a system requires two separate substations. A second alternative is a combined light and power system using the same substation for both power and lights. System B (Fig. 10.5) includes a 480-volt substation with small step-down transformers located throughout the plant for supplying the lighting load a t 120 volts. This system has the advantage of using only one substation instead of two smaller ones, hut i t still uses 120 volts for lighting. The third system goes one step further and uses a higher voltage for the fluorescent lamp ballasts. System C (Fig. 10.6) uses the same substation for supplying power t o the plant using a 480Y/277-volt three-phase system. The power load is supplied a t the line-to-line voltage of 480 volts. The lighting load is distributed on the three phases but connected from one line t o neutral a t 277 volts. These three systems are considered being used in a typical industrial plant of a given size and shape to form the basis of comparison. A lighting load of 3.1 va per sq f t is assumed. In addition, it is assumed that % va per sq f t would be required for 120-volt convenience outlets. Summary of Economic Comparison. Representative first costs of systems A , B , and C are shown in Table 10.2. System A has a total installed cost of about $82 per kva. System B shows some savings relative to system A in that the total installed cost is about $59.50 per kva. System C is the least expensive, costing only $46.75 per kva total. On the basis of

FIG. 10.4 system A.

T 7'

UNIT SUBSTATION FOR 4 8 0 V POWER

CGH,lNG

%'

.il

I'

FiXTmES

T

Y!

I'

.I

.,'

One-line diagram of system With reporate substations for power and light service-

&--7

SECONDARY FEEDER 100 KVA LOAD

I

i(:

1'

I,

I~

,t_

'"1.''

\A*&.

?,

7 r---

_ r _ -7 i- -

ll---------

300 MCM 3-CONDUCTOR

CL?+SIDERED IN COST COMPARISON NOT CONSIDERED IN COST COMPARISON

_---

UNIT SUESTATION FOR LIGHTING (AND 120 VOLT POWER)

---

2.4 TO 13.8 KV-PRIMARY FEEDER-

One-line diagram of system with one substation for power and light service and 4 8 s 120-volt step-down transformers to provide.120-volt power for lights-system B.

FIG. 10.5

!swlEn

(SIZE I COMBINATION MOTOR STARTER1

''''

BOAR0 MISCELLANEOUS WWER AT 120 VOLTS

+FORMER -PbNEL

~ L C I R U I T BREAKER

BASIC SU0STAlION COMMON TO LI6HTlNG AN) 480 V POWER LOAD

One-line diagram of system with one substation for power and light service with general. area fluorercenf lighting a t 277 volts-system C.

FIG. 10.6

1T !T !T 'i

I

>

Ad.., aT }

F

Y

<

v)


i

d

ti

588

SELECTION

OF

SYSTCM VOLTAGES

system A being 100 per cent, system R costs only 75 per cent as much and system C only 57 per cent as much. The savings of the combined light and power system will justify the selection of this system where the total 120-volt load is as high as 50 t o 65 per cent of the total load a t 480 or 6M) volts. Although cost comparisons of fluorescent,-lighting systems are beyond the scope of this book, it is well to point out that the aforementioned figures must be considered in relation to the specific fluorescent-lamp layout. The first cost of the distribution system is small compared with the first cost of the lighting fixtures. The carrying charges on capital investment of the power system are small compared with the operating cost of the lighting system. Even taking these factors into account, however, will still show that combined light and power systems are less expensive than the separate substations of system A. Selection of Voltage for Combinations of Different lamps. Many plants may have high bays and low bays. The low hays are generally lighted by fluorescent lamps, and the foregoing discussion applies specifically to such areas. I n high-hay areas, mercury and incandescent lamps are often used in combination. The mercury lamps have ballasts like the fluorescent lamps and are, therefore, suitable t o higher voltage. As shown in Fig. 10.7, the combined light and power system C can be used to supply these combinations. At A , Fig. 10.7, is a combination incandescent and small mercury unit. The small mercury unit can he operated off 277 volts and the incandescent lamps a t 120 volts through a small step-down transformer connected to the 480-volt system. The large 3-kw units may he operated a t 277 or 480 volts and the companion incandescent lamps a t 120 volts off the small transformer connected to the 480volt system, as shown in B , Fig. 10.7. The fluorescent lamps are connected line-to-neutral as shown in C, Fig. 10.7. SYSTEM CONSIDERATIONS WITH COMBINED LIGHT AND POWER SYSTEMS

While the following discussion does not hear upon the problem of selection of voltage, i t is, however, included a t this point to consolidate the various engineering items that must be considered in designing combined light and power systems. Short-circuit Currents. The short-circuit currents on the lighting system are actually quite low when the fused combination motor starter is used for controlling the lights. The starters are equipped with silversand current-limiting fuses which very definitely limit short-circuit currents to low values. Not more than about 5000 amp short-circuit current can flow to a fault in the lighting system even though the available short, circuit current on the 480-volt system may be as high as 50,000 amp.


589

Twenty-ampere fuses are used in the controllers to give proper protection to the fluorescent-lamp ballasts. In this regard, it should be noted that a 480Y/277-volt system, for example, requires better protection than a 120-volt system. With the lower voltage system, it is difficult, if not impossible, t o get a restrike when a short circuit occurs. The short may often burn itself free on a 120-volt system. However, with higher voltages it is entirely possible to get restriking, and therefore the best protection is indicated. The silver-sand current-limiting fuse protects the entire branch circuit with system C . Actually, the short-circuit protection generally furnished with system

1 POWER LOADS

(A)

(81

ICI

FIG. 10.7 One-line diagram showing how combined light and power rubrlcltions using system C can supply various types of lamps which moy operate at different voltages.


590

A is less adequat,e than t h a t descrihed ahove for system R or C . I t is recognized that the circuit breakers in the pane1 hoards do not have s u f f cient interruptiiig ratiiig (only 5000 amp) for this service. Hoivever, there are no available devicej which meet the space requirements for this service and which have adequate interrupting rating for the short-circuitcurrent duty imposed ou them. Voltage Flicker. Voltage flicker duc t o fluct,uating power loads will not canse ohjectionable light flicker iii the average factory. Large resistance welders or large motors mhich start frequeiitly should not he coiinected t o the 480-volt hus t h a t serves the plant lighting. Resistance melders are being used iii increasing quantities in manufacturing plants, particularly in those mhich fabricate metal products. I n many larger plants employing load-center pomer distribution systems, the welders are connected t o a separate transformer whirh supplies nothing but weldirig load. In these plants there is no prohlem of lamp flicker due t o resistaiice welders, unless t h e plant i s supplied by a very meak primary system, in mhich case the Ricker prohlem will bc esseiitially thc same whether or not cnmhined light and p o n w syst,erns are used.

v

Y

A

Ali

T' +-3&

2 2 Y Y

WELOERS ANO 'OWER

NO

x 2 J

42.1

POWER

LIGHTS

v

$

TO WELOING L O A O ONLY

-Y

Y

L .,

hL

T

AT 7 + 7

NORMALLY

2 2 d-0PEN-d Y Y Y WELOERS ANO POWER

T

A 1 f

7 '

Y4L

7'

d

V I POWER LIGHTS

FIG. 10.8 Secondary relective orrcingemenl of load-center power dirtribulion ryrtern wilh reiiitonçe welderr on m e 480-volt bur and lighti on the companion 480-volt bur. Motor load may be on both burer.

I n manufacturing plants where the total resistance-welder load is not sufficient t o justify separate ivelding transformers or where there are no euceptionally large welders, the secondary selcctive arraiiscment of the load-center distribution systrm may he used, with oiie of the tivo substat,ions (Fig. 10.8) for general power load aiid resistance w l d e r s and the othcr for comhiiied light and pomer. Diiring normal operation the tivo suhstation biises are iiot in parallel, and the flicker on the hus t o which the welders are connected will iiot affect the light,s coiinected t,o the other hus. When diiring an emergency one of the primary feeders or oiie of the transformers is out of servire, the buses may he paralleled, and during t h a t emergeiicy period the flicker may bc iiot,icerhle. I n general there has


591

been no difficulty with comhined light and power systems where the principles mentioned have been followed. Reliability. With either system B or C , separate feeders should be used for serving lighting load only. There are many more outages on power feeders than on lighting feeders because of the greater numher of changes made in power feeders. Furthermore, the power feeders may have more voltage drop because of fluctuating loads. Hence separate lighting feeders should he used for greater reliability and t o reduce voltage drop. With any of the systems discussed in t,his hook, consideration should be given t o overlapping the area fed by one substation, so that an outage 011 that substation mill not cause complete outage of lamps in any one area. One method of accomplishing increased reliability economically is wit,h the secondary selective load-center power svstem shown in Fig. 10.9. Small plants with demands of 500 t o 1500 kva usually require only oiie substation. It has been common practice t o install a power transformer hank and a lighting transformer bank. The welders and power load are connected t o the larger bank, and the lights t o the smaller one. By usiiig the arrangement shown in Fig. 10.9, with two transformers equivalent in total kva t o the lighting plus the power demand, considerable increase in service reliability can he ohtained without incurring objectionable flicker from the welders. With two transformers of equal size, both stepping down t o 480 volts, the plant may he kept in at least partial operation with lights and major power apparatus in service with one transformer

FIG. 10.9 Secondary selective arrangement of load-center power distribution system with overlapping oreor of lights supplied from differenl substation burer to prevent complete lois of light should one bur b e deenergized.

592


deenergized. Where separate lighting and power banks are used, this cannot he done. Increased service reliability for the same reason can he obtained in larger plants by using the secondary selective system with substations in pairs, both stepping down to 480 volts, instead of individual light and power substations scattered through the working area. Other Than General-are0 Lighting. In systems A (Fig. 10.4) and R (Fig. 10.5), 120 volts are available for general-area lighting, as well as miscellaneous 120-volt power and lighting. Therefore, all load other than 480-volt load can be taken off the same circuit in those two systems. With system C (Fig. 10.6), 265-volt ballasts should he used only in fixtures for general-area lighting. These fixtures must he more than 8 f t from the floor and must not have switches mounted in them. The National Electrical Code requires that all lighting for individual benches, machines, and so forth should he operated a t 120 volts. Since there is always some load of this nature around any plant, it is necessary to obtain the power for that load from the 480-volt feeders of system C . Factory-office Power Supply. Frequently, industrial plants are designed with a sizable office area a t one end of the factory area. Here again system C (Fig. 10.6) can he used for general-area fluorescent lighting within the office area. In such a building, system C as described above would he used for the factory area. One or more feeders would also be run into the office area for general lighting, Fig. 10.10, For further discussion of officepower and lighting systems, see Chap. 14.

OFFICE FOR FACTOR"

FIG. 10.10 One-line diagram showing how combined light and power iubrtotion using system C supplies some factory power, some factory lights, and factory-office generalarea fluorescent lamps.

SELECTION

OF SYSTEM VOLTAGES

593

Selection of Small Dry-type Tronsformers for Combined Light and Power Systems. Selection of the proper kva and voltage rating of the

small dry-type transformers for t,he combined light, and power system R (Fig. 10.5) is essential. These traiisformers are a11 si,ngle phase and may be used as separate single-phase units or iii ùanks t o supply three phase, 240 or 208Y/120 volts. In general, single-phrtse units are less expensive and more adaptable than three-phase hanks. Three-phase hanks are used when three-phase 208-\dt small power as well as single-phasc 120-volt power is needed. The single-phase units are connected line-to-line ori the 480-volt systém and arraiiged for 120/240, three wire, one phase, on the seaondary (Fig.

' 480 VOLTS 3-PHAÇE 3WIRE

n-

20ôVOLTS 3 PHASE

1201240 VOLTS I-PHASE 3 - W I R E SINGLE PHASE LOADS ONLY

ia) n-

120 VOLTS IPHASE

2 0 ô Y I 120 VOLTS 3-PHASE 4 - W I R E 3-PHASE ANO SINGLE PHASE LOAOS (bl

FIG. 10.11 Variour lighting tranrforrner connections for stepping power to 120 "Olt..

down from 480

SPA

SELECTION

OF

SYSTEM VOLTAGES

10.11a). Three-phase units are connected delta on the 480-volt side and 208Y/120volts, threephase, four wire, on the secondary side (Fig. 10.11b). Most economical sizes of step-down transformers depend upon the load t o he served. In general, single-phase units are limited to 25 kva maximum and three-phase units t o 50 kva maximum. These sizes limit the short-circuit current on the 120-volt system to about 5030 amp maximum, thus the short-circuit current available is within the rating of most of the better grade of lighting panel-hoard protective devices. Many dry-type transformers available for lighting service have even ratios, i.e., 480-120 (4:1), 240-120 (2:1), etc. Where no taps are provided in the primary windings, these units do not in most cases have the correct ratio to supply proper voltage for a lighting system from the usual 480-volt power system. The difficulty arises because the ratio between motor and lamp name-plate ratings is not even multiples. Motors are rated 440 volts, while lights are rated 120 volts. It is obvious that, if the voltage is correct for the motors it will he too low for the lamps using a standard 4: 1 ratio lighting transformer. With 440 volts applied t o such a transformer, there would he only 110 volts less voltage drop for the lights. T o remedy this condition, lighting transformers with two full-rated 5 per cent taps or four full-rated 255 per cent taps below normal in the primary winding should be used. This 480-120-volt normal rating will have a tap 5 per cent below (456 volts) and 10 per cent below (432 volts). Thus, i t really has three ratios: Normal Tap-5 per cent below Tap-10 per cent helow

480-120 456-120 432-120

4:l 3.8:l 3.6: 1

If, for example, the power system voltage is correct for the motors, i.e., 440 volts, theii by using the 10 per cent below tap on the lighting transformer the voltage will he more nearly correct for the lights. The lamp voltage will he (440 X 1)/3.6 = 122 volts. Therefore, by having proper taps in the lighting transformers, i t is feasible to have the correct voltage for 440-volt power equipment and for 120-volt lights all supplied from the same power system. I n actual practice, voltages are seldom exactly equal t o the name-plate rating of power-consuming apparatus because of voltage drop and primary circuit regulation. A report by the AIEE Industrial Power Systems Committee (see Chap. 4) established a maximum desirable range of 480 volts maximum t o 420 volts minimum for 440-volt systems and 125 volts maximum t o 114 volts minimum for lamps. I n t h e case where the voltage varies from 440 t o 475 on the 480-volt power system, the voltage would vary on a lighting system supplied from this power system


595

in this way. If the 5 per cent below normal tap is used in the lighting transformer, the maximum voltage on the lighting system would be 475/3.8 = 125 volts. The minimum voltage would be 440/3.8 = 116 volts less the voltage drop in the lighting wiring system. These are within the limits set above. With 4: 1 ratio lighting transformers without taps, the voltage on the lighting circuit would have been 475/4 = 118.5 volts maximum and 440/4 = 110 volts minimum. The minimum value is too low for best illumination level from the lamps. The taps below normal in the primary of the lighting transformers permit raising the lighting voltage t o more nearly the lamp voltage rating while maintaining proper voltage on the 480-volt system for power apparatus. Similar transformers are available for supplying lights from 600-volt power systems. SELECTING VOLTAGE FOR POWER SUPPLY TO ELECTRIC FURNACES

Wherever practical, electric furnaces should he designed for operation on 480-volt systems. This provides the lowest over-all cost of the powersupply system and, in general, does not increase the cost of the furnace. As a matter of fact, in some cases, the cost of the furnace control may actually be reduced. As an example, one manufacturer installed four 2000-kw enameling furnaces. A comparison of the design of the furnaces showed that they could be designed with equal operating characteristics and life when rated either 230 or 460 volts. Selecting the 460-volt rating for operation on a 480-volt system reduced the power supply system cost by over $50,000. Some furnaces, particularly the smaller ones, cannot be designed for operation on 480-volt systems and, hence, must be designed for operation on 240-volt systems. Larger furnaces, that is, ones of the order of 500 t o 1000 kw which must operate on 240-volt systems, may have a separate unit substation to supply the furnace. In such cases the unit substation is considered a part of the furnace and will generally be moved with it. A system with 480-volt substations for supplying general load and 240volt substations for supplying furnace loads is shown in Fig. 10.12. Smaller furnaces, less than about 500 kva, can, in general, be most economically served off 480-volt systems with transformers to step down from 480 to 240 volts for the furnace. The problem of supplying the smaller furnaces is much like the problem of supplying combination light and power loads, as discussed earlier in this chapter. Such a system would be like that shown in Fig. 10.13. Even though there are enough smaller 240-volt furnaces in one area to justify a 240-volt substation, i t is still hest nractice in many cases t o put

596


in all substations with 480-volt secondaries and then supply the 240-volt furnace load through 480-240-volt transformers. These smaller furnaces are more readily movable and may be used t o form a part of the production line. Hence, power supply, because of voltage, will be no problem when the smell furnaces plus their auxiliary 480-240-volt transformers are moved from one location to another. Likewise, 480 volts will be avai!able 0 2 4 0 VOLT SUBSTATIONS FOR LARGE ELECTRIC OVENS

0480 VOLT SUBSTATIONS FOR GENERAL POWER AND LIGHTS

-om

FIG. 10.12 Schematic one-line diagram rhowing 480-volt substations for general-purpose power and lights and 240-volt substations for large spot loads consisting of electric ovens and furnaces which must operate a t not higher than 240 volts.

SELECTION OF SYSTEM

_-__-,

\ - - I<, -7 ~Ti. ~\

~.-.

597

VOLTAGES r-------

-----------1

I

h

T,

240V

FIG. 10.13 One-line diagram showing one substation to serve 080-volts general power load ond 230-volt furnoce through 480-240-volt transformers.

whenmachine toolsaremoved into areas which may originally be occupied by electric furnaces. SELECTION OF VOLTAGES FOR SYSTEMS 2.4 TO 15 K V Voltages in this class are used mainly for primary power distribution in industrial plants. All plants using a primary voltage employ voltages of this class except some of the very large chemical plants, steel mills, etc. The latter may employ subtransmission voltages above 15,000 volts. When the utility voltage is below 15,000 volts, there is no problem of selecting the primary voltage. Since the National Electriral Code does allow 15,000 volts in the buildings, there is generally no reason for transforming voltages of the order of 13,800 volts to, say, 2400 or 4160 volts for transmission through the building. The higher voltage ran be carried in cable and conduit or in interlocked-armor cable t o the load-center substations and there transformed to utilization voltage. When the utility-voltage supply is above 15,000 volts, then in all evcept a few very Iarge steelmills, chemical plants, etc., transformation to some voltage less than 15,000 volts is desirable, both economically and technically, for transmitting the power through the plant to the load-center sub-


598

stations where it is stepped down t o utilization voltage. I n this case the plant engineer must select a plant primary voltage. The trend in the past several years has been t o select higher voltages to obtain greater economy and flexibility for expansion. Many studies have shown that either of two voltages will suffice in the majority of the cases, Fig. 10.14, i.e., 4160 volts for plants having a supply transformer and/or generating capacity of 10,000kva and less and 13,800 volts for plants having a supply transformer and/or generating capacity of 20,000 kva or greater. For the range in between 10,000 and 20,000 kva, either 4160 or 13,800 volts may prove t o he most economical. However, when selecting the voltage in this range, sight should not he lost of the fact that all plants grow in size and that while 4160 volts might, for example, be slightly less expensive for a 15,000-kva plant, if that plant should happen to grow to 25,000 or 30,000 kva, then 13,800 volts would be most economiral. I n that case the selection of the higher voltage would undoubtedly pay off in the end. The basic advantages of higher voltage are accounted for by the improved economy in primary distribution circuits, in lower primary

1

13800 V

13200

$1

6900

1 4160 V

4160p {

24001 1000

5000

15,000

10,000

20,000

25,000

KVA

SUPPLY TRANSFORMER

FIG. 10.14

GENERATING CAPACITY

Chart rhowing most economical primory system voltage for indurtriol plants where there is a choice of primory voltage.


599

switching equipment costs, and in improved economy and simplification of the main station bus arrangement. The effect of high-voltage motors will be discussed later. Because of the wide variation in magnitude of load and in the powersupply systems, i t is not feasible to draw a simple curve of primary-system cost as it was for the secondary systems. I n a small plant of 2000 kva, for example, the cost difference between a 2400- or a 4160-volt system may be very small, but in a system of 15,000 kva the cost differential may be several dollars per kva in favor of the 4160-volt system. As a further example, in some plants the power may be supplied a t 13,800 volts; thus no transformation is necessary before carrying the primary voltage to the load centers in the plant. I n other cases a master unit substation may be required. I n still others power may be generated or generated and purchased. 13,800 VS. 4160 VOLTS

The data in Fig. 10.14 show that 4160 volts is most economical for smaller plants and 13,800 volts most economical for larger plants. A few examples will explain this. First assume a 10,000-kva plant. The source switchgear costs about $1.10 per kva a t 4160 volts and about $0.75 per kva a t 13,800 volts. Primary-feeder switchgear costs about $1.90 per kva at 4160 volts and about $2.70 per kva a t 13,800 volts. Primaryfeeder cable costs about $2.00 per kva per 1000 ft a t 4160 volts and about $1.60 per kva per 1000 ft a t 13,800 volts. Adding these, the total cost of the primary system is $5.00 per kva a t 4160 volts and $5.05 per kva a t 13,800 volts. This spread would be considerably widened should there be many motors that could operate directly a t 4160 volts and could not operate directly a t 13,800 volts. The difference will be widened in any case by about $2.00 per kva because of the lower cost of load-center unit substations with 4160- vs. 13,800-volt primary windings. This makes the total in favor of the 4160-volt system about $2.05 per kva. Assume as a second example a 20,000-kva plant, the cost a t 4160 volts would be about $2.10 per kva for the source unit, $1.90 per kva for the primary switchgear, and $2.00 per kva per 1000 f t for the primary cable, making a total of $6.00 per kva. For the 13,800-volt system the approximate costs are as follows: source unit, $0.35 per kva; primary switchgear, $2.20 per kva; and primary cable, $1.25 per kva per 1000 ft, making a total of 93.80 per kya for the primary system. Adding $2.00 per kva extra charge for the load-center unit substation, the savings for the 13,800-volt system is $0.20 per kva. This load is about a break-even point. Should the plant have been about 30,000 kva, the saving for the 13,800-volt system would be about $6.10 per kva.

600

SELECTION O F SYSTEM VOLTAGES

2400 VS. 4160 VOLTS

One of the problems that arises quite often is, why select 4160 volts instead of 2400 volts? There are two major reasons for doing this. The first is that 4160-volt systems have lower cost, and second, they have greater allowance for expansion. I n so f a r as cost is concerned, 4160-volt switchgear for a given interrupting rating costs less than 2400-volt switchgear. Another advantage costmise in switchgear is that often a 1200-amp circuit breaker will oarry the full-load current a t 4180 volts whereas a 2000-amp circuit breaker would be required at 2400 volts. There is a marked difference in the selling price of 2000-amp and 1200-amp circuit breakers. Often more kva per circuit can be carried at 4160 volts than a t 2400 volts, resulting in fewer circuits, thereby further reducing the cost of the switchgear. Cable costs are generally less at 4160 volts than at 2100 volts because less copper is involved. The largest metal-clad circuit breakers made for 2400-volt service have 150-mva maximum interrupting rating. The maximum rating circuit breaker of similar construction for 4160 volts is 250 mva. This allows a much larger system t o be built a t 4160 volts than at 2400 volts without going t o complicated synchronizing buses, etc. Influence of Motors on Choice of 2400 vs. 4160 Volts. Tho use of motors at primary voltage does not, in general, alter the selection of 2400 vs. 4160 volts for generation and distribution of power in industrial plants. This condition arises in certain industries, typical of which are petroleum refineries and paper mills, where a sizable percentage of the total electric power is consumed in motors rated above 40 hp. Since these motors are readily available from most motor manufacturers with a voltage rating of 2300 volts, motors rated above 40 hp may be operated directly connected t o 2400-volt power distribution systems without intervening transformers. Hence, i t is often thought that the lowest over-all electric-system cost, including power generation, distribution, motors, and control, can be obtained by generating power or transforming i t from the electric utility system and distributing i t through the plant a t 2400 volts primary voltage. A review of the factors which influence system costs indicate that such is not the case. Studies of specific systems show that i t is nearly always less expensive, so far as over-all electric system costs are concerned, to use a primary or generation voltage higher than 2400 volts and t o operate motors, 200 hp and less, on 480- or 600-volt circuits. I n studying the effect of primary motors on the selection of 2400 vs. 4160 volts in more detail, certain basic equipment and system elements were considered.


601

These are discussed in the following. A paragraph is included to show the effect of other types of motors and motor starters. I n some areas, 4800 volts is the standard primary voltage. The basic factors presented are not substantially altered whether the voltage is 4160 or 4800 volts. Since 4160 volts is far more common than 4800 volts, the former will be used as a basis of comparison. Motors and M o t o r Starters. As standard squirrel-cage induction motors are by far the most commonly used motors, they have been considered to make this study as representative as possible. The curves in Fig. 10.15 are plotted for 1800-rpm motors. It is equally as important to select motor starters with adequate short-circuit interrupting rating as it is to select feeder circuit breakers with adequate interrupting rating, thus only modern combination motor starters are considered. System Factors. To use 440- or 550-volt motors in a plant in which the primary or generation voltage is higher than 600 volts requires a transformation to less than 600 volts for these motors. Therefore, when comparing 440-volt motors vs. motors operating directly a t primary voltage (i.e., 2400 or 4160 volts), the installed price of a step-down substation must be included with the 440-volt motors and control. This substation 70

60 a

$

50

w (0

LL

0 I 40

a

0 k 0

2" 30 ul

20 _I

A

x

4000 VOLTS - MOTORS AN0 CONTROL ONLY

10

0

50

100

150

200

250

300

350

400

450

500

INDUCTION MOTOR HP RATING

FIG. 10.15 Approximate cost of induction motors and control and approximate system cork as a function of motor horsepower.


602

is included a t $8 per hp. That figure is based on the installed price of a unit substation consisting of a transformer and one secondary breaker and 4 4 0 VOLT MOTORS assuming that, because of diversity, 1 kva of transformer capacity will care for 134 hp of connected motors. ~ ~ - 7 0 0 8 Since branch circuits to 440-volt motors require larger conductors than T T T STARTERS branch circuits to 2300- or 1000volt motors, $1 per hp was included in curve A , Fig. 10.15, to cover this. L, The one-line diagram of the circuit A ] U N l T SUBSTATION elements for 440- or 550-volt motors operating from 2400- or 4160-volt primary systems is shown in Fig. A “cy\ 2400 OR 10.16. The cost of the low-voltage 4160 VOLTS motors, control, and the step-down FIG. 10.16 Schematic one-line diasubstation is plotted as a funcgram of 440-volt motors and control kion of motor horsepower rating in and rtep-down substation. curve A , Fig. 10.15. Motors Operating Directly at Primary Voltage. The approximate prices of high-voltage motors and control only are plotted in Fig. 10.15 on curves B and C. When selecting the primary voltage of the power distribution system, it is necessary to consider more than just the prices of the mot,ors and control which operate a t primary voltage, i.e., a t 2400 or 4160 volts. The cost of the primary system must be considered as well as motors and control. In factoring the primary-system cost i t is assumed that if 2300-volt motors are used the power will he distributed a t 2400 volts and if 4000-volt motors are used the power will be distributed a t 4160 volts. In industrial plants with a total demand of a few thousand kva, the primary power system (i.e., the generating plant or main step-down substation from the utility system) and the primary switchgear and cable cost about $4 more per kva for 2400-volt primary voltage than for 4160volt primary voltage. This cost will increase to about $8 or $10 per kva for systems about 10,000kva and larger. Therefore, if a 2400-volt system is arbitrarily chosen in place of a 4160-volt system, the extra cost of the entire 2400-volt distribution system plus the cost of 2300-volt motors and control must be compared with the cost of 4000-volt motors and control for those motors which operate a t primary voltage. This comparison can he made by referring to curves C and D i n Fig. 10.15, which show that the 2300-volt motors and control plus the extra cost of the 2400-volt power system is more expensive than 4000-volt motors and control. For the general case then, curves C and D and not curves B and C should he used

0 00 $ 2 2


603

when selecting the primary-system voltage for plants where a sizable portion of the total connected load consists of motors larger than about 200 hp. The total extra cost of the entire 2400-volt power system compared with a 4160-volt power system has been included a t $8 per hp of highvoltage motor, i.e., $8 per hp was added to curve B to obtain curve D. This figure ($8) is arrived a t by assuming that complete 2400-volt power systems cost $4 more per kva than 4160-volt power systems and that an equal kva capacity of power system is required for serving motors operating a t primary voltage and to serve all other load which is operated a t 600 volts or less. If these motors require only one-half the system capacity for supplying them, that means that 2 kva of more costly 2400volt system capacity (at $4 per kva extra) is required per horsepower of high-voltage motor. That is the equivalent of $8 per hp of high-voltage motor. The actual range of extra cost of the 2400-volt system is from about $2.50 per kva a t 5000-kva plant load to about $7.25 per kva a t 15,000-kva plant load. The system costs are given in dollars per kva but are added in Fig. 10.15 as dollars per horsepower. This is done on the basis that the kva and horsepower ratings of high-voltage motors are considered equal. 480- or 600-volt Systems M o s t Economical for Motors Rated 200 Hp or Less. By referring t o the curves in Fig. 10.15, it can he seen that below about 175 hp the 440-volt motors enable the least investment in the electric equipment regardless of the primary-system voltage. Thus, generally speaking, all motors rated 200 hp or less should he operated on systems rated 600 volts or less, regardless of the primary voltage selected. Therefore, there is no economic justifi- 2300 VOLT MOTORS cation for ever arbitrarily selecting 2400 volts as the primary voltage so far as any considerations of motors 200 hp and below are concerned. A close examination of the detailed prices will show that a t 200 hp the cost of 0:s 440-volt motors and starters and step2400 VOLTS down substations, Fig. 10.16, is about FIG. 10.17 Schematic one-line equal t o the cost of 2300-volt motors and diagram of 2300-volt motors and starters, Fig. 10.17. This is the reason CO"tl0l. for placing the arbitrary division a t 200 hp rather than 175 hp as pointed out by the curve. These curves represent approximate prices and do not follow detailed variations.

U

4160-volt Primary Systems Usually G i v e Greatest Over-all Economy.

Assuming that there will he a transformation from either 2400 or 4160 volts t o supply motors rated 200 or 250 hp and less, respectively (see


604

Table 10.3), the choice between these two primary voltages is based on primary-system costs and the cost of 4000-volt motors and control for those motors rated above 250 hp, Fig. 10.18. As pointed out previously and as shown in curves D and C , Fig. 4000 VOLT M O T O R S 10.15, the selection of a 4160-volt system and 4000-volt motors and control (for those motors rated 250 hp and above) CR-7008 enables an over-all less expensive instalSTARTERS lation than the selection of a 2400-volt system and 2300-volt motors and control. BUS I n areas where the larger motors (i.e., 4160 V O L T S those above 200 to 250 hp) constitute FIG. 'O.'* One-line more than about 50 t o 60 per cent of the diagram of 4COQ-rolt motors and total load, 2400 volts may be less expenCO"tlOl. 4 v e provided that the total load of the plant is not more than a few thousand kva or distances are not too great. In general, an equally satisfactory yet a lower cost power system can be obtained by selecting 4160 volts a8 the primary voltage rather than 2400 volts.

9-Fa P-Fa Phj

TABLE 10.3 Selection of Motor Horsepower Ratings as Function of Voltage

480 2400 4160

I I 440 2300 4000

a

Smallest to 250 hp 200 hp up 250 hp up

Where 2400 Volts Is Particularly Applicable. I n plants which are served a t 2400 volts directly from the utility system, then, of course, i t would be more economical to use 2400 volts directly and t o place all motors rated 200 hp and above directly on the primary feeders. However, in these cases, almost all motors rated less than 200 hp should be operated on a 480- or 600-volt system, stepping down from the 2400-volt primary system. I n some cases there are existing 2400-volt systems in industrial plants, and i t is sometimes more economical t o extend these a t 2400 volts rather than changing to 4160 volts. However, i t may be more desirable to expand the system a t some higher voltage sncb as 13,800 volts rather than t o extend the system a t 2400 volts. Where the primary voltage is above 5000 volts, 2400 volts is applicable t o concentrated loads which supply motors rated 200 hp and above. I n


605

these cases it is preferable to step down from the higher voltage t o 2400 volts rather than to 4160 volts, if the capacity per 2400-volt bus is limited to less than about 7500 kva and nearly all this load is utilized a t this hus directly by motors rated above 200 hp. For single loads larger than 7500 kva per bus with motors above about 250 hp, it is often necessary to go t o 4160 volts to enable the switchgear to handle the short-circuit currents available from such large systems. A typical application where 2400 volts would be most economical is a pumping station, Fig. 10.19, in which the control may be of the fused combination type to give fast shortcircuit protection or of the vertical-lift power circuit-breaker type where H I G H VOLTAGE SYSTEM

1 I

2400VOLTS

I

&

7" :zz

:o'tTS FOR M l S C POWER A N 0 L I G H T S

B A T T E R Y O F S E V E R A L 2 0 0 H P OR L A R G E R M O T O R S FIG. 10.19 One-line diagram of 2&O-voh power-supply system for large spot load ar in (I pumping station.

606


the vertical-lift feature is desirable and where the frequency of starting is not too great. If this preference is followed, it will be necessary to make a transformation for these large motors. The cost of the transformation is substantially the same whether it is made to 2400 volts or t o 4160 volts. Thus the power-system cost is not a significant factor as all the load is concentrated in one place. The 2300-volt motors and their starters are less expensive than 4000-volt motors and their starters. This can be seen by comparing curves B , C , Fig. 10.15. Effect of Various Types of Motors and Control. By using 2300-volt motor starters rated 50,000 kva interrupting rating, some reduction can be obtained in the over-all cost of the 2400-volt installation compared with the lower voltage installation for motors below 200 hp. This merely reduces the crossing points of curves A and B , Fig. 10.15, t o about 150 hp. Even if plain motor starters (the use of which is not generally recommended) without built-in short-circuit protection are considered for both the 2300- and 440-volt motors, there will be little change in the relationship of curves A and B. These variables, together with the effect of explosion-proof requirements, have little bearing on the basic factors presented bere. As a matter of fact, most variables such as special induction motors or special starters will tend t o make curve B cross curve A at a higher horsepower rating than shown in Fig. 10.15. Although Fig. 10.15 is plotted using 1800-rpm motors as a base, other motor speeds over the range 600 to 3600 rpm have little effect on the general relationship of the curves. Synchronous Motors. There is less price differential between lowvoltage (600 volts or less) and 2300-volt synchronous motors and control than there is between induction motors and control for these voltages. This differential will tend t o make curve R cross curve A a t about 150 hp instead of 175 bp as shown for induction motors, Fig. 10.15. Motors Larger than 200 Hp Operating at 600 Volts and Below. I n many industrial plants t h e great majority of the motors are rated less than 200 hp and therefore are operated from 480- or 600-volt distribution systems. In these plants there may be one or two motors which are considerably larger than 200 hp. I n general, it is more economical t o operate the larger motors directly from the 480- or 600-volt system than i t is t o provide another transformer to step down to 2400 or 4160 volts where the primary voltage is above 5000 volts. Summary of Advantages of 4160- vs. 2400-volt Systems. I n general, it is most economical to operate all motors rated 200 to 250 hp and less on systems rated 600 volts or less, regardless of the primary voltage (see Table 10.3). Where a choice of primary voltage can be made, it is, in most cases,

SWCllON OF SVSlEM VOLTAGES

FIG. 10.20 Std-mill motor room *lowing large a-c-motor-driven d-c generators at the D-c motam are shown in the center.

left.

2. From an electrical standpoint more satisfactory designs of smaller motors (less than about 2000 hp) can he made a t 6.6 kv than a t 13.2 kv. The problem of satisfactory designs of smaller motors (under about 2000 hp) losea its significancewhen the over-all utilization-voltage picture is looked at critically. While it is beyond the scope of this hook t o analyze the utilization-voltage picture critically, other studies have ahown that an intermediate voltage in the 5-kv range is desirable for medium-size motors too large for the 600-volt class system and too small for the 13.8-kv class systems where 13.8 kv is the main distribution voltage. When this intermediate voltage is adopted, the line-up of motor horsepower vs. voltage is as shown in Table 10.4, and illustrated in oneline diagram form in Figs. 10.21 and lQ.22. Basidly, the intermediate voltage is essential to care for those motors economicslly and technically in the 200 to 2OOO-hp range. Even though this intermediate voltage has not been used in many 6.9-kv systems, it is eseential in the 13.8-kv system. This does not represent an economic penaIty on the 13.8-kv system vs. the 6.9-kv syetem. The intermediate voltage is desirable from an economic standpoint as well aa from a motordesign standpoint in the 6.9-kv system to cover the application of motors in the range of about ux) to 1000 or ZOO0 hp. In the smaller horsepower

609


ratings, under about 1000 hp, a 6.6-kv motor and starter costs as murh as or more than 6.9-2.4-kv transformer capacity arid 2.3-kvmotors and motor starters. I n either the 6.9-kv or 13.8-kv systems this intermediat,e volt,age provides grcater flexibility of system design, and motors more suitable

13.8 K V

LARGE MOTORS 2000 HP AN0 LARGER

5 0 0 0 K V A MAX.

MOTORS HP O R LESS

200

MOTORS

200 T O 2 0 0 0 H P FIG. 10.21 horsepower.

One-line diagram.rhowing selection of voltage for motors Primary distribution voltage 13.8 kv.

01

a function of

Generator transmission Medium-sire motor* Small-sire motor utilization voltmge and large-motor utilization voltage ".allage (200-2000 hpl up lo about 200 hp

13,800 volh

2400 volt.

I

480 volh

Where the capacity per bus exceeds about 6000 t o 7500 kva and motors are 250 hp or larger, it may be desirable to use 4160 volts for these intermediate-voltsge motors t o permit more economical use of standard interrupting-rating switch gear and motor control.

TABLE 10.5

Motor Horsepower Range and System Voltage as Functions of Motor Voltage Motor voltage

440 volt, 2,300 volh

6.600 v d h 13,200 V d h

Horsepower range

System voltage

Up to 200-500 hp 200-2000 hp Above 1000-2000hp Above 2000 hp

480 W1h 2,400 volh' 6,900 volt. 13.800 volh

* 2400-volt systems are referred to here rather than

4160-volt systems because this

is really a voltage for supplying only local motors and not for extensive distribution of power. Hence there are not generally sufficient distribution systems savings a t 4160 volts t o pay for the extra cost of 4000-volt motors. See preceding text for 4160- vs. 2400-volt system comparison, However, far very large installatione of intermediate voltages motors, 4.16 kv may he used instead of 2.4 kv. See footnote on Table 10.4.

With the intermediate voltage in the over-all motor-voltage picture, it works out that so far as utilization-equipment voltage rating is concerned the only significant economic penalty of the 13.8-kv system vs. the 6.9-kv system is the $1.50 to $2.00 per hp higher initial cost of the larger motors above about 2000 hp. 13.8-kv vs. 6.9-kv Power Distribution Systems. The next important point of consideration in 13.8-kv vs. 6.9-kv systems is to see if there is


61 1

enough saving i n the 13.8-kv system compared with the 6.9-kv system to pay for the extra cost of the larger, more expensive 13.2-kv motors. As a fundamental principlc of selecting power-system voltage, i t is gcnerally true that the higher the voltage i n a given class, such as 13.8 kv vs. 6.9 kv in the 15-kv class, the lower the system cost. This will be borne out in the following discussion. First, a general comparison of the basic system components of the power system, i.e., transformers, switchgear, unit substations, cable, and hus must be ronsidered. These comparative costs were outlined prcviously and are summarized in Fig. 10.23. An euamination of this figure indicates that one would not expect a wide difference

13.8 K V

MOTORS 2 5 0 H P AND L A R G E R

MOTOR 2 5 0 H P

OR LESS

FIG. 10.22 One-line diagram showing selection of voltage for motors CIS CI function of horsepower where large concentration of motors may require 4160 volts instead of 2400 volts to enable we of available metal-clad rwitchgear and motor starters.


612

in cost per kva of 13.8-kv and G.9-kv systems because there are maiiy items whirh are suhstantially thc same regardless of system voltage, and the items which are more expensive at one voltage than for t h e othcr voltage more or less balance earh other. There are, however, factors in addition t o those shown in Fig. 10.23 whirh make the larger 6.9-kv systems increasiiigly more expensive thaii 13.8-kv systems. These factors wili be studied hy first rniisideriiig small systems, ttieii medium-size systems, and then very large systems.

TRANSFORMERS ABOUT SAME PER KVA

T

-METALCLAO

v 8

SWITCHGEAR COSTS SLIGHTLY MORE AT 6 . 9 KV

$

I $ TO $ 2 PER H P MORE AT 13.8 KV

CABLE COSTS 25 TO 65 CENTÇ PER KVA PER M F T MORE AT 6.9 K V

J

-.*

UNIT SUES COST ABOUT 4 0 CENTS TO $ i PER KVA MORE AT KV

FlG. 10.23 Chort rhowing comparotive coiti of ryrtem eqvipment a t 6.9 m d 13.8 kv.

SELECTION O F SYSTEM VOLTAGES

613

Small Systems. A one-line diagram showing the typical elemelits of a small system, for example, of ahout 20,000 kva or less is outlined in Fig. 10.23. I n such systems there is little, if any, economic penalty of 6.9 kv vs. 13.8 kv. Insystems of this size, short-rircuit duties are gcucrallyless than 500 mva, which permits the application of modern metal-clad switchgear usith removable air-type circuit hreakcrs. I n t,his t,ype of metal-clad switchgear, interrupting ratings are the same for both 6.9 arid 13.8 kv, and there is only a very small extra cost for the 6.9-kv sivitrhgear comparrd with 13.8-kv smitrhgcar. I n general, i n these smaller systems 1200-amp continuous-rated circuit breakers are adcquat,e berause, as a rule, smaller feeder circuits go with smaller plants. In ?able cirixiits, for example, there is more copper but less insulation required for G.9-kv circuits than for 13.8-kv circuits. The balance hetween these two represents a small price differential in favor of the 13.8-kv rahle. I n the smaller plants t,he lower cost of 6900-480-volt substations plus the loiver cost of any 6.6-kv motors more often offsets t,he higher cost of 6.9-kv sn-itchgear and 6.9-kv cahle so that there would he little cost differential betwell the two voltages. If a n appreciable portion of the poiver in these smaller systems is used in large motors, then there may he a saving a t 6.9 kv, depending upon the feeder circuit lengths arid sizes. However, since all systems gron, it is better t o choose 13.8 kv instead of 6.9 kv for the smaller systems t o allow for easier, lower cost future expansion. Hence, the syst,em voltages for small-size mills would he as outlined in Table 10.5, using 13.8 k v as the plant, primary voltage. Choice of 13.8 kv vs. 6.9 kv in Larger Systems. The larger the system becomes, the more factors there are that increase the cost of a 6.9-kv system faster than for a 13.8-kv system. The first, factor is t,hat standard 1200-amp and 2000-amp power circuit breakers will carry twice as much kva at 13.8 kv as a t 6.9 kv. Circuits larger than about 14,000 kva at 6.9 kv require a 2000-amp circuit breaker whereas circuit,s as large as 28,000 kva can he carried on a 1200-amp circuit hreakcr at 13.8 kv. This factor alone does not add much t o the cost of the system, but it is nevcrtheless a step in the direction for inrreased cost of (i.9-kv systems. For example, if a source unit was rated 18,000 kva, it would require a 2000amp circuit breaker at, 6.9 kv and only a 1200-amp circuit breaker a t 13.8 kv. This would add ahout 20 cents per kvatotheG.S-kvsystemcost,. In smaller systems there would he perhaps only one or two 2000-amp circuits at 6.9 kv, hut in larger systems there may he an increasing number of 2000-amp circuit breakers required at 6.9 kv where 1200-amp circuit breakers would do at 13.8 kv. An illustration of one of these systems is shown in Fig. 10.2i. All circuit breakers marked A4would be rated 2000 amp at 6.9 kv and the others 1200 amp. A11 circuit hreakers would be rated 1200 amp at 13.8 kv.


614

The auxiliary equipment, such as synchronizing buses, reactors, etc., would of necessity he of higher ampere rating a t the lower voltage. On the basis of the circuits and ratings shown in Fig. 10.24, the 6.9-kv system costs about $1 per kva more than the 13.8-kv system for switchgear and synchronizing bus reactors. The extra cost of these items a t 6.9 kv nearly offsets the extra mot,or cost. arid extra step-down unit substation costs a t 13.8 kv; and as systems become larger, cable or circuit costs go up rapidly a t 6.9 kv. I n these larger circuits of several thousand kva each, more than one conductor per phase is usually required, and such circuits rost about two-thirds more per kva per 1000 ft a t 6.9 kv than a t 13.8 kv. For interlocked-armor-cahle circuits this differential runs up t o ahout 10000 KVA

20000 KVA

9

Y

1 &\

15000 KVA

&

15000 KVA

UL.2

$A

$A

Y

Y

A -&&'A M I L L BUS

FIG. 10.24

CANNOT PARALLEL FEEDERS

One-line diogrom of power system wring synchronizing-bur arrangement and 15 kv. Synchronizing-bur system reducer short-circuit duty to less tho" 500 mva. a distribution voltage leis than


615

65 cents per kva per 1000 ft. This item is another factor for making large G.9-kv systems disproportionally more expensive than large 13.8-kv systems. For an average length of cable circuit of 2500 ft in Fig. 10.24, the extra cable cost would add about another $1.25 per kva t o the system rost at 6.9 kv. The saving of $1.00 per kva in powerhouse switchgear and synchrouizing hus reactors plus the extra $1.25 per kva saved in cable circuits at 13.8 kv more than pay the extra cost of the 13.8-kv load-center unit substation arid 13.2-kv motors in medium-size plants up t o about 40,000-kva demand. There would he some additional saving in the 13.8kv snzitchgear and cable circuits in the mill area. Higher Momentary Duties Adversely Affect 6.9-kv Systems. Since the momentary currents are usually the most l i m i h g factor in applying power circuit hreakers in power systems, the use of the steel-mill-type 60,000-amp momentary 500-mva 13.8-kv metal-clad switchgear enables slightly larger systems t o be built without synchronizing buses than does the 70,000-amp 7.2-kv 500-mva breakers on 6.9-kv systems. The maximum ampere interrupting rating of the 500-mva 7.2-kv circuit breaker is 44,000 amp, arid the maximum momentary rating is 70,000 amp or about 1.6 times the maximum interrupting-rating amperes. This ratio is only euough t o care for the d-c component in 6.9-kv systems. The maximum ampere iriterruptiiig ratiug of the 13.8-kv 500-mva steel-mill magne-blast cirruit breaker is 25,000 amp, hut its momentary rating is 60,000 amp, or 2.4 times the maximum interrupting rating. I n this circuit breaker, as in the 7.2-kv circuit breaker, only 1.6 times the maximum interrupting rating is required t o care for d-c component. Therefore, the difference between 1.6 and 2.4 is available t o care for motor contribution. I n systems in which the short-circuit duties are helow the momentary ratings associated with 500-mva magne-blast circuit breakers, there is no significant difference in the circuit arrarigement or number of circuit components required t o care for short-circuit current whether the system voltage is 6.9 or 13.8 kv. There is a factor in favor of the 13.8-kv system because of the higher momentary rating of the 13.8-kv breakers. Where the mill systems become so large that the short-circuit duties increase t o above the momentary rating of the 500-mva metal-clad magne-blast rircuit breakers, synchronizing bus arrangements are required. While these are workahle and often used for keeping short circuits within hounds, synchronizing bus arrangements do have certain limitations. Among these :imitations are that any transfer of kilowatts through the reactors represents a small Z2Rloss, and it is difficult t o transfer large amounts of kvar between bus sect,ions without creating wide differences in voltages between the bus sections. When one source unit is out of service, all load from that bus section must he fed through a reactor with the attendant voltage drop. The synchronizing bus arrange-

616


meiit imposes rirruit desigii aiid operat,iiig restrictioiis too. For example, feeders from diffcrciit hus sectioiis caiiiiot he paralleled because this will short out thc syiir,hroiiiziiig bus reartors, Fig. 10.24, aiid iiicrcase the short-circuit duties of the system. Therc are aiso drtiiiite limits in the relatioiiship hetweeii the size of t h r soiirre uiiit aiid the iiiterriiptiiig rating of rirruit breakers whirh raii he iised i i i syiichroiiiziiig bus arraiigemeiits. For example, ivheii usiiig 500-mva switrhgear, the largest source uiiit per bus sertion is limited t o ahout 25,000 kva. ‘I’his is ofteii too severe a limit for larger systems. Whilc syiichroiiiziiig hiis arraiigemerits are expedieiits t h a t solve maiiy system short-rircuit prohlems, it is, however, desirahle iii larger mills to operatc with as many souice units as possihle oii one hus t o allom more freedom of system operatioii and desigii. When more source uiiits are operated oii one hus, it requires higher iiiterruptiiigraparity stat,ioii-type switchgcar. This statiori-type switchgear is availahle x i t h circuit breakers for iiiterruptiiig dut,ies as high as 2500 mva a t 13.8 kv. Thcse circuit breakers can he used u p t o nearly their full interrupting rating on 13.8-kv syst,ems. They i a n , hoivever, be used t o orily part of their mva int,errupt,ing ratirig on 6.9-kv systems because of inoment,ary and maximum interrupting-ratiiig limitations. This can he seen hy referring t o Tahle 10.6. TABLE 10.6 Station-type Circuit-breaker Interrupting Ratingr a t 13.8 and 6.9 kv Intorrvpting roting, mro ~

At 6.9 kv

At 15.8 kv

iO00 1500 2500

.

77,000

1 O00

l i 5.000 190,000

1500 2500

1

575 860 1440

Csing a 1000-mva station-type circuit breaker at 13.8 kv enables the capacity per hus t o he nearly doubled compared with using the 500-mva magne-hlast-type metal-clad switchgear, but a t 6.9 kv the use of the 1000-mva circuit breaker eriables ari iiirrease of only 75 mva or 15 per cent per bus compared with metal-clad switchgear. The serious limitations of 6.9 kv using other interrupting-rating circuit breakers can be seeri from Table 10.6. It will be noted that only by going to the 2500-mva circuit breaker can a real gain he made at 6.9 kv. Not only is there a limitation in the interrupting duty and therefore the size of a system t h a t can be put on one bus, there is also a very severe economic penalty on the use of


617

these larger circuit breakers at 6.9 kv. This can be noted from Table 10.7, which shows that this larger switchgear costs almost twice as much per mva of interrupting duty at 6.9 kv as it does at 13.8 kv. TABLE 10.7 Approximate Cost of Station-type Metal-enclosed Switchgear per M v a per Circuit-breaker Position Approximate price per my.*

cirruitbreaker rating, m*a

&t 13.8 kv

I

At 6 9 kv

$16 16 16 22

2500

$16

28 28 38

* Interrupting rating. Summary of System Cost Differences. T o combine the factors that have been discussed in the foregoing, Fig. 10.25 has been chosen to illus40 MVA

25MVA

25MVA

-$-

&

0 0 0 ~

~

~

7A

TRANSFORMERS 25 MVA EACH

I

Hd!H 1

I

IS00 FT EACH

$1

SHORT CIRCUIT DUTY 1250 MVA

'3

LOOFT';-.

2100 F T

2300 F T

3 $ 3 3 3 3 3 3 P P

'r

J~ 30 MVA MILL 'A"

F

18 MVA MILL"B"

12 MVA MILL"C"

A

C

T

O

22 MVA MI LL"D"

X 8 MVA MILL"E"

500 MVA METALCLAD SWITCHGEAR AT A L L M I L L S

FIG. 10.25 Large power system with distribution v o l t a g e less t h a n 15 kv using rtotiontype switchgear.

61 8


TABLE 10.8 Ratings of Equipment in Fig. 10.25 Item

.............................. ........

Switchgem. slotion lype.. Cable cirwih, eoch cable three-conductor, 500 M C M .

1

13.8 kv

1500 mvm 4 per circuit

................................... ........................... T o l d difference infavor of 13.8 k v . . ................................

Cost difference in favor of 13.8 kv.. Less cost diAerence of substotions and motors.

6.9 kv

~

2500 mvo 8 per ~ircuit

+$12.00

- $ I.SO

per kvo per kva

110.50 per kvo

trate the system components used to determine the difference in system cost. While this may not represent any particular plant, it could be typical of a steel-mill system of the order of 100,000-kva capacity. Table 10.8 gives the detailed ratings of the switchgear and circuits. These have the same source and same load, the difference being in voltage only. For the entire power system down to but not including the step-down substations and utilization equipment, the difference in favor of the 13.8-kv system is $12 per kva. From this must he subtracted the extra cost of the 13.8-kv step-down substations and large 13.2-kv motors. This will be about $1.50 per system kva, leaving $10.50 per kva of total system rapacity as a saving of the 13.8-kv system. Different system designs could vary the relative cost of these systems$1.00 to $2.50 per kva. For example, the cost of the 6.9-kv system may be reduced somewhat by the use of a 1590-mva circuit breaker in all but the 50-mva generator position. This, however, would require the use of synchronizing buses and the added complication in system design and operating problems associated with the synchronizing- bus arrangement. This typical esample will shov that for mills with loads above 30,000 to 40,000 kva the 6.9-kv s3stems are considerably more expensive and more complicated than 13.8-kv systems. Where 6.9 K v Is Particularly Applicable. The 6.9-kv systems are particularly applicable in certain large systems supplied with a primary voltage above 15 kv and in some testing stations. I n large mills, such as steel mills, where the basic transmission voltage is ahove 15 kv, 6.9 kv may be used as the utilization voltage at eachload center, such as at a hot-st,rip mill or a cold-strip mill, etc. In this case the advant,ages of 13.8 kv may not he so great economically because there is not the length of transmission involved to make the saving in the distribution system to offset the increased cost of motors rated 13.2 kv


619

compared with those rated 6.6 kv. The suggested voltages for mills of this type are outlined in Table 10.9. TABLE 10.9

Large Mills-Transmission

G e n ~ r o t o r *.nd trO".miSSiO" "OltDge

33 kv or 69 kv

Voltage above 15 K v

Lo,ge-,ize motor utili.otion Medium-size motor Smoll-size motor YtiliZOtion "tiliZ.ti0" voltage "Oltage "oltage "p to 2000 hp ond (200-2000 hp) about 200 hp 1.rger

6900 volts

2400 volts

480 mlh

USE OF SYSTEM VOLTAGES ABOVE 15 KV I n many large plants and mills, transmission voltages above 15 kv are used. Such voltages cannot be generated directly or used directly on motors; hence a transformation up from generator voltage and down t o suitable voltage for the motors is required. This will add one or two extra steps of power transformation not required with the 13.8-kv system shown in Fig. 10.25 a t an extra cost of $3 t o $6 per system kva for the additional transformer capacity required because of the higher voltage. This extra transformer cost must be offset by savings in switchgear and circuits to justify the use of transmission voltages'above 13.8 kv. This, however, can often be justified where long distances or very large loads are involved or where part of the power is purchased from the utility a t some voltage above 13.8 kv.

620


SELECTION OF TRANSMISSION VOLTAGE ABOVE 15 KV

There are, in general, four classes of voltages which have been used for this service. These are 23 kv, 34.5 kv, 46 kv, and 69 kv. These voltages can be subdivided into two classes, i.e., 34.5 and 69 kv, for the sake of discussion. The 34.5-kv distribution for very large systems enables the use of compact indoor metal-enclosed 34.5-kv station-type cubicle switchgear similar to that shown in Chap. 13. This occupies less space than opentype outdoor switching structures and, since it is all metal enclosed, the problems associated with cleaning the many exposed insulators that are found in outdoor stations normally employed at this voltage are eliminated. Where space for locating substations is no serious limitation and the insulator cleaning problem is not considered a major obstacle, larger, simpler systems can he built for lower cost a t 69 kv than a t 34.5 kv. The higher voltage systems require fewer circuits where there is bulk transmission of large amounts of power from one point to another. The general criterion for the selection between 34.5 kv and 69 kv would hinge on whether or not metal-enclosed indoor station-type cubicle switchgear for this higher voltage is to be used. If it is, then 34.5 kv solidly grounded Y is the upper limit because that is the maximum voltage for which such switchgear is now available. There are, of course, other factors which will enter into the selection of voltage in this class, such as utility supply voltages, voltage already available in the mill, kva demand, etc.

FIG. 10.26 One-line diagram of (I large system using high-voltoge transmission (above 15 kv) and a high-voltage load center arrangement to step down to voltages below 15 kv.


621

USE OF THE LOAD-CENTER PRINCIPLE

When higher primary voltages such as 34.5 or 69 kv are used, the loadcenter principle may be employed t o reduce the cost of these systems. This principle involves primarily t,aking power at the higher voltage t o the load areas and there stepping i t down t o utilization voltage. I n this case the step-down substations supplying iudividual load centers for mill areas should have a maximum rating of about 30,000 t o 40,000 kva for the 6.9- or 13.8-kv utilization bus arid a maximum kva rating per single threephase transformer of ahout 15,000 t o 20,000 kva each t o limit the shortcircuit duty t o within the ability of 500-mva metal-clad switchgear. These substations will be located at t.he various mill areas. For loads larger thau 30,000 or 40,000 kva conrent,rated in one area, more than our suhstation may he required at that load area. Figure 10.26 is a one-line schematic diagram of such a load-center system. I t is well t o point out that when the higher primary voltages are used the system voltages can he either 13.8 or 6.9 kv for distribution in mill areas. See Chap. 13 f o r more details on primary distributiou systems.

Chapter 11

by Donald Beeman

L oad-center Power Systems and Circuit Arrangements c The technical system data and economic information on selection of voltage discussed in Chaps. 1 t o 10 form the foundation for the design of the industrial power distribution system. That information, together with the necessary load dat.a and information oil power sources, enables the po\~er-systemdesignert o lay out the plant power syst,cm. The fundamental principles upon which modern power distribution systcms rest are based on the “load-center power distribution” principles. The load-center power system is only a part of the over-all plant distribution system, as shown in Fig. 11.1. For discussion of the primary distribution system, see Chap. 13, and for discussion of the secondary distribution system, sce Chap. 12.

LOAD-CENTER POWER DISTRIBUTION SYSTEMS 600 VOLTS AND BELOW

As will be seen later, the& principles apply t o higher voltages too, but for the sake of detailed explanat,ion, the load-center power distribution system for serving loads 600 volts and below n.ill be used here. The loadcenter power distribiition system has two distinctive features which set it apart from other methods of delivering electric energy t o points of use within a building or plant. They are as follows: 1. Power is distributed a t primary voltages of 2.4 t o 13.8 kv t o unit substations located close to the centers of electrical load. Here it is st,epped down t o utiliaat,iori voltage (600 volts or less) arid delivered by short secondary feeders t o points of ose. 2. Where t,he total plant load cannot be met by a single small substation of 750 t,o 1500 kva or less, two or more small load-center unit substations 622

LOAD-CENTER POWER SYSTEMS A N D CIRCUIT ARRANGEMENTS

623

are used rather than a single large suhstation a t the point of incoming SUPPIY. The advantages of the load-center power distribution system are manifold. First, it costs less. For example, in a 250,000-sq f t single-

PRIMARY SYSTEM

I LOAD CENTER POWER DlST SYSTEM

BUS TIE

SECONDARY DISTRIBUTION SYSTEM E-----

BUSWAY

LOAD

LIGHTS MOTOR

FIG. 11.1

120 VOLT PANEL BOARD

One-lme diagram showing primary system, load-center power dirtributiun system, and recondory distribution system in on industrial plant.

614

LOAD.CENTER POWER SYSTEMS AND CIRCUIT ARRANGEMENTS

floor factory the old method of power supply would have a large piecemeal substation located outside the building and long lo\\~-voltagefeeders to serve the load. This power system from the 13.2-kv incoming line to the plug-in husway would cost installed about $143,000. A modern loadcenter system for the same plant would cost about $121,300, a saving of $21,700, or 15 per cent. The breakdowri of the costs is shown in Fig. 11.2. The load-center system has 3000 kva of substations to allow for diversity; yet these substations cost about the same as the single 2500kva piecemeal substation. The significant factor in Fig. 11.2 is that the major saving is made iu the cable. This stresses the importance of

150

FIG. 11.2 Comparative cost of complete old-style system and load-center power dirtribution system ond major components.

LOAD-CENTER POWER SYSTEMS AND CIRCUIT ARRANGEMENTS

625

always looking a t the over-all system, as a change in one part may have a significant bearing on some other part. While transformer costs are higher in the load-center system, the secondary switchgear cost is materially reduced because only 15,000-amp interrupting-rating feeder circuit breakers (in cascade) are needed for the small 750-kva load-center substations whereas 50,000-amp interruptingrating feeder circuit breakers (in cascade) are needed for the large 2500kva substation. Figure 11.3shows how the short-circuit current increases with transformer-bank kva rating a t 480 volts. The cable costs are reduced because the power is carried most of the

I-

z

w LL

K

a

0

6

I0

t

FIG. 11.3

Chart showing how short-circuit current increaser

01

a function of transformer

size.

way in a small high-voltage cable instead of a multiplicity of heavy lowvoltage cables. This is shown graphically in Fig. 11.4. There are other savings too. For example, losses are usually much lower in the average l o a h e n t e r system than in the old large low-voltage substation systems.


626

As the plant grows, extra savings are made because the load-center power distribution system offers great flexibility in providing for new loads in existing buildings or for evpansiou into new buildings. Load-center power systems are less costly t o evpand than old-style systems where large substations are put in in anticipation of load growt,h. I n the load-center systems the system capacity is built up in small units as

VOLTAGE

COST

4160

2400

480

13%

23 %

100 %

7

I

!

I

I I

I 80"

I I

I I

I

I I

L.a*D.RC&

LFIG. 11.4 voltoger.

~

- --I

I

1

UUOLIIT.

L-----J

I

'

LOAOARL. L-----J

1

Chart showing the cornparotive cost for circuiti to transmit power a t different


627

the load grows. Growth of the power system is closely coordinated with actual plant growth. As shown by the upper part of Fig. 11.5, when the old-style system is used a large initial investment is made iri a large substation. Secondary feeders are added as required until the capacity of the substation is reached. Then another substation is built a t great expense. If the load

ADDITIONAL INSTALLED

LARGE SUBSTATION

YLE POWER

SYSTEM

MORE FEEDERS

>

k

SYSTEM COSTS

0

2a

0

2

w + m %

\

I

DITIONAL UNIT SEVERAL SMALL UNIT SUBSTATIONS INSTlfLLED I

I

SUBSTATIONS

INSTAL LED

YEARS*

LH-kk LOAD-CENTER UNIT SUBSTATION

FIG. 11.5 The shaded part shows the accumulated rovingi for load-center power dirhibution as the plant expands, using a load-center system instead of Iorge substation systems.

628


for the new substation does not materialize, a large investment is wasted. As shown by the lower part of Fig. 11.5, the initial investment in the load-center system is lower because onlv sufficientcapacity for the present load plus a small reserve is installed. If load growth stops, there is no large idle investment to carry during the years ahead. The shaded area between the curves shows the over-all saving. Lower voltage drop means less voltage spread. Proper voltage a t all parts of the plant means better performance of the production equipment, hence more production and less spoilage. This improves production efficiency. Because secondary feeders are short in a modern load-center system, normal load voltage drop from each load-center substation to the end of the secondary feeder is held to around 2 per cent maximum, Fig. 11.6. In contrast in old-style systems with long low-voltage feeders, a voltage drop along the length of the feeder of 10 to 20 per cent is not uncommon. Indeed, it may be necessary to raise the voltage cousiderably above normal a t the substation bus to maintain enough voltage a t the end of the long feeders to operate the utilization devices. See Chap. 4 for more data on the effect of voltage spread and how t o reduce it. The load-center system provides greater flexibility. With small units they may be added when and where needed to care for increased load as it develops. Then, too, there are many types of systems or circuit arrangements with varying degrees of service continuity and flexibility, as outlined later, which permit adoption of the type of system best suited to various production processes. Continuity of service can be greater because of a small number of scattered substations instead of one concentrated large substation. A substation failure affects a much smaller part of the total production area. Duplicate circuits down to the load area can he employed with a minimum A

T

LOAD CENTER SYSTEM

tL

T

FEEDER

DROP

FEEDER

FIG. 11.6 Load-center pow;; distribution systems with short secondary feeders have far less secondary-feeder voltage drop than old large substation systems with long secondary feeders.


629

of extra cost as, for example, in the secondary selective system descrihed later. The factory-assemhled and coordinated load-center unit substation is the heart of the load-center power system. These substations are available in standard arrangements with catalogue numbers assigned. Hence, it is easy to purchase them as only a catalogue number or, a t the very most, a simple functional specification is necessary. Engineering and planning are reduced as the substations are factory designed; so the plant planner has only to provide a flat surface to mount these on and t o bring the leads to and from the substation after installing the two or three sections of the suhstation. Since the substations can be located indoors almost anywhere or outdoors also, planning of the power system can be simplified. This flexibility permits ordering Substations before final production layouts are crystallized. The substation components are coordinated by the manufacturer; hence that burden is removed from the system planner. Since the equipment is factory assembled and coordinated, it is only necessary to bolt together the two or three sections of the substat,ion in the field. It is shipped in sections for convenience in handling and shipping. Should processes change and radically different power-system layouts be required, the snhstations have high salvage value in that they may he readily moved in a few sections from one location to another. Factory-assembled equipment has established selling prices and also requires a minimum of labor for installation in the field. These two factors permit far greater accuracy in estimating power-system costs than with piecemeal assembled substations with many separate parts to order, coordinate, and assemble in the field. The large design engineering and purchasing work content and the large amount of field assemhly make estimating of the cost of field-assembled piecemeal Substations far less accurate.

LOAD-CENTER-SYSTEM CIRCUIT ARRANGEMENTS Because there are many small substations in load-center power distribution systems, there is the possibility of many circuit arrangements. While a great number are possible, those commonly used can he resolved into four basic arrangements or combinations of them. These basic arrangements are: 1. Radial 2. Secondary selective 3. Primary selective 4. Secondary network

630


These may be compared on many bases, but some of the more important characteristics are: 1. cost 2. Safety 3. Flexibility 4. Service reliability 5. Ease of expansion 6. Simplicity 7. Diversity 8. Voltage regulatioii 9. Short-circuit-current level Comparisons of these arrangements do not require detailed discussion of the primary-feeder system. In most cases the load-center unit substat,ions are supplied by radial primary feeders. Since looped primary feeders are sometimes considered for serving load-center systems, a discussion of the looped primary feeder as it, specifically applies to load-center systems will he given in this chapter. This discussion is concerned mainly with the part of the power system from the primary-feeder bus to and including the substation secondary bus. Although radial secondary feeders are used in the majority of the load-center power systems to feed the load supplied from the secondary hus of the unit substation, certain variations are sometimes employed, and these variations have a bearing on the over-all characteristics of the particular circuit arrangement that may be employed. Transformer and Primary-cable Failure Rates. As a background to the circuit-arrangement, discussion, the operating record of askarel transformers is about one failure per 2500 transformer years and primary-cable failures average about one failure per mile per 100 years. The trend toward the use of rubberlike insulation on primary cables, which eliminates the necessity of potheads, mill improve the service reliability of primary-cable circuits even more. Service reliability depends as much on using adequate modern equipment and itist,alling it properly and carefully as it does depend upon circuit arrangement. I t is misleading to try to represent the variations in t,he characteristics of these circuit arrangements by any numerical or alphabetical index because the difference between these systems on many points is so very small. Effect of Interrupter Switches on Safety and Overcurrent Protection.

While the application of interrupter switches on the primaries of loadeent,er unit suhst,ations does not affect the hasic one-line diagram of the circuit arrangement, it does have a serious bearing on the safety of the over-all system and on the overcurrent protection requirements. Many system engineers consider the use of interrupter switches in unit substations as a partial compromise hetmeen an ideal solution from the


631

safety standpoint and a practical solution from the economic point of view. The ideal solution would he t o have a power circuit breaker as the incoming-line interrupting device in each unit substation, as illustrated in Fig. 11.7. This is, of course, relatively evpensive and has resulted in the use of incoming-line interrupter switches, as shown in Fig. 11.8, as the practical compromise system in thousands of industrial plants t o provide a disconnecting means without primary overcurrent protection at each xubstation transformer. I

1

8 T

I

P

I

T y &&&A T 9

T

J J J 2

J J J J

1 3 2 2

*)J 1 J

Y Y Y Y

Y Y Y Y

Y Y Y Y

Y Y Y Y

FIG. 11.7 Ideal load-center power distribution system wilh individual protection for each bod-center substation provided by an adequate power circuit breaker.

Function of Interrupter Switch on Unit Substation. By far the most common function of the interrupter switch is t o provide simple isolation of the unit substation from its incoming feeder, as indicated in Fig. 11.8. Here t,he maximum intended duty of the switch is t o disconnect the transformer from the energized incoming line, breaking only the magrletizillg current of the transformer. This is a reasonably safe operation, provided that the interrupter switch is key-interlocked t o prevent making or breaking high currents caused by such mishaps as a lorn-voltage feeder circuit breaker failing to clear a secondary-feeder short circuit or excessively high inrush current from a group of large motors, etc. Where only one primary feeder is brought t o each substation (Fig. 11.8). this simple isolating function of the interrupter switch is all that is required. It should be pointed out, however, that in reenergizing a transformer the safest procedure is t o operate the switch with the primary feeder dcenergized. Otherwise, should something have happened in the transformer t o cause a fault, the switch would be subjected t o closing in an energized primary feeder on a fault. I n t.he secondary selective SYStem i t is easy t o drop the primary feeder long enough t o close the sjritch, because the load can tie transferred t o another secondary bus during the switching period.

632


t 750 KVA

300 KVA

480V

8

FIG. 11.8 Radial load-center dirtribulion system with interrupter switches ot primary of each load-center unit substation. Short-circuit protection provided by primary-feeder circuit breaker

Load-center System Overcurrent Protection. The National Electrical Code, Paragraph 4512b, allows for omission of individual primary overcurrent protection by the following rule: “ A transformer having an overcurrent device in the secondary connection, rated or set a t not more than 250 per cent of the rated secondary current of the transformer, or a transformer equipped with a coordinated thermal overload protection by the manufacturer, is not required to have a n individual overcurrent device in the primary connection provided the primary feeder overcurrent device is rated or set t o open a t a current value not more t.han six times the rated current of the transformer for transformers having not more than six per cent impedanre and not more than four times rated cufrent of thc transformers having more than sis but not more than ten per cent impedance.” Explaining that another way, overcurrent protection at each transformer (6 per cent Z maximum) can be omitted when the piimary circuit breaker is set to trip when the current flowing through it, is six times t.he primary current rating of the m~allesttransformer connected t o the feeder. For example, if a feeder fed two 750-kva and one 300-kva load-center unit substation transformers, Fig. 11.8A, and all were unprotected a t their primaries, the feeder circuit breaker would have to be set t6 trip a t 1800 kva, i.e., 6 X 300 kva-the smallest transformer on the feeder. This


633

setting is equal t o the sum of the transformer kva connected to the feeder. There may be sufficient diversity of loading of the substations that the total load on the feeder would never reach 1800 kva, thus the 1800-kva setting of the main primary-feeder circuit breaker would not result in a trip-out on normal load conditions. The maximum setting of the primary-feeder circuit breaker of six times the transformer rating for 6 per cent impedance transformers without individual transformer overcurrent protection has been selected to give reasonable assurance that the primary circuit breaker will trip for auy fault on the secondary of the transformers, poifits 1 to 4, Fig. 11.9. Where individual transformer primary overcurrent protection is omitted, the main secondary circuit breaker is required t o provide some degree of overcurrent protection for the transformer in so far as overloads are con-

THIS BRElKER MUST TRIP FOR F4ULTS A T 1,2,3 OR 4

MUST 4ND

FIG. 11.9

One-line diegram of a load-center system showing overcurrent protection required where individual overcurrent protection is omitted from the primary of the loadcenter unit substation transformer.

634


cerned. The primary-feeder circuit breaker provides short-circuit protection for the transformer, and the main secondary circuit. breaker provides protectiomagainst severe overloads. With the high set primaryfeeder circuit breaker the main secondary circuit breaker provides much better backup protection for the secondary-feeder circuit breakers should one fail to trip for a secondary-feeder fault, see Fig. 11.9. Also the main circuit breaker permits a,ny radial arrangement to be easily expanded into a secondary selective arrangement. I t also provides a means of quickly deenergizing a substation secondary bus in case of emergency. If all transformers have the same kva rating, then four to six or more transformers may be connected t o the same primary feeder without individual overcurrent protection. When as ma,ny as six or more transformers are connected to one primary feeder, there must be substantial diversity in the loads on the various substations; otherwise the primaryfeeder circuit breaker will trip out on normal overloads. I n the average system such as that shown in Fig. 11.8B, about four unit substations of equal kva rating can be connected to a given primary feeder with adequate over-all short-circuit protection. If the kva size of one of the unit substations is less than about one-fourth to one-sixth the tot,al load of the feeder, as indicated in Fig. 11.10, then a circuit breaker or an interrupter switch and fuse combination should be used on this smaller unit t o provide short.-circuit protection for that unit. The simple interrupter switch without the complication and expense and coordination problems of fuses is adequate for the other units on the same feeder, provided that the setting of the primary-feeder circuit breaker is not more than six times the kva rating of the smallest unprotected transformer. COST COMPARISON

I n all the following cost -comparisons, primary switchgear is omitted as i t is assumed that primary power will be availahle t o the site. The

FIG. 11.10

Load-center power system primary feeder serving both large and small loadcenter unit substations. Small load-center unit substations fused on primary.


636

the equipment being maintained. Many engineers feel that the deenergieation during maintenance is far more of a handicap than forced outages. Because forced outages would be so rare, they feel that temporary connections could be made to keep essential load going. If sufficient substation capacity is used in the radial arrangement to provide about 10 to 15 va per square foot of floor area, experience has shown that it mill adequately care for almost any diversity that will be encountered hecause of shifting of loads in most metal-fabricating plants where machine tools are frequently moved to rearrange production lines. This system appeals to plant engineers and maintenance men hecause

400 AMP PLUG IN BUS

---- --I

2 5 0 FT UNIT SUB

I

I

I

APPROXIMATE INSTALLED SYSTEM COST $39 KVA(O0 PERCENT) (1951) (LESS PRIMARY SWITCHGEAR) F!G. 11.1 1 Typical rodiol-circuit orrongement load-center power distribution system: (a), one-line diagram; (b), layout in plant.

LOAD-CENTER POWER SYSTEMS AND ClRCUll ARRANGEMENTS

!

637

I

FIG. 11.12

Stondard single-ended load-center unit substotion typic01 of t h e used in mdid load-center power dmtribution systems.

with adequate equipment properly installed it is d e , simple, easy t o @perateand easy to expand by merely extending a high-voltage feeder or ddding a new one and installing the load-center unit substation, This in no way afIeets substations or secondary circuits already in service. It 'des good voltage regulation because of its short secondary feeders, d when moderate-size substations, i.e., 500 to 1500 kva, are nsed the circuit levels are kept to limits which permit economical switching ent to be used in the secondary system. The equipment required ple and thus easy for the average electrician to care for properly. re 11.12 shows a single-ended load-center unit substation for a radial rcuit arrangement. The radial load-center system has been used in every type of plant there is. It has met the service reqnirements most exacting and critical processes. The excellent operating of this system is proof of its ability t o satisfy almost any power tion problem. NDARY SELECTIVE CIRCUIT ARUANGEMENT (Fig. 11.131

ere greater flexibility and the ability to take a primary naformer out of service and still maintain power on all second essential load are desired, the simple radialsystem can bemodified with

630


a normally open tie between pairs of substations or double-ended suhstations used to form a secondary selective arrangement, Fig. 11.13. The pairs of substations normally operate indepeudently of each other just a s they would in a straight, radial load-center system. However, in case of a forced or a scheduled outage of one primary feeder or transformer, the main cirruit breaker of the deeuergized substation is opened and the tie circuit breaker closed to supply power from the remaining energized substation bus of the pair of substations operated as a secondary selective pair. The secondary selective load-center system is in effect two radial systems with a secondary tie between them and, therefore, has all the appealing qualities of a radial system, i.e., low cost, safety, ease of operation, simplicity, good voltage regulation and can he cared for by the average electrician. Additional benefits are flexibility for operating purposes and theoretically somewhat greater service reliability because of the normally open secondary tie between suhstations. Complication is increased only very slightly hecause the ties are normally open. Short circuits are not increased hecause the transformers are not paralleled. Cost is increased about 10 to 31 per cent, depending largely upon the relative substation sizes and locations, the amount of reserve transformer capacity provided, and the exact form of the secondary selective arrangement used. It should be noted that with the secondary selective circuit arrangement, Fig. 11.13, the service reliability is a function of the system arrangement in contrast with some circuit arrangements, notably the primary selective, where the service reliability is largely dependent upon the type of equipment used. In popularity, the secondary selective load-center system ranks second t o the radial load-center system and accounts for most of the other loadcenter power distribution systems which are not radial. It has been used in ahout every kind of manufacturing plant and has met the most exacting requirements of the most critical processes in chemical plants, paper mills, steel mills, petroleum refineries, drug-manufacturing plants, etc. Because there is a tie between pairs of substation secondary buses in the secondary selective arrangement, there is a possibility of one transformer being connected t o two secondary buses to serve the loads normally supplied hy t,wo transformers. This may be t,rue when one of the pair of transformers operated in a secondary selective arrangement is out of service. Hence, the problem arises whether reserve transformer capacity should he provided in the secondary selective load-center system to allow one transformer to carry its own load plus its partner's load when its partner is out of service. Engineering judgment on the part of the power-system designer should determine how much, if any, reserve transformer capacity is required in any part'icular secondary selective load-center system. To help the sys-

LOAD-CENTER P O W E R SYSTEMS A N D CIRCUIT ARRANGEMENTS

639

-

1

TWO SINGLE

TIE

2i"

3250 F T

101 250 FT

00

FT 600 F

iCOSTS 0 0 NOT INCi APPROXIMATE INSTALLED SYSTEM COST 1 4 5 PER KVA 1115%1 S U B S T I T U T I N G 2.1000 UVA SUBSTATIONS INSTALLED C O S T . ON 1500 NVA TOTAL LOAD B A S I S - I 5 1 PER XVA( 1 1 1 XI-119511 Ibl

L

PRlM4RY SWlTCHGE4RI 4PPROIlM4TE SYSTEM COST '143 PER K Y 4 I l l 0 %I SUBSTITUTING 2-1000 L V 4 SUPST4110NS. I N S T A L L E D COST ON 1500 KVA T C T 4 L LCAD B 4 S l S 1 4 8 5 0 PER K Y 4 119511 1 1 2 4 % 1

I< I Typical secondary selective circuit arrangement load-center power' dirtribution system; (01, one-line diagram; Ibl and (
FIG. 11.13

640

LOAD-CENTER P O W E R SYSTEMS AND CIRCUIT A R R A N G E M E N T S

tem designer arrive at a sound decision on this problem, certain hasic facts should be considered. Pirst, examine the need for reserve transformer caparity over and above that needed for normal load growth. Referring to operating records of the major components that may fail and cause a forced outage, it is evident that a forced outage would be so rare (one transformer outage for 2500 transformer years and one primary cable outage per 100 circuit miles per year) as t o warrant little if any extra investment just t o provide substantial reserve subsbation transformer capacity. In most plants there is some vitally essential load like life tests, lights, continuous processes, ovens, etc., which it is desirable to keep going even though a primary feeder or a transformer serving this load may be out of service because of a forced outage or for service and maintenance. I n all except a very few cases, such as main water-supply pump motors or boilerhouse auxiliary motors, the percentage of vitally essential load to total load is small, of the order of 25 to 50 per cent of the total load. If such load is even as much as 50 per cent of the total load, then one transformer can carry, in an emergency, the vitally essential load on two substation buses without overloading the transformer remaining in service. This of course requires dropping the nonessential load on both buses of the pair of secondary selective load-center suhstat.ions. Even though this vitally essential load should he as high as 60 to 75 per cent of the total load, the one remaining transformer might still be able to carry all the vitally essential load on two buses for short periods of time. This would represent a n overload of 20 to 50 per cent on the remaining transformer for the short duration of the emergency. I n such a n emergency, supplemental cooling may be applied to the transformers and secondary switchgear by means of portable fans, etc., to keep temperature rises from reaching damaging levels. While it is true that large overloads do tend to reduce the life of insulation at an accelerated rate, in general, the total time in the life of the average suhstation during which it will he subjected to such overloads in secondary selective load-center systems is so small as t o he negligible. Therefore, it would seem that there can be little economic justification when considering the problem of reserve transformer capacity from a purely insurance viewpoint for spending much if any money for reserve transformer capacity in a secondary selective loadcenter system to take care of all load in the case of forced outages. In the few cases where it is felt that some reserve transformer capacity is necessary in a secondary selective load-center substation, it can be obtained at low cost if it can be obtained on an incremental basis, i.e., by using two 1000-kva substations instead of two 750-kva suhstations. I n Fig. 11.136 it is shown that providing nearly 50 per cent reserve transformer capacity increased the cost from $45 t o $51 per kva, or about 13 per cent.


A

641

t-I L

4.0. UORMALLY OPEN TIE1

LOAD CENTER UNIT SUBSTATIONS

RADIAL SYSTEM NO. I

L

Secondary selective load-center system made by placing normally open tie between series of radial load-center unit substation secondary buses.

FIG. 11.14

642


I n view of the foregoing, the most common way to design a secondary selective load-center system is to approach it on the hasis of loading the transformers just as in the case of a radial load-center system, providing little if any reserve transformer capacity except that to care for normal load growth. After that is done, normally open tie circuits are provided between pairs of secondary buses, Fig. 11.14. This approach keeps the system investment t o a minimum and yet provides the most basic and important advantage of the secondary selective circuit arrangement, and that is to have emergency power available a t every important secondary bns from another substation bus through the normally open secondary tie. In larger systems, where the secondary selective load-center system finds its greatest application, there is generally more than one primary feeder and many substations; so the advantage of the secondary selective circuit arrangement can he obtained for only the cost of the normally open secondary tie. I t further permits a complete division of the system from source to secondary bus, as shown in Fig. 11.14. Hence, any part of the system from the source through the load-center substation main secondary circuit breaker can fail or he taken out of service for maintenance, inspection, adding new circuits, etc., and yet power can he available a t every secondary bus by merely opening the proper substation main circuit breaker A , Fig. 11.14, and closing the proper tie circuit breaker B . Spot Secondary Selective Arrangement. The secondary selective arrangement can be put together in several forms. A one-line diagram of the spot arrangement is shown to the right in Fig. 11.13. Pairs of singleended substations located side by side with a bus tie asinFig. 11.15 may also be used. The spot secondary selective arrangement is simple and compact as all interconnections and switching are a t one location for any pair of unit substation transformers. It is generally slightly lower in cost (109 per cent compared with 115 per cent or more for the distributed secondary

GOAHON OPERATING AISLE

FIG. 11.15 One-line diagram of a spot secondary selective arrangement using two single-ended loadcenter unit rubrtations connected together through CI tie bur.


643

selective arrangement) than other forms using distributed single-ended substations. The two single-ended units provide the advantage that extra secondary rircuits can be added easily. In the double-ended arrangement, provision must be made in the beginning for the extra secondary feeder positions that may be required in the future. Distributed Secondary Selective Arrangement. Another form is the distributed secondary selective arrangement shown to the left in Fig. 11.13 and in Fig. 11.20 under comparison of secondary selective and secondary networks. In the distributed arrangement the substations serve smaller load areas, thus reduce the secondary feeder cable required. This saving is partially offset by the extra tie cable required. Many operators prefer this form for a new plant with the idea that as load grows another transformer can be added a t each existing substation to form a spot secondary selective arrangement to provide for a load growth of 100 per cent in a given area. This arrangement will cost about 115 per cent. In some cases a reduction in first cost of the distributed secondary selective load-center system has been achieved by merely making provisions for an emergency secondary tie circuit to be installed when the need for the tie circuit arises. This is particularly applicable in mills where complete periodic shutdowns are made for inspection and maintenance of process equipment. At the same time, the maintenance of the service facilities and the electric power system equipment is taken care of. There is practically no load in such mills when shut down. Such provision ran he easily made by merely having one drawout circuit-breaker compartment in each unit substation complete with all connections but without the removable circuit.-breaker element, Fig. 11.16. .I few (two a t least) spare circuit-breaker removable elements t o fit into the empty emergency tie circuit-breaker compartments and enough cable to make the emergency tie circuit must he kept on hand ready for immediate use. When the need for the tie circuit does arise, then the circuit-breaker elements are placed in the proper substation tie breaker compartments and a temporary or permanent tie circuit cable is connected between the outgoing terminals of these tiecircuit-breaker positions. The time required t o do this of course depends upon proper preplanning for such a contingency and upon having the proper circuit-breaker elements and enough of the proper size and voltage-rating cable readily available and accessible for immediate use. Automatic-transfer Secondory Selective Arrangement. Automatic transfer can be used to disconnect quickly a deenergized secondary bus from its transformer and connect the bus t o its energized “partner” in the secondary selective arrangement. The automatic transfer can be designed t o reconnect the system t o normal when proper voltage is

644


TIE CIRCUIT

WHEN T I E BETWEEN SUBSTATIONS IS NEEDED INSERT TIE BREAKERS IN POSITIONS TI AND T 2 AND INSTALL TEMPORARY (OR PERMANENT) TIE CABLE TC. FIG. 11.16 Low-cost form of distributed secondary selective load-center ryrtem with the tie circuit installed as needed.

restored t o the transformers that were out. The complication to do this is generally not warranted, and manual restoration is used. There are factors t o consider in connection with the automatic throwover secondary selective arrangement. The system is complicated and costly. Where the total load of one transformer is transferred to another transformer, depending upon the sum of the two loads, this may overload the one transformer too greatly. It usually takes a t least 1 sec to make the transfer. This will allow motors t o slow down or stop, and when they w e all reenergized a t once, there may he a very large inrnsh current which could cause tripping of the feeder circuit breakers. This can he overcome by putting enough time delay in the transfer means to allow all motor starters except


A45

those on very important motors t o drop out. The automatic transfer will then quickly reestablish lights and esseutial service. This has two advantages, the load transferred is less and the inrush current is materially rednced. If there are only a few smaller motors that must be kept on, their inrush current may not be heavy compared with the transformer rating supplying the system. A manual means of transferring for normal maintenance where both

LT

cli

T 'p

CJ-

I +>+ ; + . CENTRAL POWER TRANSFORMER

EREAKERS A E ANO C M A N U A L L I CLOSED ELECTRICALLY ANO MANUALLY TRIPPEO

-

m

XY

"A" AUXILIARY SWITCHES ON CIRCUIT BREAKERS

A B + H H

C

H

R SELECTOR SWITCH CONTACTS EREAKER T R I P COILS

H R = HESITATING RELAY TO AÇSURE PROPER ÇEOUENCE OF OPERATION. OPERATION: PLACE SELECTOR SWITCH TO BREAKER TO BE TRIPPED ( C I N THIS CASE). THEN MANUALLY GLOSE BREAKER WHICH IS OPEN (E IN THIÇ CASE ).AS SOON AS E GLOSES A L L AUXILIARY SWITCHES ARE CLOSED AND EREAKER C IS TRIPPED AFTER SOME TIME DELAY INTRODUCED BY RELAY H R .

FIG. 11.17

Connectionr for manual tranrfer of load from m e recondory bur to the

othei in a secondary selective ryrtem to permit removingone tranrformer or primary feeder without dropping the recondary service.

646


secondary buses are t o he kept energized is shown in Fig. 11.17. I n this schemc the tie circuit breaker is closed first; and as soon as it closes, the previously selected main secondary circuit breaker is opened. This scheme does momentarily parallel t,he transformers, and should a heavy feeder fault occur at the instant the transfer is being made, the shortcircuit,-current duty would be that. of two transformers instead of one. To transfer when one hus has been deenergized because of a primarysystem fault requires that, t,he main secondary circuit breaker feeding the deenergized bus be opened manually before the transfer is made. If this circuit hreaker were not opened, both secondary buses may be deenergized as soon as the tie circuit hreaker is closed as the dead bus would he directly connected t o a fault on the primary feeder because the main secondary r,ircuit, breaker on t,he dead bus mas closed. There are many other forms of the secondary selective arrangement. However, t,hose described cover the common usage. SECONDARY N E T W O R K ARRANGEMENT (Fig. 11.18)

Serondary network systems were originally developed by central station utilities for distributing electric power in the downtown or commercial arcas of large cities. When electric systems and equipment specifically developed for a given purpose in one field are heing adapted for use in another field. it is logi(,al to review thoroughly the requirements of the nrw field t o determine wherein the rcqnirements differ. Serondary network arrangements, Fig. 11.18, differ from the radial, secondary selertive, arid primary select,ive arrangements in two respects: (1) the transformer secondaries are operated in parallel, and ( 2 ) in the event, of a primmy feeder or transformer failure, the faulty circuit is switched off aut,nrnatiidy by opening of t,he proper primary-feeder circuit hreaker and the autornat,ic operation of the network protector in the secondary circuit of !he transformer. T h e operation of the industrial secondary network system is as follows: The net~vorkprotertor, if it has a full complement of protective equipmeut, consists hasically of an elertrirally operated air circuit breaker automatirally ront,rolled hy a power directional tripping relay and reclosiug relays. In addition, overvurrent protection is afforded either by fuses in serics with the network protector (for 208Y/120-volt systems) or direct-acting trips on t,he air-circuit-breaker portion of the network protrctor. \Vhen a primary feeder or transformer failure occurs, the directional relay closes it,s contacts on reverse pmver flow (i.e., power flow from the secondary system t o the t,ransformer) and trips all net,\rork protectors wniiertcd t,o the faulty feeder. This automatically isolates a primary fan!t lvithout dropping load. Iluring the several cycles or so it takes the


647

600 F T

500FT

(b)

APPROXIMATE INSTALLED SYSTEM COST $ 7 8 00 PER K V A ( 2 0 0 X ) (LESS PRIMARY SWITCH GEAR)

FIG. 11.18

1951

Typical recondory network arrongement load-center power distribution system; la), one-line diagram; (b), layout in plant.

648


primary-feeder breaker to open, the voltage on the network may go down to almost zero. It will then rise to 50 per cent or more until a second or so later when the proper network protectors open. When the fault is repaired and proper voltage is restored, the power directional relaying acting as a voltage relay in conjunrtiori with a phasing voltage relay causes t,he protector to close automatically only when voltage conditions are such that power will flow from the primary to the serondary system. The above description of the operation of the network protector is very abbreviated. Complete Network Protector Controls Are Not Necessary in Industrial Plants. The use of network protectors and all the associated relays as

developed for the utility network systems are not neressary for itidustrial network systems. Since distances are not great and all power usually comes from one source, the automatic reclosing feature can he eliminated and manual closing used. Further simplification can he done by using standard electrically operated air circuit. breakers in metal-enclosed switchgear with standard reverse-power relays instead of network protectors and the highly sensitive network-type reverse-power relays. Advantages and Disadvantages of Parallel Operation of Transformers in laad-center Systems. Recause the transformers are operated

in parallel and automatic relaying is used, the secondary network arrangement offers slightly greater service reliability than radial-type load-center systems. The advantage is very slight because even the radial arrangement has a service-reliability record of over 99.99 per cent plus. Because power can be transferred over the secondary ties from one substation to another in the network arrangement, this in some cases reduces the amount of transformer capacity required compared with the capacity required in the radial arrangements. The parallelling of the secondaries of the transformers increases the short-circuit-current duty and therehy necessitates higher interrupting capacity and therefore mbre expensive secondary circuit hreakers than the comparable radial system. It also requires considerable additional secondary cable to interconnect the secondaries of all transformers t o form a secondary network. These tivo factors generally more than offset the saving in transformer capacity and as a result increase the first cost of the secondary network arraugement up t o 50 per cent or more compared with radial or secondary selective arrangements. Parallel operation generally adds materially to the complexity of operation of the secondary network arrangement. There are tivo sources of power on the tie circuits. Both ends must he opened to deenergize the tie rirruits i:omplet,ely. More complicated relays are required on the network protectors than on conventional air circuit hreakers used in radial load-center systems. This arrangement rail be made as safe as any by

LOAD-CENTER POWER SYSTEMS ANO CIRCUIT ARRANGEMENTS

649

the use of unit substations consisting of adequate interrupting-rating switchgear of modern metal-enclosed construction, properly installed. Steady-state voltage drop is about the same in the secondary network arrangement as in the other systems because transformers carry about the same loading and secondary radial feeders are about the same length and loading. Because the transformers are in parallel and the short-circuit currents are higher, there will be less voltage drop when starting motors or using welders. This may be significant when motors of severa1 hundred horsepower and larger welders are used ou low-voltage systems. Because all transformers are in parallel, voltage variations will show up over a wide portion or a11 over the system. I n the other arraugements the critical loads can be put on one bus and noncritical loads on another bus (see Chap. 10). Secondary networks have two hasic forms, i.e., the distributed network, Fig. 11.18, and the spot network, Fig. 11.19.

ult

L

350MCM CABLE

__ __ _ _ _ _ _

NO.OO .-- -CABLE - - - - L _ __ _ - - - $ J 1500KVA SPOTI

I

f

NETWOAK 2-750TRANS

250 FT

APPROXIMATE INSTALLEO SYSTEM COST $48 PER KVA fi227,) (LESS PRIMARY SWITCH GEAR)

FIG. 11.19 Typical spot sccondary network arrangement lood-centsr powcr dirhibutim ryrtem: (top), one-linc diagram; (bottom), IayOut in plant.

650


Spot Network (Fig. 11.19). The spot network, if limited to two 750kva 480-volt transformers per bus, keeps short-circuit currents within 50,000 amp and provides high service continuity for station auxiliaries, etc. Larger trarisformers in t,he spot, network require 75,000-amp interrupting-rating feeder circuit breakers, thus make this syst,em cost prohibitive. A secondary bus failure drops all load, whereas in a secondary sclect,ive arrangement it drops only half the load. The cost of the spot network using two 750-kva t,ransformers per bus is 122 per cent compared with 110 per cent for the spot secondary selective system. No reserve transformer capacity is considered i n either case, and directional tripping and manual reclosing are considered only in the spot network. COMPARISON OF SECONDARY SELECTIVE AND SECONDARY NETWORK LOAD-CENTER SYSTEM ARRANGEMENTS

A comparison of the secondary network and secondary selective arrangements will he made on ot,hcr major points than those mentioned in the preceding paragraphs. Cost Comparison. While the cost of an industrial power distribution system represents hut a few per cent of the total plant investment, the cost of one system vs. another is still of prime significance if the systems t o he compared give substantially the same operating characteristics and meet the requirements of a particular plant. A secondary network arrangement, Fig. 11.18, and a secondary selective arrangement, Fig. 11.20, have been compared on a first-cost basis. The secondary network load-center system consists of siu 500-kva substations and serondary tie circuits. The load-center unit substation size is limited to 500 kva to keep the short-circuit duty within 50,000 amp on the feeder circuit breakers. Fewer larger network substations would have made it difficult to keep the interrupting duty down t o 50,000 amp. The secondary selective load-center system consists of six 750-kva substations loaded to an average of 500 kva each. The cost of the secondary network load-center system, Fig. 11.18, is $78 per kva (200 per cent on the hasis of the radial system, Fig. 11.11). The cost of comparable equipment for the secondary selective load-center system, Fig. 11.20, is $63 per kva (161 per cent). A word of explanation about the various system costs is in order. The secondary selective arrangement of Fig. 11.20 is almost exactly like that of Fig. 11.13. In Fig. 11.13 there was assumed no reserve transformer capacity when 750-kva substations were used. The cost was $45 per kva (115 per cent). Assuming full load on each transformer, i.e., 750 kva in Fig. 11.20, the cost per kva would be almost the same. However, in Fig. 11.20 the substations, for comparison with the secondary network

IS9

A 091b

S I N 3 W 3 3 N W W llll3813 a N V SW31SAS a3MOd 8 3 1 N 3 3 - a V 0 1

652


arrangement, are assumed to he loaded t o an average of only 500 kva each; thus this automatically increases the system cost per kva of load by nearly 50 per cent. The load areas are 500 kva each, whereas the substation capacity is 750 kva per load area-50 per cent over the actual average load. That accounts for the cost going up from $45 (115 per cent) for Fig. 1 1 . 1 1 t o $63 (161 per cent) for Fig. 11.20 per kva for the secondary selective arrangement. Even so, the secondary network loadcenter system costs 25 per cent more than the secondary selective loadcenter system. There is 50 per cent more transformer capacity in the secondary selective arrangement, which will be far more than is required in the average plant t o care for diversity. Could that he reduced to 25 per cent more transformer capacity, then the difference in cost would be widened to where the network load-center system would cost about 50 per cent more than the secondary selective load-center system. This has been verified in several checks of actual system designs. The figure would probably he between 25 and 50 per cent for the secondary network load-center system in the average case and where the same quality and adequate rating equipment was used throughout. Caring for Load Shift. One of the inherent characteristics of the secondary network load-center system is that, if the load a t one transformer substation increases beyond the rating of the transformer supplying that bus, the adjacent substations will supply part of this load over the secondary tie circuits provided that the adjacent substation is not overloaded. There is, however, a limit as t o how much load can be placed on one suhstation bus. This is determined by tie-circuit impedance, transformer impedance, and the nature of the load. To compensate for this load shift, larger transformers (750-kva transformers with an average load of 500 kva) are required in a secondary selective load-center system. This is taken into account as outlined under cost comparisons. The ability of the secondary network load-center system to divide all loads among all substations is desirahle from the standpoint of economics where there are large shifts in the load a t any one location in the plant from day to day or week to week. A dock or a pier for servicing ships is a typical example of where such large load shifts may occur. A ship may tie up a t one end of the dock and require power service somewhere in the neighborhood of the total transformer capacity on the pier. The next ship that requires service may tie up on an opposite corner of the pier. I n the average manufacturing plant, such violent day-to-day or weekto-week load shifts are not often encountered. Installations of secondary selective load-center systems in hundreds of industrial plants of all types have shown that this system can adequately care for the average shifting of loads on a day-to-day or week-twweek basis. Therefore, the great majority of industrial users have selected the simple secondary selective


653

load-center system in preference to the secondary network load-center system because it will meet their requirements a t a saving in first cost of about 25 to 50 per cent as compared with secondary network load-center system. Service Reliability. The secondary network arrangement offers a somewhat higher degree of service reliability than does the secoudary selective arrangement hecause a faulty primary feeder is removed automatically by the network protectors, whereas in a secondary selective arrangement it is necessary t o operate two or three feeder circuit breakers to transfer the load at one substation to another. During this emergency condition, the overload on a transformer in the secondary network arrangement will be less than in the secondary selective arrangement because the load dropped by one or two transformers in the network arrangement can be absorbed by all the remaining transformers. In the secondary selective arrangement,, the load is carried by the “pardner” of the transformer that is out after switching the necessary secondary circuit breakers. I t is not uncommon during such emergencies t o overload transformers considerahly. This will generally enable full production to be maintained. However, in some cases, this has necessitated dropping some nonessential load where secondary selective load-center systems are used. The failure rate of transformers and primary cable is so low in the average industrial plant that most operators feel they cannot justify anything except the lowest. cost radial load-center system or at the most the secondary selective load-center system. They reason, and justifiably so, that most outages to service are caused by faults in the serondary branch circuits and the next worst offender is secondary feeders. There are many more outages in these parts of the system than in the transformers or primary cables. Since the secondary feeders and branch circuits are radial in both the secondary network or secondary selective load-center systems, neither will have any effect on reducing the secondary-feeder or branch-circuit outages. On the other hand, some operators feel that a major advantage of duplication of sources down to the secondary buses is to enable power to be maintained at all secondary buses when primary circuits are out for test, maintenance, or expansion. While there are many more outa.ges due to secondary-feeder faults than due to primary-cable or transformer failure, there is not such a large block of power affected when a secondary fault occurs as when a primary system fault occurs. Because of the larger blocks of power affected and to allow for normal maintenance, primary circuit testing, or expansion, many industries considered it economically justifiable t o have some means of getting power during a primary circuit emergency or during maintenarice t o all secondary buses.

654


The secondary selective arrangement provides this at minimum cost. Only a few minutes outage is encountered on the rare occurrence of a primary-circuit failure-just time enough t o make the transfer by operating two or three circuit breakers a t the affected substations. For normal maintenance the transfer can be made on the secondary selective load-center system without dropping any load. The advantage of the secondary network arrangement from a service-reliability standpoint is t o automatically isolate a faulty primary circuit on the occurrence of a primary fault. Ease of Operation. The secondary selective arrangement consists of only the simplest type of eqnipment, that is, transformers and manually operated drawout air circuit breakers. Operating personnel usually require no special training for the operation and maintenance of this equipment. The secondary network arrangement, however, consists of many parallel ties and secondary network protectors with complicated relay arrangements. When servicing a part of the secondary network load-center syst,em, at least. t,wo, and somet,imes as many as five, circuit breakers must be operat,ed to deenergize any portion of the system. I n the secondary selective load-center system, opening one circuit breaker will deenergize completely any one part of the syst,em. Specially trained maintenance persorinel are required t o care for the relays on the network protectors. I n some plant,s where secondary network load-center systems are employed, the network relays are cared for by the local utility maintenancc personnel if t,he local utility happens t o have a secondary network system and will agree to maintain industrial plant property. Voltage Regulation. The normal voltage regulation in either type of system is comparahlc. The network arrangement having a higher shortcircuit level on the secondary huses, therefore, being a “stiffer system,” can accommodate the starting of larger motors for a given drop in voltage. With the secondary selective system with transformers of the order of 500 t o 750 kva, motors as large as 50 t o 100 hp may be started across the line without causing enough’ voltage drop t o produce objectionable light flicker. Secondary Tie Capacity. An important prohlem in the design of secondary network load-center systems is selecting the amount of tie capacity required between suhstations. If no load is tapped off the ties, then it is generally considered adequate from a system design standpoint t o provide tie-?able capacity equivalent t o about two-thirds of the capacity of the largest substat,ion in the secondary network load-center system. If load is tapped off the ties, then the tie capacity should be increased t o 100 per cent, of the rating of the largest substation in the system.


655

PRIMARY SELECTIVE ARRANGEMENT (Fig. 11.21)

The primary selective arrangement differs from the radial and the secondary selective arrangements in that two primary feeders are brought to each substation (Fig. 11.21). Since two primary feeders are available, a switching means is necessary to select the desired feeder t o connect to the substation. The switching function is different also in that there is a transfer furictivn to be performed arid not the mere disconnecting of a primary feeder as in the other arrangements which have only one primary feeder coming to each substation. In discussing any engineering problem such as this, there are always two things that are important: (1) what are the inherent engineering aspects of the system as dictated by the one-line diagram, and (2) what kind of equipment is necessary to enable these characteristics to be obtained with safety? Safety, particularly in SO far as human life is concerned, is of first importance. Reliability is important but is secondary to safety t o human life. The safe way to make the transfer is with adequate power circuit breakers, as shown in Fig. 11.21. However, the cost of such an arrangement may not be justified; particularly if used on a high-capacity 13.8-kv system or if very small unit substations are used. Should the cost of such a primary selective load-center system with power circuit breakers for the transfer means he unreasonable, it is suggested that the secondary selective arrangement be used which has many advantages over, and on an average rosts less t,han, the primary selective arrangement with primary power circuit breakers (sce Fig. 11.22 for comparison). In an attempt to reduce the cost of primary selective load-center systems, interrupter switches are often proposed as a substitute for power circiiit breakers, Fig. 11.23.

T

T c

+,

A o , r . I p o

)

I

)

)

+ + + +

Y Y Y Y

A

+ + , r e >

1

&

)

+

)

J

+

Y Y Y Y

FIG. 11.21 Primary selective circuit arrangement load-center power distribution system with power circuit breakers for tronrferring from one p r i m ~ r yfeeder to the other.

656


s

A

$: A

f* ?*f + ' F *,

** , h* , *T ?+

Y Y Y Y

Y Y Y Y

SECONDARY SELECTIVE

SEC

SYSTEM C O M P A R I S O N I. C O S T - _

-_

_ _ _ - _ _ _ _ _ -----_-_

2. SERVICE RELIABILITY

PR I

SEL

SEL

- - $45KVA

845-655KVA I I I 5 7 -140-/d

I I 15%)

_ --- __

_

* O N E PRIMARY FEEDER OUT- _SAME CAPACITY ONE P R I M A R Y FEEDER OUT75% loox b * O N E TRANSFORMER O U T - - _ - -. 75% 0% c. 'MAIN BREAKER OUT 75% 0% d. SECONDARY BUS OUT 50% 0% *OUT EITHER FOR MAINTENANCE OR DUE TO FAILURE 3. SAFETY SAME 4. SPACE _ - -- - - - - - - - - - _ -LESS 5. LOSSES _______-SAME 6. VOLTAGE DROP I S T E A O Y STATE)--__ SAME _. 7 VOLTAGE DROPIMOTOR S T A R T I N G ) --I12 8. COMPLEXITY OF OPERATION __- SAME 0.

__ __

______ _______________ _______ ___ --___

______________________________

_ _ _ _ _ _ _ _ _ _ __ _ _ _ ______--_--_____-

____ ___- _ _--

_-

_____________ -

_ _ _ __ _ _ _ _ - _ _

FIG.1 1.22

___

Comporiron of primary selective and secondary selective circuit arrongemenlr.

POTHEADS


657

Whether the service-reliability feature of having two primary feeders available is or is riot realized when interrupter switches are used depends upon the design arid arrangement of the interrupter transfer switrhes. Two types of switches are available: ooc is the double-throw type; the other consists of two separate single-throw switches. Most double-throw switches necessitate bringing both feeders into the same compartment, whereas two separate switches enable termination of the feeder cables in separate compartments. This accounts for the real difference in service reliability and a major portion of the difference in safety obtained from the two different interrupter switch transfer means. Both arrangements involve a real hazard should a transfer be made with a fault on the load side of the transfer switch with the primary feeder energized. These differenres are now to be considered in detail. With both potheads adjacent t o each other, a failure in one will more than likely involve both and shut down both feeders. Furthermore, there is no possibility of isolating one feeder and its pothead through an air break, so that it can he worked on in safety with the second feedcr energized. In any throw-over scheme with interrupter switches, there is always the hazard of operating under fault conditions. To overcome this hazard the operator should, after a tripout, shut down both feeders, make the transfer, and then rerlose the good feeder. Although the transformer failures are rare, when a primary feeder trips out, the operator may not know whether the fault is in a transformer or the cable. If this failure should happen to he in the transformer and if the operator attempts t o restore service by throwing the double-throw primary switch t o the opposite position without first deenergizing the good primary feeder, he will connect the faulty transformer to a live feeder through the interrupter switch. Operation of interrupter switches under such conditions may result in fatalities and a complete shutdown of both feeders. Contrasted with this, the interrupter switch on a radial or secondary selective arrangement is for maintenance purposes only and not for operation or for switching of the system. The hazard of transferring a faulty transformer is not present in these arrangements because there is no second feeder available. The cost of the radial arrangement and the primary selective arrangement with double-throw interrupter switches cannot be compared directly. The cost of a primary selective substation with double-throw switches varies compared with radial substation from slightly more, percentagewise, for larger substation t o substantially more, percentagewise, for very small substations (those of approximately 200 kva or less). The real increase in cost of the primary selective arrangement compared with

658


the radial arrangement is in the duplicate primary feeder cable (at a cost of $5 to $15 per ft) required for the primary selective arrangement. The operation and safety characteristics of the primary selective arrangement with double-throw switches and a straight radial arrangement are summarized in Fig. 11.24. Based on the foregoing, it is felt that either the radial or the secondary selective arrangement should be used instead of the primary selective arrangement with double-throw interrupter switches. The secondary selective arrangement has better service reliability, and the radial arrangement has,essentially as good service reliability. Neither arrangement has the safety hazard of the primary selective arrangement with a doublethrow interrupter switch. The sole reason for using a primary selective instead of a radial loadcenter system is to provide service to the load-center substation primary in the event of a primary-cable failure. As pointed out in the beginning of this chapter, primary-cable failure rates are about one per 100 miles of cable per year. This failure rate is low enough to question the extra investment of the added primary cable of a primary selective load-center system compared with a radial load-center system. The trend toward

__._._..-.---SLIGHTLY

FIG. 11.24

Cornparimon of

MORE COMPLICATED

radial and primary selective circuit arrangements.

LOAD-CENTER POWER SYSTEMS A N 0 CIRCUIT ARRANGEMENTS

660

which permits, with proper precautions, the servicing of either feeder pothead with the other feeder energized. T o be the more certain of getting this higher quality equipment, the

FIG. 11.26

Primary selective loed-center system with two single-throw interlocked pri-

mary switches of the type shown in Fig. 11.25.

PRIMARY SELECTIVE ARRANGEMENT

TWO SEPARATE SWITCHES

ONE DOUBLE-THROW SWITCH TWO SWITCHES

COMPARISON

......................

I. COST E. SERVICE RELlABlLllY o.CABLE FAILURE b. POTHEAD FAILURE c.SWlTCH FAILURE c. TRANSFORMER FAILURE f. MAINTENANCE OF SWITCH 3. SAFETY IN OPERATION 4. SPACE 5. LOSSES AND VOLTAGE DROP 6. COMPLEXITY OF OPERATION

---__.SLIGHTLY

MORE

OK OK OUT OUT DUT POOR SLIGHTLY MORE SAME SAME

_________________--___

___________-__________

________-__-_____ _____________________----____________________

FIG. 11.27

ONE OT.

sWlrCH (TO MAKE OUT REPAIRS1 OUT OUT OUT OUT POOR

Comporiron of primary selective load-center systems using two single-throw

interrupter switches VI. one double-throw interrupter switch.


661

one-line diagram of the circuit should always be drawn as shown in Fig. 11.26 t o indicate clearly two single-throw switches. However, two singlethrow swit,ches st,ill involve the hazard of transferring a faulty transformer t o an energized feeder by means of the interrupter switch. Although the probability of this may be remote, iu the interest of safet,y the transfer from one feeder t o another with any kind of interrupter switch should always he made wit,h hoth feeders deenergized. As poiuted out previously, the highest degree of safety in the primary selective arrangement can be realized only by using incoming-line power circuit breakers. Therefore, the reduction in cost by using interrupter switches instead of power circuit breakers may very likely be a t an undue sacrifice in safety. However, if after a review of all facts it is felt that a primary selerhve system with interrupter switches is required, then an arrangement as shown in Figs. 11.25 and 11.26 with two interlocked single-throw switchcs should be used in preference t o the arrangement with double-throw switches, for the reason outlined previously and summarized in Fig. 11.27. LOOPED PRIMARY FEEDER LOAD-CENTER SYSTEMS

The systems so far considered all have radial primary feeders. l n some cases, looped primary feeders are proposed for load-center power systems such as that illustrated in Fig. 11.28. See Chap. 13 for a more complete discussion of looped primary systems. Sectionalieation of the loop requires switching means a t the points indicated numerically. The only really safe way t o sectionalize a n energized looped primary is with properly applied power circuit hreakers at these points as shown in Fig. 11.29. As in the primary select,ive load-center system, the cost of power circuit breakers for the switching furiction where small substations are used may he prohibitive, and thus there is often the tendeucy t o cheapen the looped primary system by using interrupter switches, as shown in Fig. 11.28, instead of power circuit. breakers. When this is done, practically all the advantages of the loop that seem apparent in the one-line diagram are lost, so far as service reliability is concerned, and safety is brought t o a very low order for the following reasons: I . A fault anywhere drops all load until the fault is found and isolated. This may require several hours. 2. There is no practical way of interlocking the interrupter switches to assure that they will not he operated under short-circuit conditions. 3. If certairi types of separately mounted oil-filled switches are used, there is no visual means of being sure the circuits are cleared when it is desired t o work on one piece of cable with the rest of the system energized. Oil may darken and make sight glasses useless.

662


Most operators will not permit working on a cable isolated only by contacts under oil because of the possibility of leakage through old carbonized or wet oil. The most serious disadvantage of this system is the hazard of operating the interrupter switches under high overload or fault conditions. While i t is easy t o consider academically what, an operator will do when a fault occurs, the result is too apt t o work out differently in actual practice. I n the first place, faults do not occur every day; so, no matter how good the operating instruct,ions, t,here is little if any occasion t o practice operating the systcm properly. Hence, in the rush t o restore service after a fault has occurred, the operator can too easily make many possible mistakes. First of all, when a one-line diagram is visible, the entire system is before one's eyes arid it is easy t o scc what t o do. But when the oprrat,or leaves the main substation t o operate a switch, he may have only a mental pic-

FIG. 11.28 Looped primary circuit bod-center system with rectionalizing switches for supplying load-cenfer unit substations.

FIG. 1 1.29

Looped primary circuit bod-center ryrtem with rectionalizing power circuit breakers for supplying load-center unit rubstotionr.


663

ture of the system and may have t o rely eutirely on memory. Because i t is easy t o become ronfused when under pressure, he may go t o the wrong switch, or he may dose instead of open the switch. For example, fallow through the case of isolating a cable fault, say a t X in Fig. 11.28, assuming of course that the operator does uot know where the fault is but only knows that hoth rircuit breakers tripped out. This is 110 help i n lorating t,hc fault. It merely indicates the occurrence of a fault somewhere. The first opcration may be t o open switches 1 and 6 and reclose both circuit breakers. Since the circuit breakers stay in, the fault has not been found. So. the operator will start through the system throwing switches and closing circuit breakers, hoping that the sequence is correct: for, if it is not, the results may be a switch blowup and a iseverely injured man. l f the operator is lucky and starts at the left side, the fault will he isolated a t the third try because the second operatiori of closing switch 1 and opening switch 2 will show that the fault is not in the transformer. The third operatiori is t o close switch 2 and open switch 1. This mill show where the fault is. 4 little study of Fig. 11.28 reveals how many more times this process would have t o he repeated before the fault is located, if the s\vitching operations had started at the right. Essentially the power circuit breakers at the source should be opened every time a w i t c h is thrown. If the circuit hreaker were not opened, because of an error or misunderstood dircetions, and the S o . 2 switch in this example was rlosed, then it would he closed on a fault. The operation is so romplicated that the probability of making a mistake is very high and is fared over and over again. Tho suggested alternative t o any kind of a loop primary feeder loadcenter system is always either a straight, radial or a secondary selective load-cent,er system. In some layouts where there are scattered small huildings (such as nurses' or attendants' quarters a t a hospital) the straight radial systcmshould serve well enough, because service continuity in these is no more essential t,han in ordinary di~ellings,most of which are fed by radial systcms. Howcvcr, some engineers may feel that, since the feeders to these isolated huildings are underground, some primary-feeder duplication is necessary. If so, and the secondary selective arrangement does not tit, then the primary selective arrangement (Fig. 11.26) may he employed, i n place of the loop feeders ivith interrupter switches for scctionalization. The primary s e l w t i v c tcm, ivhen using interrupt,er switches, has one hazard, as pointed out earlier: hut that, hazard is far less than the hazard of tem lvith int,errupter switches. \!'hen au outage occurs in the primary selectirr arrangcmcnt, the operat,or knows on which feeder it is, and orily one-half the load is dropped. I t is desirable t o drop both fceiiers to make the transfer safely. If the fault is riot in a transformer, all service

664

LOAD-CENTER VOWER SYSTEMS AND CIRCUIT ARRANGEMENTS

is restored after one operation-which is much quicker than is probable hy thr loop primary feeder load-center system with interrupter switches. There is only one s\vitching seqnence for the primary selective arrangement; hence t,he chanres are less for making a serious operating error with the interrupter switches. However, when specifying the unit suhstations for the primary sclert,ive load-renter system for these applications, it is most importaiit,, as pointed out previously, t o specify two interlocked single-t,hron. interrnpter sn.itrhes instead of one double-throw interrupter switch. This type of primary selective load-center system compares favorahly i n first, cost with loop primary feeder load-center system when t,he same types of int,rrrnpter switches are used in both systems. SELECTION A N D APPLICATION OF LOAD-CENTER UNIT SUBSTATIONS There are many factors t o cousider in selecting a load-center unit suhstation, two of whirh are very important to the system design engineer. These are (1) the kva rating of the substation and (2) the arraiigemcnt of the snhstation romponents. Aft,er selei:ting the substation, it is important, t o choose a good location for it. FACTORS AFFECTING SELECTION OF SUBSTATION KVA RATING

For the sake of this discussion it is assumed that the substation voltage ratings have heen determined as outlined in Chap. 10 and that t,he system circuit arrangemcnt has heen agreed upon. The kva size of the loadcent,er unit suhstat,ion is then largely a matter of economics in plants where several small substations are w e d , the desire being t o select the suhstation k v a rat,ing that \ d l effect the lowest over-all system cost. 1hree major system romponents affecting the over-all system cost are the primary cahle, unit substations, and secondary cable. These factors ~ v o r kconbrary t o one another; so t,hc most eronomical system as affected I)y substation kva rat,ing ran be obtained only hy considering all three at once. A s the number of substations increascs in a given area, the length of primaryfceder cahle required t o serve these substations increases. Conversely, as the nnmhcr of load-center unit substations in a given area increases, the amount of secondary-feeder cable rcqiiired decreases. The load-center unit substation cost per kva varies depending upon the size of the substation. If the= fact,ors are combined in their proper proportions for a typical indnstrial plant, the result is shown i n Fig. 10.1. This curve shows very definite minimiim system cost as a function of substation size for 208Y/

,.


665

120-volt, 240-volt, 480-volt. and GOO-volt systems. For 208Y/120 volts the most economical sizes are in the range of 300 to 750 kva, with the bottom of the curve being a t 300 and 500 kva. For 480-volt systems the most economical range is 500 to 1500 kva, with the bottom of the curve being at 750 kva. There are other factors which have an influence on substation rating, too. For example, the higher the primary voltage, the larger the kva rating unit substation should be so that a greater kva per primary feeder can be handled without unduly complicating t,he load-center unit substation overcnrreut protection problem. For example, when the primary voltage is 13.8 kv, it is desirable t o have a loading of 4000 to 7500 kva per primary feeder. With total primary feeder loadings of this magnitude using 500-kva units, individual primary overcurrent protection would be required for each substation. If 750-, 1000-, or 1500-kva units were selected, then primary-fender loadings of the order of magnitude just sperified could be obtained without individual primary overcurrent protection. Therefore, it would be expected that where primary voltages are of the order of 13.8 kv the range of substation sizes would be in the 750- to 1500-kva bracket. With lower primary voltages of the order of 2.4 t o 4.16 kv, the most economical primmyfeeder loadings would be of the order of 1000 to 3000 kva per feeder. Thus, for these primary voltages, subst,ation kva ratings could be smaller and not require individual primary overcurrent protection. I n any event,, substations larger than 1500 kva with 480 volts secondary are generally not used because of the over-all increase in system cost for the larger substations. Large spot loads can sometimes justify larger unit substations. For example, a single furnace, a single large oven, a single large welder may justify the use of substations of the order of 2000 to 2500 kva a t 480 volts secondary because there is no secondary distribution. Therefore, there is nothing t o be saved in that part of the system t o offset the lower cost per kva of the larger substation with a very few secondary circuit breakers attached t o it. Where these larger spot loads are encountered, it is well to consider using two smaller unit substations rather than one larger unit substation, particularly if the number of feeder circuit breakers for the secondary exceeds three or four. For general factory area load-center power supply, two 1000-kva loadcenter substations will nearly always provide a lower over-all system cost than one 2000-kva unit substation. Many times a load-center power system using two 750-kva load-center substations per load area will be less expensive over-all than a system using one 1500-kva unit substation per load area. The space available for substations sometimes dictates larger substation kva ratings. For example, one 1500-kva substation instead of two 750-

LOAD-CENTER P O W E R SYSTEMS A N D CIRCUIT ARRANGEMENTS

I

667

I FIG. 11.30

Typicol arrangements of load-center unit rubrtotianr.

INCOMING-LINE SECTION / >

I he primary incomiirg-line section may consist nf a metal-clad srvitchgear section with one or t n o readily removable rircuit hreakers as the circuit arrangement demands, Fig. 11.31A and 8. This gives the uliimate


668

in protection and safety for the incoming-line section. I n other cases the incoming-line sivitchgear may be remote, in which rase thcre would be a cable connection between the feeder circuit breaker and the load-center unit suhstation because of the desire to get the load-center substation closer to the load center. In this case only a pothead may he reyuircd, Fig. 11.31C. The pothead is often mounted on a junction box. I t is often desirable to place more than one small load-center unit substation on a primary feeder. I n that rase many operators feel that there should be a disconnecting means het,ween the feeder and the transformer t o permit isolation of the transformer for maintenance or other reasons without completely deenergizing the feeder. This, as pointed out previously, is done for the sake of economy at some slight sacrifice in safety. There are numerous types of interrupter switches and interrupter switch and fuse combinations. .Classes of Interrupter Switches. The t,erm interrupter switch is defined iu the Power Switching Equipment Standards of NEMA, Paragraph SG50-658, as follows: “An interrupter switch is a switch combining the functions of a disconnect.ing switch and a circuit interrupter for interrupting at rated voltage currents not exceeding the continuous current rating of the switch.” By t,his broad definition both the so-called “mag-i” and “load-break” switches are classified as interrupter switches, but in applying such switches the current-interrupting Capabilities of each class must be factored. There are three general classes of interrupter switches available for use

,a>

(8,

(C,

I01

IF1

FIG. 11.31 One-line diagram illurtroting application of the more commonly used loodcenter unit rubitotion incoming-line sections.


671

whioh alloivs the use of ilcse cutouts without being i ~ ~ s t a l l eirid a fireproof vault. For unit-sulist,ation applications t,hrre does not, x'rm i n lie any prartiml requirmwnt for makiiig i t i t interrupter su-itclr ai:trially operate under any more severe duty tliaii tire miigrietiziiig r i i r r w t of ilir trarisformcr. :is i i matter of fact, thiart: aro vcry good ~CRYOIIS for interloi'king svitche.; regardless of their ititrrmptiiig aliility. so tliat oidy triliisi'ormcr niagri(:tizing ci1rri:iit is ii~t(;rrupteil. I t is vrry simplit t o s:ifoly koy-itlrc~rlor~k the iiiterrupter sn-itvh ivith the main lon.-voltnge srcoii(lary circuit breaker so that, this circuit I>reaker must lie oprn (and t,hcrefore all load dropped) before the npcning or closing of tlir ititi>rrttptersnitrli. .is meiitioned earlier, tile u s e of thi. i i i i upiiLr sn-itch oii unit srtl~stationsis a t liest a cornpromisc, aiid r v c i i w h coiistriii.trd as \vi:Il as ~iossil)lt~. it docs riot h a w h i l t - i n iivrrlo:td or short-circuit protecti erefore, it is still the weakest, part of the unit srihstntioii. Thrse f a der-elop tiro hasic iiitf:rliii~kiiigintt~rriiptcrswitches: ety t o personuel arid (2) protectioli to service corrtiiluit,y. THE TRANSFORMER SECTION

Tht: trimsformer section of the liiail-r~~iiter iiiiit, sulistaiioii may ronsist of an oil, askarcl, opeir-vetitilntrd dry-typr, or a scviltd dry-type transformer. The h s k compnrisoils of thesr four t,rnnsformers are givim i n Talilc 11.1. For iiidoor iiistallatioiis thr rhoicc is iisua11y tirtivrtw tlic iiskarel a1111 the operr-veirtilatrd dry-typi: Iiiiits. 'l'hr tinsic considerations of sclr are ootliiicd i n Tatili. 11.2, a1111a pictorial comparison of tliese rrriits is given in Fig. 11.31.

..

I

' I

a

r

\FIG, 11.34

Typical bod-center unit substation tronrformer sections and arkarel-filled transformer, left; conventional dry-type transformer, center; and sealed dry-type tromformer, right.

672


TABLE 11.1

Comparison of Load-center Unit Substation Transformer Sections

I

Liquid filled ___

~

Type of Ironsformer

.......

Impulse strenglh.. Toto1 IDIS o t 75 C.. Insulation.. Temperature rating.: Average rise.. Hottest spot rise. , , , Audio sound level.. Weights. Dimensions: F I space.. ~ ~ ~ Height Normally available for .pplic.tion: Indoor or Outdoor. Submersible.. Fire and exploiion resistant. Maintenonce required: Liquid.. Intcmd cleaning, Extcrnol cleaning ond pointing expense.. Special precoutionr before energizing either initiolly or after shutdown.

....... ............. ........ ... ...... ............... ......... ...............

Oil

Atkars

Open ventiloted

Sealed Clor, B

100%

100% 100% Clots A

50%' 100% Clor, B

SO%*

50%

100% Ci.% B

100%

100% Class A

c

80 110

100%

65 C X db 125%

1007~ 100%

55 C 65 C X db

...

.....

:

c c

+ 10 db'l

120

c

140 C :+10db*

Cl0.S

* H

120 c * 140 C* k'+lOdb*I

80%

125%

125%

100% 100%

100%

120%

90%

110%

120% 110%

All

Indoor only

Ail

No

Ye,

Yes

Normal None

freque,

NOW

NOW

None

Frequent

NOW

None None

Normal

Normal

Subnormol

Minimvm

Minimum

None

None

Yes

None

None

..... Outdoor ......... Submersible

.............. ......

55

Sealed Clas H

Yes

t

All

IPlUlI

Yet (plurl

* S o t yrt covered by industry standards

t Applicahlc for all typrs of installation assuming no ~ r p o s u r et o lightning or assuming ad-quate protcetion ageinst impulse voltages can be provided.

SECONDARY SWITCHING SECTION

There are three hasic types of secondary switching equipment availahle for load-center unit substations. These are 1 . Drawout low-voltage switchgear with large-type air circuit breakers 2. Stationary-mounted molded-case circuit breakers for light-duty unit substations 3. Centralized motor control The drawout-t,ype switchgear and rentralieed motor coiit,rol are sometimes used in combination on the secondaries of the load-renter unit substation.


TABLE 11.2

673

Principal Points of Camparison* between Askarel and Open-ventilated Dry-type Transformers Type of tronsformer recommended

I Exposure lo lightning

Atmospheric conditions

Where transformers ore connected lo circuits exposed to lightning ond the usual protection i s provided Where study has determined mat the amount of lighming expos~re i s negligible. or the possible reivltont voltage stresses can be adequately taken care of by lightning protection Where atmospheric conditions are clean, such (IS in plants producing aircrah. instrumenh, precision parts and in certain types of machine &hops, assembly plants, food-processing plmts, and in &on. dry vovlb Where dirt conditions m e severe, such (IS in foundries, steel mill%, Row mills, cement mills, or other similar dusty or dirty lxations Where moisture conditions are rovcre,ruch (IS in geographic Iocalions of high humidity and where hansformers may be subjected lo p o r t i d or total submersion Where unih ore mbiected to acid, oil, or cormolive vopon Where units me wbjcded to oil or inflammable vapors and locw (See Article lion i s classifled (IS horardous or semihazardous. 500 of Notiond Electrical Code for Guidance) Where Ironsformen arc located overhead on platforms or in roof

Open dry

NO

Ye*

Yes

NO

NO NO NO

YU

IW*S**

Future applicatio

Where possible rearrongemont may ruse Ironsformers to be moved outdoors at D l d e r dote

NO

I

'Inaca of installa

ensed guide such as this, it is possible to cover only the more u81

.-

T I

LL"-"c,.-"

-L....I,l

".I

-""""":-"A

&L"& &I."-"

"-~

types

"..-"."I

1 " " . .

674


cuit breaker may be taken out and another one put in its place and the replaced circuit breaker taken to a clean dry location for maintenance under favorable conditions. This switchgear offers a high degree of safety hecause all parts are metal enclosed and interlocking is provided t,o prevent drawing out or inserting a properly adjusted large air circuit breaker with contacts in the closed position. The molded-case circuit breakers are limited in interrupting rating to 25,000 amp a t 480 volts and 40,000 amp a t 240 volts and are therefore limited as to the sizc of substations on which t,hey may bc employed. Thcse circuit breakers, while adequate \Tithin their interrupting rating, are not so sturdily constructed as the large circuit breakers and are not drawout in construction; therefore they lack the flexibility and ease of maintenance associated with the drawout switchgear. The moldedcase circuit breakers, however, being small in size, can be mounted as a complete unit on the end of the transformer section of the light-duty loadcenter unit substation. They are not employed on larger substations used in industrial areas as it is very desirable from the maintenance and operation standpoint t o have circuit breakers drawout, a feature not incorporat,ed in the molded-case circuit-breaker switchgear. In substations for commercial buildings where light-duty unit substations with molded-case circuit breakers find their widest application, the drawout feature, while highly desirable, may not be so essential as in an industrial plant where there are many more circuit changes and circuit-breaker operations than there are in commercial buildings. The third form of secondary switching is the centralized motor control as shown in Fig. 11.35. In many cases it is desirable to have a drawout switchgear section and a centralized motor-control section both in the same secondary line-up to provide feeders to other switching centers as well as to the centralized motor-control units associated with the unit substation. In such cases where there is more than one secondary feeder from the secondary of the load-center unit substation, a centralized motor control section should be connected to the switchgear bus section through a circuit breaker as shown in Fig. 11.36. The centralized control section on unit substations may be employed where it is desired to centralize motor control in one location such as might be the case in oil refineries where the substation and control are placed in a ventilated room, free from explosive or corrosive atmosphere. Always Use a Main Secondary Circuit Breaker. It is always preferable to include a main secondary circuit breaker in every secondary switchgear section regardless of t,he type of secondary switching equipment, used whether it is drawout switchgear, centralized motor control, or molded-case circuit breakers. The main secondary circuit breaker is

LOAD.CENTER POWER SYSTEMS AND CIRCUIT ARRANGEMENTS

FIG. 11.35

675

Typic01 centralized motor-control section for secondary of load-center unit

substotion.

requird by tlie Satioiial I~iectricalC'o:le whrre there is no indiaidrlal primary ov<'rcurrent protection of the unit siihstation transformer sertion. Where thore is irrtlividunl oimvurrent primary proteetion provided hp fuses, tire xsorrdary ?iri.nit breaker ofTcrs m1ir.h better protection to the secoiidnry INIS a t i d other equipmciii in the secondary sivit,rhg:ear than do the nsiial primary fiises w1iic.h hnvr t o he selected w i t h fairly high current renit hw:iker permits poasihle future expansion

FEEDERS

CENTRALIZED MOTOR CONTROL USE B R E A K E R HERE

FIG. 11.36 One-line diogrom of application of circuit breoker to feed centralized motorcontrol section to permit deenergiring of the control section for maintenance without deenergizing feeders lo lights and other load.

67 6


of a radial load-center syst,eIn into a secondary selective load-center system t o ohtain thc: advantages of the secondary selective system. Main secotidary circuit breakers permit cascading of large drawout air circuit breakers t o obt,ain lower cost of secondary snitchgear equipment where the slightly lower service reliahility offered by cascading is not ohjectionablc. The main secondary circuit hreakers offer better backup prot,ect,ion t o the feeder circuit hreakers, should a feeder cirnuit breaker fail t o trip on a serondary feeder fault remote from the suhstation. The main secondary circuit, breaker permits complete deenergization of the Ilus quickly i n case of an emergency, and it, greatly simplifies the interlocking prohlem where primary interrupter discounectirig switches with or without, fuses are employed. Where there is a power circuit breaker on the primary of each loadcenter uuit substation, it will provide most of the features, in so far as overcurrent protection is concerned, that t,he main secondary circuit hteaker will provide. Even t,hough there is an individual primary circuit breaker, if the maitt secnndary circuit breaker is omitted, it, nil1 not he easy to i:onvert, a radial load-center system t,o a secondary selective loadcenter system in the future. The primary rircuit breaker mdl not serve as a backup I)rcdwr for cascade operation of the secondary feeder circuit hreakers \!-here t.his t,ype of owration is desired.

FIG. 11.37

Double-ended load-center unit rubrtotion installed on the factory floor.


FIG. 11.38

677

Single-ended lood-center unit substation instolled on grating in roof trusser.

SELECTION OF SUBSTATION LOCATION

To get the most orit of the I o i ~ d - ( ~ r i ipo\vi.rr t ~ r distrihiition load-center uiiit sntistation should lie at the ceiiter of the load a sic pon.er may l i e coilsidered I i h r a n material, substntioiis a r c more akin t o a mitchiiio tool t h u i tlic’y are t o a service: tlwefore, they s1ioi:ld be givcii the same consideriitioii ii-heti laying out, the factory area arid spa(’eprovided oil the floor if it is iiecrssary t o iiist,all t,he load-~,eiitcruiiit substation on t,he factory floor. If they arc installed on t,lie favtory lloor. they take n p less than one-tenth of 1 per cerit of t h e tot,al floor area. There are many other locatioiis of load-center unit snbstations nhirh do not owiipy floor s p x e and yet permit the load-center nnit substatioii t o he close t o the miller of the load area. Some of these are iiloiig aisles, over Tvashrooms, on balconies, 011 the roof, iii the roof trusses, jrist outside the buildings, ctc. h pict,orial presentat,ion of some of tliesc locations is given in Figs. 11.37 t,o L1.10.

678


FIG. 11.39

FIG. 11.40 D-C

Lood-center unit r u b s t d o n instolled in penthoure on roof

load-centei unit substations instolled on plotform inounted on colurnnr

a b o v e factory floor.


679

LOAD-CENTER POWER DISTRIBUTION SYSTEMS FOR LOADS SERVED AT ABOVE 600 VOLTS Some loads, particularly large motors above about 200 hp, can often be served more economically a t voltages above 600 volts. Where the plant primary voltage is suitable, these large loads may be served directly a t primary voltage. For example, motors over about 200 hp may be served directly from 2400-volt primary systems, motors over ahout 250 hp from 4160-volt systems, motors over about 1000 hp from 6900-volt systems, aud motors over about 2000 hp from 13.8-kv (11.0 to 13.8 range) systems. In cases where oue or two motors or severa1 scattered motors, whieh would lend themselves to be supplied by a separate medium-voltage substation, ine toa large for the low-voltage system and yet too small for the primary voltage systcm, they may be served from the high-voltage primary feederp through a transformer for each motor. For example, an 800-hp motor ir a factory whose primary voltage may he 13.8 kv eould be served from th(

A LOAD CENTER r - q n

b

SUBSTATION

INDIVIDUAL TRANSFORMER FOR ONE LARGE MOTOR

I

6"

SMALLER MOTORS

FIG. 11.41

One-line dicigram rhowing large motorr fed through individual tronrformen and rmaller motor$fed from load-center unit rubrtationr.

680


13.8-kv system through a step-down transformer, Fig. 11.41. The most economical motor voltage would be 2300 volts in this case. I n many plants there are concentrated groups of large motors or other loads such as arc furnaces which may be served directly from the primary power system, and the primary voltage motor or are furnace rating is compatible with that primary voltage. Such may be the rase in plants with 4160 volts primary voltage or in mills with 6.9 or 13.8 kv primary voltage, Fig. 11.42. However, many mills have 13.8 kv primary voltage, which is too high to apply directly to the average large motor in the mill. Other mills may have 34.5 or 69 kv primary voltage, which is too high to apply directly to any motors. Some are-furnace transformers may he fed a t 34.5 kv. Where the primary voltage is too high for the large motors and where there are large groups of these motors, the load-center principle may be used to supply the large loads or motors with the same advantages as the load-center system supplying loads a t 600 volts or less. The master load-center substations, Fig. 11.43, for the larger loads merely step down to some voltage in the 2.4- to 13.8-kv range. A typieal case may he a paper mil1 with 13.8 kv primary voltage, Fig. 11.44. The larger motors

LOAD CENTER SYSTEM '

MOTOR A 1 PRIMARY VOLTAGE

L &

FIG. 11.42 Typicol one-lhe diagrrim showing Iarger motom operated directly a t primary voltage and maller motorr fed from load-center unit rubrtotionr.

LOAO-CENTER POWER SYSTEMS AND CIRCUIT ARRANGEMENTS

68I

FIG. 11.43 Morter load-center vnit wbrtotion rtepping power down from 13.8 to 2.4 kv for wpplying rl bonk of lorge 2300-volt motorr.

13.8 K V SOURCE

d~,

MASTER LOAD CENTER UNIT SUBSTATION

* fBBBf

?j

KVA LOAD CENTER UNIT WBSTATION

J.

Y

2.4 K V

66A6A MOTORS

RATEO 350

FIG. 11.44

SMALLER MOTORS

Typical one-line diogram of a mil1 rhowing groups of larger motorr fed from a m a t e r load-center uni1 rubrtaiion ond rmoller motori fed from a lood-center unit rubrtotion.


682

are supplied a t 2.4 kv through 13.8-2.4-kv substations and the smaller motors a t 480 volts through 13,800-480-volt suhstatioos. Another case would be a steel mill where the primary voltage is 69 kv (see Fig. 11.45). The power is transmitted to load centers a t 69 kv and there stepped down to 6.9 or 13.8 kv. I n each area larger motors operate at 6.9 or 13.8 kv and intermediate motors from 2.4 kv through 13.8-2.4 (or 6.9-2.4) kv master unit substations and smaller motors a t 480 volts through 13,800480 (or 6900-480) volt load-center substations.

1

I

13.8 KV BUS

1500 HP MOTOR RATED 13.2 KV MOTORS UP TO 250 HP

Y

,d

2,4,K:

MOTORS 2 5 0 TO 2000 H P

I

I I I <;n;> SUBSTATION

FIG. 11.45

, I,

;3.; KV BUS

Typical one-line diagram of a large mill showing large load orem fed by load-center ryrtemr through packaged substations. The large motors are supplied directly ot primary voltage, i.e., 13.8-kv ond medium-sized motors from o 2.4-lrv master unit substation ond smaller motors from a 480-volt load-center unit substation.


603

The fundamental approach is to use a relatively high voltage for transmitting of power to the load centers where it is stepped down to utilization voltage. The primary voltage varies as the function of total plant load and the secondary voltage as a function of motor size. See Chap. 13 for more details on primary distribution systems.

Chapter 12

by Shelby C. Cooke, Jr.

Secondary Distribution Systems The secondary distribution system discussed in this chapter is that part 6f the system beginning a t the load terminals of the secondary feeder breakers and extending all the way t o the utilization equipment. This secondary system is common to all sizes of plants that use voltages of 600 volts and below. For example, the small plant that buys power direct from the utility a t this voltage level would have protective and switching equipment, a t the service entrance point, as shown between dashed lines D and E of Fig. 11.1. A small plant that buys primary power (higher voltage power) mould have several load-center substations as shown between dashed lines C and E . Larger plants would have a primary system and load-center units as shown between dashed lines A and E . In the larger systems there may be various types of interconnections in the primary circuits and between the load-center substations resulting in different circuit arrangements as discussed in Chaps. 11 and 13, but each substation would then have a secondary distribution system beyond the low-voltage feeder breaker terminals. Therefore, the discussion in this chapter is applicable to any size industrial plant from the very smallest to the very largest. The secondary distribution area of the over-all system is the area to which the codes of various organizations principally apply. Therefore, i t is essential that, in designing this portion of the system, particular attention be given to the applicable codes. Mention will be made in this chapter of the National Electrical Code where it imposes limitations on the design of the secondary aystem. The discussions in this chapter assnime that the low-voltage feeder breakers have been properly selected to provide adequate secondary feeder protection, both from the standpoint of overload protection and short-circuit protection. 684

SECONDARY DlSTRlBUTlON SYSTEMS

685

FACTORS TO CONSIDER

The design of a secondary distribution system requires proper consideration of several factors, the more important of which are as follows: Short-circuit protection Overload protection Voltage drop Safety SHORT-CIRCUIT PROTECTION

The need for adequate short-circuit protection for primary industrial power systems is widely recognized and readily accepted today. Consequently there are very few, if any, new industrial plants built which do not have adequate short-circuit protection in their primary systems. The recognition of the necessity of adequate short-circuit protection has spread during recent years from the primary system down to the loadcenter unit substation secondary feeder breakers, and i t is relatively uncommon to see new plants designed with inadequate short-circuit protection a t this point in the system. However, adequate short-circuit protection is often entirely neglected beyond the secondary feeder breaker terminals down through the branch-circuit part of the system. A review of available accident reports indicates that considerably safer electric power distribution systems could he made if the same degree of engineering attention was given to the short-circuit-protection problem in secondary distribution areas as is given to this problem in the primary and load-center distribution areas. Secondary distribution systems with complete adequate short-circuit protection are economically feasible when using modern short-circuit protective equipments and observing proper system design fundamentals. The most important factor in obtaining adequate short-circuit protection in the secondary distribution area is that of proper design. When the system is so designed, presently available shorkircuit protective devices are adequate. Reference is made to Chap. 3, which outlines the characteristics and ratings of the major devices used in secondary systems and also to Chaps. 1 and 2, which describe the methods of making shortcircuit-current calculations. Use the load-center System. Although this chapter is entitled and deals with secondary distribution, i t will sometimes be necessary to refer hack to the system ahead of the secondary system since the design of the primary system will have considerable effect on the characteristics of the area beyond the secondary feeder breakers. The use of the load-center system of distribution as discussed in Chap. 11 will result in reasonable

686

SECONDARY DISTRIBUTION SYSTEMS

values of short-circuit current on the secondary feeder bus so that standard types of equipment will be adequate and can he readily applied. Table 12.1 shows thc maximum kva size of transformer for the shortcircuit-rating levels of standard short-rircnit protective equipment used in secondary distribution systems. TABLE 12.1

Maximum Kva Size of Transformer for Standard Short-circuit-rating levels of Secondary Protective Equipment Tronsforrner size, Ina

Shoe-circuit ,1. " .1 omp 208 volts

480 volts

112.5 225

15.000

50 100 150

25,000

300

50,000

750

5,000 10.000

300 750 1500

One of the important factors in selecting short-circuit protective equipment for the secondary distribution system is that such equipment usually has t o he small in size. This is due t o the conditions that exist around machine tools and other places where this equipment is located and the fact that there are so many of these devices that it is essential from a space-saving standpoint t o have them as small as possible consistent with obtaining adequate performance. Reference to Chap. 3 will show that, in order t o have small compact short-circuit protective devices in the secondary system in the form of circuit breakers, the short-circuit level must he kept helow 25,000 amp. This means that on a 480-volt system, for example, the maximum size of transformer is limited t o 750 kva. Many devices used in the secondary system require that the short-circuit current he less than 15,000 amp, and in these cases the masimum size traiisformer at 480 volts is 300 kva. As pointed u u l later i n this chapter, however, a long run of cable may introduce sufficient impedance to permit the application of these breakers t o systems supplied by larger kva transformers. Select Short-circuit Protective Devices with Established Short-circuit Ratings. There are several t,ypes of short-circuit protertive devices com-

monly used in secondary distribution systems. They are (1) inagnetic circuit breakers in individnally mounted cases, (2) molded-r,ase circuit breakers, (3) fuscd safety switches, and (4) panel hoards of various types with either fused switches or molded-case circuit breakers. Typical examples of these devices are shown in Figs. 12.1 t o 12.4, respectively.

687

SECONDARY DlSTRlWTlON SYSTEMS

FIG. 12.1 Enclosed monuolly operoted large air circuit breaker for individual mounting.

FIG. 12.3

FIG. 12.2

Molded-care air circuit

breaker.

Modern-type fusible rofety witch


688

Proper attention should be given to the ratings of these devices to be sure that their application is correct for the system under consideration. Care must he exercised in the selection of cable, bus, and other parts of the secondary distribution system in so far as short-circuit capabilities are concerned as well as in the selection of the short-circuit protective device itself. The principles discussed in Chap. 3 should be followed in the application of these items to be sure that they can withstand the possible short-circuit current during the interval required for the short-circuit protective device to clear the circuit. The foregoing comments show that i t is possible to design a system so that available equipments can he readily applied within their short-circuit ratings. However, some occasions may arise in which the load-center system with its small transformers cannot be used, or in which the shortcircuit currents are not within the interrupting ratings of available equip ment. In such cases some other means must be used to reduce the shortcircuit current t o a lower value. The addition of impedance in some form or another to the system is necessary to reduce the amount of short-circuit current. Every circuit element has a certain amount of impedance and therefore has some effect toward limiting the short-circuit current. Long cables or bus bars can be particularly important components in this respect.

FIG. 12.4

Typical fusible-type diitribution p o d board.


689

For example, even though the available short-circuit current on a loadcenter substation bus is 25,000 amp, the feeder cable may have sufficient impedance t o reduce the short-circuit current t o within the ability of a 15,000-amp interrupting capacity device a t the end of the cable. This effect is particularly noticeable on the lower voltage circuits such as 208 and 240 volts. This characteristic was discussed in detail in Chap. 1. Application of Reactors in Low-voltage Systems. In cases where the circuit impedance is not high enough t o limit the fault current sufficiently, reactors can be added to limit the current to the desired value. A typical example of the use of such reactors would be between a load-center bus with 50,000 amp short-circuit current available and a grouped motorcontrol equipment with a 25,000-amp short-circuit-current rating. Reactors are generally considered as undesirable in low-voltage distribution systems because of certain inherent disadvantages, namely, (1) voltage drop, (2) power loss, (3) cost, and @) spsce.., Hourever, when properly applied, they can be very usefd tools in system design. Proper application cannot be overemphasized because a reactor improperly applied can result in excessive cost, as well as poor system operation. For example, assume a circuit with a 50,000-amp short-circuit level of fault current. For such a system, a 15,000- or 25,000-amp interruptingcapacity breaker, plus a reactor, is more expensive and requires more space than a 50,000-amp interrupting-capacity breaker. I n addition, such a reactor adds to the voltage drop of the system and increases the system losses. I n general, reactors should not be applied for the sole purpose of reducing the cost of switchgear equipment since such applications will prove t o he uneconomical in the final analysis. Care should he given t o the location of the reactor for best over-all operation. The location of the reactor in the system can have considerable effect on the principal disadvantages listed previously, that is, voltage drop and power loss. For any given number of ohms, the voltage drop varies directly with the current and the power loss varies directly with the square of the current. Therefore, the reactor should he placed in as low a current circuit as possible in order t o minimize the effects of voltage drop and power loss. For example, refer t o Fig. 12.5, which shows two possible arrangements of reactors to limit short-circuit current on feeder circuits. Figure 12.5A shows the reactor in the main circuit where it must carry the full output current of the source system. Figure 12.5B shows the reactors in the individual feeder circuits where they are subjected only t o onefourth of the source system current. I n this particular case the voltage drop would be one-fourth and the power loss would be one-sixteenth of those experienced with the reactor in the main circuit. For the example just studied, it was assumed that all the current flowing to the short circuit was supplied from the power source. I n the more


690

SOURCE

SOURCE

Y

Y

n

n

T11

I =I00PERCENT

T

FEEDERS TO LOADS

FEEOERS TO LOADS

REACTOR VOLTAGE DROP = I

REACTOR VOLTAGE DROP =

REACTOR POWER LOSS = I

REACTOR POWER LOSS =

iA1 REACTOR IN MAlN CIRCUIT

(E) REACTORS IN FEEDER CIRCUITS

FIG. 12.5

b4

t6

Alternote reocior locaiionr for limiting rhort-circuii ~urreni.

typical case, some of the cnrrent may be contributed from induetion or synchronous motors connected to the feeder circuit. This makes the feeder-reactor arrangement even more advantageous since a11 motor contributions pass through the additional reactors. There is one appiication where a reactor can be appiied effectively in a large circuit and that is where the total short-circuit current in the lowvoltage system without reactors is over 100,000 amp. There is no standard a-c switchgear availahle to handle short-circuit currents in excess of 100,000amp. Modern systems properly designed do not encounter such high levels of short-circiit current. However, where conditions do not permit conventional designs and something must be done t o reduce the short-circuit current, large reactors ran often he effeetively applied in a synchronizing bns system similar to those outlined in Chap. i 3 on primary distribution systems. The calciilation t o determine the amount of reactance required for any application should give dne consideration to the current contributed to the short circuit directly from the load apparatus, which is not affected by the reactor impedance. Figure 12.6 illustrates this point and shows a sample calculation to determine the total fault current on the utilization bus for the installation of a particular reactor. Similar applications sbould be made using calculating procedures such as outlined in this

691

SECONDARY DISTRIBUTION SYSiEMS

figure. In some cases the load is of a static nature and does not contribute to the short-cireuit current. In these cases the same procedure can be used, neglecting the load contribution. The effect of the reactor on the system voltage regulation should not he considered lightly. The effect of the voltage drop due to the reactor on the system voltage conditions is, of course, a function of the power factor of the load current (see Chap. 4). Therefore, in applying reactors, consideration should be given to the circuit power factor and the resulting effect of the steady-state voltage drop on the system voltage conditions. In addition to steady-state voltage drop, reactors can have a very serious effect ou the system voltage dip which occurs during the starting of motors. A typical motor will draw about six times normal current a t a power factor perhaps as low as 20 to 30 per cent. If this current is drawn through reactors in the system, the resulting voltage may be so low as to he intolerable. The low voltage might result in an ohjectionable flicker or outage of lights and/or the undesirable drop-out of contaetors. I t might even be so severe that motors requiring a high starting torque would fail to start, since the motor torque developed is proportional to the square of the voltage. For additional information on the effect of motor starting on voltage conditions, refer to Chap. 4. The discussion thus far ou short-circuit protection has been based SAMPLE CALCULATDN SYMMETRICAL FAULT IMPERES

A

AS:~~È?%?CAL AMPERES AVAILABLE

G

AT eus

=- loo'ooo i80,000 AMP

480

+

I

1.25 EOUIV 1MPEDANCE'- 480 X

q

- 0.00350HMS

I 80.0W

IMPEDANCE PLUS REACTOR '0.00354 0 . 0 2 ~ 5= 0.0320 OHMS SYMMETRICAL FAULT CURRENT THROUGH REACTOR =* x1 = e670 AMP 0.032

+

CONTRIBUTION FROM SOOAMP MOTOR LOAD [ASSUME 4 TIMES NORMAL)=4X600;2400AMP

REACTOR TOTAL SYMMETRICAL FAULT CURRENT 0.02850HMS i2 4 0 0 A M P f 8 6 7 0 AMP : I 1 0 7 0 A M P 600 AMP TOTAL ASYMMETRICAL FAULT CURRENT

FIG. 12.6 Sample calcvlation to determine effect of reactor on rhort-circuit current.

692


principally on major secondary feeders and branch circuits where magnetic or molded-case breakers or larger fused switches may be employed. Practically all panel boards for 208Y/120-volt or 120/240-volt circuits have maximum interrupting ratings of 5000 amp. Where large 120-volt systems are used, the application of panel boards within their interrupting rating for branch-circuit protection will require extremely careful system design, since it is impractical to attempt to reduce short-circuit currents from values in the order of 50,000 or 25,000 amp down to the 5000-amp level required. Modern industrial plant power systems solve the problem of adequate branch-circuit protection by limiting the size of transformers supplying the 120-volt system to a maximum of 25 kva. Such transformers are located as needed around the plant and connected into the 480-volt system. Based on the foregoing discussion it can he concluded that, if proper attention is given to the design of t,he over-all power system, adequate short-circuit protection can be provided for all parts of the secondary distribution system. OVERLOAD PROTECTION

The provision of proper overload protection for secondary distribution circuits is very important. These circuits are widespread throughout the plant and generally are accessible to operating personnel. For this reason, additional loads are often connected by personnel not familiar with the charact.eristics of electrical systems. Such additions can easily cause the overloading of circuits, which in turn results in overheating of circuit elements and possibly can create fire hazards. Therefore, close overload protection should be provided for these circuits. The National Electrical Code specifically defines the degree of overload protection required for various types of branch circuits. For example, one section states that the fuse protecting a conductor should have a rating equal to that of the conductor. However, where standard fuse ratings and conductor ratings do not correspond, the next larger fuse may be used, but in no case shall the fuse exceed 150 per cent of the allowable current-carrying capacity of the conductor. Other sections of the National Electrical Code outline the overload protection required for motor branch circuits and state allowable limits for the overcurrent protective device ratings. Similarly, overload protection for transformers is also outlined in the Code. The foregoing is not intended to cover all the sections of the Code pertaining to overcurrent protection, but to point out some of the limitations imposed by these standards. Even in cases


693

where compliance with the NEC is not required, its requirements ran be used as a good hasis for providing adequate overload protection. VOLTAGE DROP

The amount of voltage drop which can be permit,t,edin any dist,rihution system depends a great deal on the type of equipment being supplied and the voltage variations which that equipment can tolerate without rnaterially affecting its operation. Chapter 4 discusses indetail the permissible voltage spreads for typical power systems and the effect of other than rated voltage on various types of utiliaation equipment. Before defining the permissible voltage drop in the secondary distribution system, it is necessary t o know what the voltage spread of the primary and the loadcenter distribution systems are. The amount of voltage drop which can be allowed in the secondary distribution system can he higher if the primary system spread is lower and must he limited if the primary system has excessive spread. Typical figures, however, might be as folloivs: Primary voltage variation could he in the order of 5 per-cent, load-center transformer voltage drops in the order of 335 per cent, secondary feeder drops approximately 156 per cent, and branch-feeder drops ahout 245 per cent. It is obvious that several of these factors could vary, giving the same total voltage spread. For example, if the pover factor were improved t o near unity, the transformer voltage drop would be less, permitting the other items t o have higher voltage drops. The foregoing values of IJ.5 per cent voltage drop for secondary feeders and 255 per cent voltage drop for branch circuits would easily keep voltage spreads within desirable limits in the average plant, assuming that the primary voltage spread is not more than about 5 per cent. SAFETY

Safety t o personnel in the secondary distribution system is perhaps the most important consideration of all, sirire this systen: is so extensive and so many operating people can come into contact with it. The safety of an industrial plant is principally the responsibility of the operating and maintenance personnel. However, the design engineer of the system also has a n important place in the safety picture. It is the design engineer’s responsibility t o provide types of equipment and system arrangements which reduce the necessity of working on energized circuits. This will promote safety in maintenance methods and also in normal operation. Modern equipment is designed with safety as one of the principal factors. It is common practice today t o locate all live parts in grounded

694

SECONDARY DlSTRlWTlON SYSTEMS

metal enclosures and t o make circuit elements such that they can be easily isolated or completely removed from the electrical system for maintenance operations. For example, low-voltage switchgear equipment is equipped with readily removable breakers which can be completely removed from the housing and taken to a remote location for maintenance and repair work. Alternate power paths should be provided by the design engineer for essential loads that cannot be deenergized so that necessary parts of the circuit can be deenergized for maintenance work. The most important part of safety in any system and particularly in the secondary distribution system is that of properly grounding the equipment. I n a secondary distribution system there are apt to be portable equipments, motors, and various other types of utilization equipment which are readily accessible to all personnel. It is extremely important that all theseutilization devices have their metallic frames solidly grounded to a good ground system in the plant in order to prevent any voltage elevation of the equipment under any conditions. Chapter 7 discusses in detail the methods of providing adequate equipment grounds, and therefore these methods need not be repeated here. Small portable tools, which are moved from place t o place periodically, and which often have no automatic means of grounding the tool frame when the tool is connected t o an energized circuit, are probably the most difficult t o keep properly grounded. Particular attention should be paid to this type of load t o assure proper equipment grounding. I n the secondary distribution system it is very common for machine tools to be connected to a plug-in type of bus run, and it is essential that proper grounding facilities be included to assure that the machine tool is grounded whenever it is energized. Chapter 7 discusses this particular point in some detail. However, it should be mentioned that both the design engineer and the operating and maintenance personnel should exercise extreme care to be sure that when plug-in devices and portable devices are used they are adequately grounded. THREE-PHASE FOUR-WIRE SYSTEMS

When three-phase four-mire systems are used, Consideration must he given t o the size of neutral conductor required t o meet all possible operating conditions. The 208Y/120-volt systems are always four-wire systems and therefore include the fourth conductor. With the growing use of higher voltage fluorescent lighting (see Chap. lo), there are more and more four-wire feeders on 480Y/277-volt systems. There are two principal factors in thc determination of the size of the fourth conductor. First, the fourth wire should be large enough to carry any current due t o unbalanced line-to-neutral loads; and second, it should be able to carry


695

any third (or multiple of the third) harmonic currents drawn by the load. I n a modern plant, practically all loads are either three phase or single phase connected line-to-line. Such loads do not cause any current to flow in the fourth wire. I n the case of line-to-neutral loads, attempts are generally made t o saitch the loads in balanced three-phase banks. When such switching is not possible, however, an adequate fourth wire must he provided t o carry any unbalanced neutral current. The use of fluorescent lights on either the 208Y/120-volt system or the 480Y/277-volt system results in third-harmonic currents, which must add u p and return to the source through the neutral conductor. A study of this problem indicates that the third-harmonic current in any one phase might amount t o as much as 30 to 35 per cent of the fundamental frequency current, resulting in a neutral harmonic current in the order of 90 to 100 per cent of the phase fundamental current. Therefore, all branch feeders and secondary feeders which supply only fluorescent lighting loads should be provided with a 100 per cent capacity neutral conductor. Practically, this means that any branch circuit with more than 50 per cent of its connected load being fluorescent lighting should have a neutral wire with 100 per cent capacity. For those circuits supplying only three-phase power loads, no neutral conductor need be included with the phase conductors since there mill not be any neutral current. As the load is viewed from points closer and closer to the source of power (load-center unit substation transformer), the line-to-neutral unbalanced load and the third-harmonic producing loads become smaller and smaller percentagewise. As a result, the capacity of the neutral conductor can be smaller in comparison to the phase conduct,ors. For example, consider a Y-connected .three-phase four-wire system. A feeder supplying only fluorescent lighting load must have a 100 per cent rated neutral conductor in order to carry the harmonic currents drawn. Another feeder might be a single-phase feeder connected from line-to-neutral, requiring a 100 per cent neutral conductor also. The remainder of the feeders supply balanced three-phase loads and do not require any neutral conductors. When this total load is viewed from the substation bus, the total neutral currents become a smaller percentage of the total phase currents, thus permitting a smaller neutral conductor, perhaps 25 to 50 per cent of full-load rating of the substation itself. This point should always be checked to ensure an adequate neutral conductor all thc way hack to the supply transformer. SERVING

PORTABLE T O O L S

There are two principal problems which arise in connection with serving portable tools. The first is to make sure of an adequate ground connec-

696


tion. This was mentioned earlier in this section and is discussed in detail in Chap. 7. The second problem in connection with portable equipment is that of providing adequate short-circuit protection and making the plngging and unplugging of such equipment safe. For small 110-volt circuits, 30 amp and below, the fuse in the branch cirSAFETY SWITCH cuit should provide adequate. shortAND FUSE OR circuit protection. The plugging of CIRCUIT BREAKER such circuits, without deenergizing the plug, is conventional practice. However, where larger machine tools are involved with their larger heavy-duty plugs, it is desirable from RECEPTACLE a safety standpoint to provide a disconnect means with individual short~ ~ ~ ~circuit A ~ protection f E for the particuLOAD lar circuit similar t o that shown in Fig. 12.7. This arrangement permits FIG. 12.7 Disconnect means with rhortinsertion of the plug while the circuit Circuit for heavy-duty plug. is deenergized. If the circuit were not deenergizcd and a fault were t o exist in the portable load ivhcn the plug was inserted, the fault would he removed in a minimum of time with the least chance of damage to the equipment if high-capacity currentlimiting silver-sand fuses werc used ahead. of the plug. For other types of fuses and circuit breakers, however, a OLUMN serious personnel hazard might exist. Therefore, the safer way of making surh 480 VOLT CIRCUIT a connection is t o deenergize the circuit SAFETY SWITCH hy opeuing the safety snitch before inAND FUSE OR serting the plug and reclosirig the safety CIRCUIT BREAKER switch after the portable load bas been . connected. 480-120 VOLT In a modern industrial plant withpracTRANSFORMER tically all loads at 480 volts and higher, one of the prohlems is to provide suffi-FUSE BLOCK cicnt 120-volt power around the plant for .,20 o,TLETS miscellaneous loads, such as portable tools, inspection lights, etc. Onemethod of providing this power is to place a small 480-120-volt transformer on each column

Qh

power.

shown in Hg. 12.8. 'The transformer


697

can he connected directly t o a 480- or 600-volt feeder circuit throllgh a safety switch and fuse comhination or a small circuit breaker. 1Io~vever, the device used should he checked t o he sure! of its short-circuit adequacy for the system t o which it is heing connected. I n most cases a trarisformer of $6 or I kva rat,ing will be sufficient t,o supply these miscellaneous loads. SECONDARY FEEDER CIRCUITS

13asicaIly, there are four t,ypes of secondary feeder circuit,s which should he mentioned. They arc (1) open wire, ( 2 ) mPta~-ericlosetibus, (3) CatJle in conduit, and (4) intcrloaked-armor cable. The open-wire type of feeder which has commonly been used in past years i s rarely used in modern plants. First, such an open-wire installatiou creates a safety hazard since the energized conductors are not enalosed and are siitijcct to contact by opcrat,ing personnel. I3rcause of the exposure of the open-wire installat,ion it,s relialrility i s c:oiisiderat>ly reduced as it is much more subject t,o failure because of forrign objects coming in contact with the conductors. Because of the lack of such inst,allatioris in present-day plants, no further reference will he made t o this type of circuit. The metal-enclosed bus is gaining wide favor in industrial plant installations because of its safety arid reliability as contrasted t o that offered by the open-wire system. There are two principal types of bus which are used. These are the plug-in type and the feeder type. The feeder h u s is often called a low-reactance hus hecause of the interleaved phasc-coriduct,or arrangement (see Fig. 12.9). This type of bus is gerierally used for major circuits carrying large currents to specific pieces of equipment. For example, this bus is often used t o feed large welding equipment. Also, i t is sometimes used to supply power to the plug-in type oE bus.

FIG. 12.9

Cross-sectional view of low-reactance feeder bur rhowing interleoved bur-bar

arrangement.

698


The plug-in bus coiisists of the necessary conductors enclosed in a suitable housing designed to permit the insertion of tap-off circuit breakers or fused-switch types of plug connectors. The tap-off points are located at close intervals along the bus run to provide a high degree of flexibility. This hits offers a safe reliable system for power distribution but still retains the advarit,age of the open-wire installation with respect t o the easr of tapping off for branch circuits. Figure 12.10 shows a typical plug-in hus iustallation. Cable in conduit i s wed widely for supplyiug all types of loads in the secoudary dist,ribution area. I n past years braided-type cables with either rubber or varnished-cambric insulation have been used for such inst,allations. I n recent years, however, the development of synthetic insulations has rrsulted in cables with improved characteristics. For example, synthetic-ruhher cahle with a neoprene protective jacket is commonly used today in conduit systems. Figure 12. I1 shows a typical cable of this type. The neoprene jacket provides a very tough and

FIG. 12.10 Typicol installation of plug-in-type bur with flexible bur drop cables.

700


FIG. 12.13

Interlocked-ormor power cable instolled in overhead rocks

of the sa,me ' s i x in steel conduit. In addit,ion the cable installation is extremely flexible, being out in the open for visual inspect,iori and easy repair, and can be relocated in case of system changes. A typical iiiterlocked-armor-cable installation in overhead racks is shown in Fig. 12.13. The prececliiig types of secondary-feeder cirmits will bc discussed in the next section with regard t o their application t o the design of the secondary feeder and branch circuits.

SECONDARY-FEEDER AND BRANCH-CIRCUIT DESIGN

The secondary-feeder and hranch-circuit design of a n industrial plant vary t o a cert,ain degree, depending upon the type of plant involved. Basically, there are two types of maehirie loads in industrial plants. There are machines wiiich are not pemanent,ly located and machiiies which are permariciitly located, Examples of plants with the first type of load are mctal-fabricating plants, automohile plant,s, etc., which consist principally of machine tools which are very suhjcct t o changes in location. The second type of load is found in a process type of plant in which all parts are more or less permanently installed. Typical examples


701

of this type of plant would be steel-mill motor rooms, paper mills, oil refineries, etc. There are many industrial plants which cannot be classified either as the metal-fabricating type or the process type but are sonic conihiiiation of the two. The power distribution syst,cm for any plant, consisting of such a combination can easily be designed if the fundamentals for t,he two extreme types are understood. Metal-fabricating Type of Plant. The plug-in h u m a y type of installation offers the most flexible system possible for distributing power in the areas where the machine tools are Iocated. As production models and designs change, these machine tools are often relocated, and it is essential that the power system supplying them be sufficiently flexible t o permit these changes with minimum disturbance. The plug-in busway is generally installed overhead, running completcly across the area at not more than 40- t o 50-ft intervals. This means that the maximum horizontal run t o reach a position over any machine tool will not exceed 25 ft. The power feed dropping down t o the machine tool from' the bus can he either by means of a flexible bus drop cable or by means of wire in rigid conduit. The flexible types of bus drop ofkr a material advaritage in flexibility in utilization-equipment movements. The most economical may t o supply power to the plug-in bus in most plaqts is hy means of interlocked-armor cahlc run overhead in racks or trays. This cable begins a t the secondary-feeder breaker at the loadcenter unit substation and terminates in the cable entrance box on the run of plug-in busway. Oftentimes cranes prohibit the use of an overhead bus system because the busway and the bus drops would he in the may of the crane hook. The flexibility of a plug-in bus installation can still he ohtaiiicd in these cases by installing the bus in a different manner. The bus can he run horizontally along each line of columns with the drop cables running down the columns and, if necessary, under the floor t o the machines. Another method would be t o run the cable feeder along the row of columns with short lengths of plug-in busnay mounted vertirally on certain columns, as shown in Fig. 12.14. When deciding on a plug-in bus arrangement, it should be rcmemhered that there is a very definite advantage to having the cable drop from overhead. I n this manner the cables are free from the oil and grease on the floor, and it is not necessary t o dig up the floor t o install or move any of the machine tools. Process Type of Plant. The other extreme in plant t,ype is that where the utilization equipment consists of machinery which will be permanent,ly located. For loads such as this, the type of syst,cmsmentioned under t,he metal-fabricating plant can also be used even though the machines never

702


R U G IN BUSWP;

FIG. 12.14

Suggested arrangement of plug-in burway where overhead crones ore

prerent.

need to be moved. The use of plug-in busway fed by interlocked-armor cable is a very economical method of supplying a number of relatively small loads from a load-center unit substation. In many cases, however, the loads are rather large spot loads and can best be served by an individual feeder. This feeder can either consist of cable or of a metalenclosed feeder type of busway. The selection of the type of feeder would be solely one of economics. In many plants there are spot loads of relatively large size, which require some voltage other than 480 volts. For example, there are many ovens rated 240 volts used in industrial plants. For such loads i t is suggested that they be connected to the 480-volt system through individual transformers stepping the voltage down to the 240-volt level required. Depending upon the size of these loads, they might either be connected into the overhead plug-in busway or perhaps would warrant their own individual feed from the load-center unit substation. It is common in a process type of plant to find areas with large numbers of motors permanently located. I n such cases it is common to install grouped motor-control equipment with a number of motor starters in a common housing. Such motor-control centers are generally supplied by means of interlocked-armor cable from their own load-center unit substation secondary feeder breakers. From the motor-control center to the

703


individual motors, cable circuits are commonly used either in conduit or in interlocked armor. It is apparent from the foregoing discussion that there are many variations possible to supply the common types of loads in the secondary distribution system. Figure 12.15 illustrates severa1 of the methods which have been discussed. It is not intended that what has been said should

I

a

L

1-r480y’277 3 \1

CAELE

INTERLOCKED ARMOR CABLE

PLUG-IN BUSWAY

4 CONDUCTOR INTERLOCKED ‘ARMOR

FOR OVEN

:&

LICHTING FEEDER

CONTROL

LARGE OVEN DR MACHINE TO WAD FIG. 12.15 Severai methodr of supplying powei to loodr in the recondory dirtribution ryrtem.

704


be a complete discussion of this suhject, but merely t o point out how modern metal-enclosed bus or interlocked-armor rahle can be used in various manners t o feed the loads safely and reliably in a modern industrial plant. Lighting Feeder Circuits. The foregoing discussion of secondary distribution feeder circuits has not included any mention of the method of supplying the lighting circuits. When high-voltage fluerescent lighting is used as discussed in Chap. 10, a four-conductor interlocked-armor feeder cable is commonly used to feed a numher of the small combination motor sbarters which control the lighting circuits. T h e lighting circuits are switched in balanced three-phase banks, which is more advantageous from a system standpoint. When small dry-t,ype transformers are used t o step down t o 120 volts for the lighting load, thrcc-conductor cable should he used t o feed several such transformers. The transformer secondary will then connect t o a lighting panel hoard. Panel hoards, which normally have a n interrupting rating of 5000 amp, are suitable for this service since the small step-down transformer sufficiently limits the shortcircuit current. The lighting load either can be on separate feeder circuits or can be taken from the power feeder circuits. Most engineers look more favorably upon having the lighting circuit completely separate so that it is not suhjert t o the outages and possible voltage fluctuations of the load circuits. Also, when using the fourth conductor in the plug-in busway as a ground conductor, it becomes impractical. t o snpply line-to-neutral lighting loads from the busway. Circuit Arrangements. The circuits associated with the secondary distribution system are practically all radial circuits. A radial circuit i.s one in which there is only one path through which the power can flow t o reach the utilization equipment. It is the simplest form of circuit and because of its simplicity is commonly used in these systems. Such an arrangement is normally adequate for this type of service and offers a sufficient degree of reliahility. Since there is a great deal of exposure t o oil, grease, dirt, mechanical damage, etc., in the jvorkirig areas, the secondary circuits undoubtedly experience more fault,st,han do other parts of the electrical distribution system. However, the amount of load affected by the outage of a secondary feeder is relatively small, and therefore additional paths for power t o flow t o the utilizat,ion equipment cannot normally he justified. Therefore, it is extremely rare t o find anything except radial circuits in the secondary distribution area. Although practically all circuits are radial, there occasionally arises a condit,ion where extra reliability is required because of the nature of certain process equipment. T o build reliability into the distrihution system for such loads, it is necessary t o consider not only the secondary distrihu-


705

tion system but the over-all poirer dist,ributioIi arrangement from the source of power down t,o the load. Chapters 11 and 13 discuss various circuit arrangements for the load-center system and the primary system in order to assure higher degrees of reliability for such essential processes. When an adequate primary and load-center system have been designed t o ensure addkional reliability, then steps can be t,aken in t,he secondary distribution system to carry this extra reliability down t o the utilization equipment.

Chapter 13

by Norman L. Hadley

Primary Distribution Systems The primary distribution system of an industrial plant is generally the higher voltage portion of the system, starting with the purchased-power service and including generators, switching equipment, circuits, and all transformers with secondary voltages higher than 600. Figure 11.1 is a simplified one-line diagram that illustrates the dividing line between the primary distribution and the other portions of the system described in Chaps. 11 and 12. The material in this chapter is presented under several headings relating t o different parts of the primary distribution system. While this is a convenient method of presentation, i t may fail to emphasize the importance of planning the parts together. Coordinated planning guided by over-all system characteristics is the only way desired objectives can be achieved. Such system characteristics as cost, safety, reliability, flexibility, and simplicity should be viewed together because they will be interrelated in varying ways. System arrangements tending to favor a particular desirable characteristic will most often tend to produce compromises in one or more other desirabl: characteristics. System cost is the characteristic that receives the most planning attention. I n comparing alternative layouts it is very helpful t o hear in mind that lower cost is by no means synonymous with better value. On the other hand, a determination of values relies heavily on judgment because it is very difficult to appraise recognized differences in such characteristics as safety, reliability, and flexibility. Service reliability is considered t o be improved when the arrangements are modified in ways that promise to reduce outage time during maintenance operations or in the event of trouble. The general attack is to provide more than one power channel around system components that need maintenance or might fail. Increased investment for such provisions may be money wasted unless the system is well planned in some 706

PRIMARY DISTRIBUTION SYSTEMS

707

other respects. The primary requirements for good service reliability are that good-quality adequate equipment will be selected, that it will be rarefully installed, and that it will be well maintained. Industrial systems should be planned with good Hexibility as a characteristic so that possible needed changes can be made readily. It is particularly important t o he able t o expand the power system, because industrial plants historically develop more and more load, even without Hoar-space expansion. There are numerous ways in which growth possibilities may be restricted. A common mistake has been the application of circuit breakers with little or no margin in interrupting rating above initial requirements. A good working rule is to anticipate the most logical future steps in a possible expansion program for the purpose of checking probable future interrupting duties. Conscious planning for flexibility will avoid much future waste and many unnecessary barriers to expansion, sometimes with little or no increase in initial investment. PURCHASED-POWER ARRANGEMENTS

Whether an industrial plant generates power or not, i t is typical to have a power connection t o a utility system. Certain characteristics and requirements of the utility service must be understood. It is therefore important t o investigate the purchased-power arrangement as early as possible because the power company must first study the request with relation to its own problems. A number of matters should be settled with the power company before even preliminary engineering of the primary distribution system can he undertaken. Whether there will be a consumer-owned main substation or not is an important matter. The service voltage that can be made available must be known, and there will sometimes be a choice. It is also desirable to know the maximum and minimum no-load supply voltages, or the voltage spread. Another characteristic always needed is the symmetrical rontribution from the utility system to a three-phase short circuit. The system engineer is better informed if he knows the actual present maximum and minimum duties for checking voltage conditions, as well as the anticipated maximum duty for selecting suitable switching equipment. Maximum and minimum line-to-ground short-circuit-current contributions should be known, but this information along with metering and line-protection requirements may not be essential for preliminary system planning. Particularly in plants with large purchased-power demands, hut also in some relatively small plants that need extra power-supply reliability, two power-company lines, and sometimes three, will be made available. It is not enough t o know simply that there will be two lines; the various

708

PRIMARY DlSTRlBUTlON SYSTEMS

possible differences in two-line characteristics can require radically different main-bus or substation arrangements. I t is necessary to be able t o classify the service as one of the following kinds: 1. The two lines can be used one a t a time only-in some cases as alternates, in other cases on a preferred-emergency basis. 2. The two lines can be used simultaneously and may be operated in parallel. 3. The two lines can be used simultaneously but must not be operated in parallel. 4. The two lines constitute a loop service with plant-owned normally closed sectionalizing switching to permit a net power transfer past the industrial substation. If there are two supply lines, the power-company contribution to substation short circuits should be inquired about with particular care. If the symmetrical contribution to a three-phase short cirnuit, is known for each line separately, nothing further is needed for nonparalleled lines. But if the lines are to be connected in parallel by the user’s equipment, then the total duty with the two or more lines simultaneously contributing to the same three-phase short circuit should be known in addition. From data in this form a suitable equivalent circuit can be derived for determining short-circuit duties a t various points and for checking voltage drops. Chaptcr 16 covers some further matters of mutual interest to the power company and the industrial plant, if it is desired to operate plant generators in parallel with the utility system. PRIMARY-SYSTEM BREAKERS

Three classes of circuit breakers are used in the primary distribution system. These are outdoor oil circuit breakers in outdoor stations, air circuit breakers in statioh-type cubicles, and oilless circuit breakers in metal-clad switchgear. Outdoor oil circuit breakers are typically used in outdoor stations where there are overhead lines operating above the 14.4-kv level. Figure 13.1 shows a 115-kv circuit breaker a t the right in an outdoor substation used by a large steel mill. Circuit breakers in this class have always become available as larger power systems develop a need for them. Listed circuit breakers include sizes up t o 330 kv with an interrupting rating of 25,000 mva. Metal-clad switchgear is the most widely used class in the primary distribution system. Figure 13.2 shows an indoor installation of metal-clad switchgear with one of the circuit-breaker units removed. This line of switchgear includes circuit breakers up to the 13.8-kv 2000-amp size with a maximum interrupting rating of 500 mva. Metal-clad switchgear is

PRlMARY DISTRIBUTION SYSiEMS

709

.. FIG. 13.1

Main substation rhowing 115-kv outdoor circuit breaker a t the right and power tronrformerr throot-connected lo 13.8-kv metol-clod witchsear in the center.

FIG, 13.2

Indoor metol-clod rwitchgeor with one circuit-breaker unit removed.

710


well suited to haiidle the main-hus arid distribution switching problems in the 2.4- t o 13.8-kv range for a11 hut a few unusually large plants. It is technically and economically sound t o employ statioii-type cubicle snitchgear, as illustrated by Fig. 13.3, in some parts of very large industrial systems. This s\ritchgear is available at volt,age ratings of 14.4 and 31.5 kv. The higher voltage level is seldom needed, hut the 14.4-kv eqnipineiit is risefill for main buses where the sources are large, especially if they are higher than 2000 amp, and the short-circuit duty cannot be limited to 500 niva. Even in such large systems, the major part of the primary-syst,em switchgear w i l l still be of the metal-clad construction out in the plant and on snhhuses. It, becomes t,oo costly to use station-type s\i-iti,hgear for the many smaller circuits, and the problem may be solved as shoii-ri in Fig. 10.25 of Chap. 10 hy the iise of ciirrerit-liinit,ing feeder reactors. As mentioired in Chap. 10, the most p~ofit,aiileiise of statioiitype sxvitchgrar is at the 1500-mva iiiterrupt,ing rating rat,her than 1000 mva a n d at 13.8 kv rather than 6.9 kv. At 14.4 kv, this class of smitchgear includes a 5000-amp k i w i t breaker wit,li a11 interrupting rating of 2.500 mva. The circuit-breaker equipment in m y primary distribution syst,em has an important ct'Eect, on over-all system cost as \re11 as on the performance. In particular, t,he availability and costs of circuit breakers help t o tfeterten1 voltage, the most practical sizes of soui the liest bns and feeder arrangements.

FiG.

i3.d

inrtoiiviion o i i4.4-iv rtotion-type cubicle switchgear.

PRIMARY DISTRIBUllON SYSTEMS

71 1

A cost sense for circuit-breaker application is a little hard to develop. because estimating data are not readily put in the familiar dollars-per-kva form used for many apparatus items. Circuit voltage and interrupting duty typically call for circuit breakers where only a small-but still widely variable-fraction of the continuous-current ability will be needed. For example, at 115 kv a single circuit breaker costs about the same as a 5000-kva transformer. However, the same circuit breaker with a current rating of 1200 amp is capable of handling a three-phase load of 239,000 kva. Some useful cost data on circuit breakers are included in Chaps. 10 and 17, and further reference to manufacturers’ schedules may be needed in making cost comparisons. TRANSFORMER CONNECTIONS*

The transformers used in main substations and a t other points in industrial primary distribution systems are typically three-phase twowinding units with delta-connected primary windings and Y-connected secondary windings. Two objectives are answered by the delta-Y connection. First, the Y connection provides a neutral connection point for system grounding purposes according t o the preferred practices covered in Chap. 6. The transformer secondary winding is a power source for the system it supplies, and the alternative of using a grounding transformer in connection with a delta source would be more expensive and less dependable. Second, the delta-connected primary winding is the simplest and surest way of stabilizing the Y-secondary neutral. Occasionally a Y-connected primary winding will be found desirable. For these a Y-Y two-winding transformer should be generally avoided in favor of adding a delta-connected tertiary winding. Among the reasons given for wanting a Y-connected primary winding is one based on a very common misunderstanding. When a four-wire service is available for supplying a substation transformer, an assumption is often made that it is compulsory t o use the neutral fourth wire, thus requiring a Y-connected transformer winding. This is not true; in fact, it is generally preferable not to use the fourth wire even though a Y-connected transformer primary winding is used for some other reason. A Y-connected autotransformer can be an unsatisfactory and dangerous system component. This connection should not be employed unless the possibilit,ies of trouble are thoroughly understood and will be guarded against in the application. Three-winding transformers may be used to interconnect parts of a * See Blume, Boyajian, Carnilli, Lennox, Minneei, and Mantsinger, “Transformer Engineering,” John Wilcy Q Sons, h e . , New York, 1951.

712

PRIMARY DlSTRlWTlON SYSTEMS

system a t two or three different voltages. These unit,s have the windings on a single core mounted in a single tank. They are t,echnically sound and will permit, a saving for an occasional applicat,ion, but they are not used so oft,en as might be expected. The reactances hetweeo each pair of windings and the loading relations are fixed by design, and t,here will consequently be less opportunity for later system modifications than if separate transformers are used. MAIN SUBSTATIONS

Xot all plants will o m and operate a main substation for supplying the primary dist,ribution system. A plant main bus serves the Same purpose if the purrhased-power voltage is suitable vit,hout transformation for the plant primary system. The principal functions of a main substation are indicated in Fig. 13.4A, which is a simple arrangement answering the requirements of a great many Emaller plant,s. More complivated substation arrangements result when there arc two or more incoming lines, two or more power transformers, or one of a number of other bus arrangements: Also in plants with power generation, the snbstation output may not supply a plant main bus but may be connected t o a synchronizing bus, as described later in this chapter. The substations in a few vary large plants with heavy loads in widely separat,ed areas may require transmission-voltage feeders connected t o the incoming-line bus, as mentioned in Chap. 10. Figure 13.4B differs from Fig. 13.4A in using power fuses instead of a circuit breaker in the incoming line. Circuit breakers are generally preferable, but fnses will be useful in satisfying over-all objectives in some of the smaller and simpler substations. When substat,iori primary fuses aTe used, it is better t o employ solid neutral grounding of the transformer secondary than t o limit the ground-fault current in the primary distribution system. The remaining substation examples all show two supply lines. I n these stations it will often be necessary t o accept some functional compromises in the high-voltage switching equipment for cost reasons. Inasmuch as smaller plants must sometimes be served from higher voltage systems, the main substation high-voltage circuit-breaker equipment can be disproportionately expensive among the other substation components. Stated in another way, a given high-voltage breaker arrangement for a given supply system will cost just about the same regardless of the substation size. The discussion is intended simply t o indicate what the several arrangements offer. Figure 13.4C shows a two-line single-transformer substation using two high-voltage circuit breakers. This arrangement might be used whether

713

PRIMARY DlSTRlWTlON SYSTEMS

the two lines are alternate, paralleled, or part of a loop. For a loop supply, the substitution of a circuit breaker for the transformer horn-gap switch would avoid opening thc loop hy the transformer protection scheme. T h e use of either two or three circuit breakers might he hard t o justify in particular cases. For alternate-line or preferred-emergmry

0 IBI I

I

(El

FIG. 13.4

Some typical main substation orrangementr used by industrial plants.

714


supply, the single circuit breaker of Fig. 13.40 with interlocked incomingline switches has a minor deficiency in not permitting an aut,omatictransfer between lines. Either Fig. 13.4C or Fig. 13.40 permits expansion by adding one or more transformers t,o the high-voltage bus. Figure 13.4E is simply an extension of Fig. 13.4C for a two-transformer substation where the two incoming lines are alternate, paralleled, or part of a loop. As illustrated with four high-voltage breakers, this substation arrangement can provide a n unusually high degree of service reliability except for a high-voltage bus fault. For the special case of two incoming lines that may be operated in parallel hut are not a loop supply, the arrangement of Fig. 13.4F is often a good solution. By omitting the high-voltage bus, and paralleling on the low-voltage side of the transformers, a saving in high-voltage breakers and structure is accomplished. The arrangement reduces the availability of the total transformer capacity because each unit has a transmission line in series. However, the station-cost reduction may be so significant for smaller substations that load curtailment during an outage becomes an acceptable risk. I t is moreover possible t o reinvest part of the circuitbreaker saving in additional size of transformer units to achieve service cont,innity for all t.he load or to reduce the amount of load curtailment during half-capacity operation. Referring again to Fig. 13.4E in connection with loop supplies only, the high-voltage part of the substation employs almost all the circuit breakers that can be fitted into a single-bus arrangement. However, a fifth circuit breaker could be added in the bus. With appropriate relaying, it would ensure continuity of service through one transformer under the condition of a high-voltage bus fault. I t is perhaps more profitable t o observe how reliability and flexibility are modified by removing circuit breakers one a t a time, as illustrated in Figs. 13.W t o 13.45. I n the three-breaker echeme of Fig. 13.4G the main functional compromise is that transformer protection requires opening the loop supply. A utility would not ordinarily consider this as a serious shortcoming, but it could Ee avoided in the alternative three-breaker scheme of Fig. 13.4H. Either of these arrangements provides service continuity through one transformer for any single fault, including a high-voltage hus fault. I n the two-breaker scheme of Fig. 13.41, operation of the protective relaying of either transformer not only opens the loop but drops the whole subst,ation load. The loop can be reclosed and plant service can be reestablished t,hrough the unfaulted transformer circuit by manual switching. A permanent high-voltage bus fault must, of course, be repaired before either circuit breaker can be reclosed. I n attempting to use a single breaker as shown in Fig. 13.4J, a problem

PRIMARY DISTRIBUTION SYSTEM

715

is encountered. Any high-voltage fault down to the transformers will be cleared by the single circuit breaker and the utility as a single-line short circuit, leaving uninterrupted plant service through one transformer, However, there will he a level of transformer fault current below which the utility cannot trip, and the faulty unit cannot be automatically disconnerted from the system a t such a level of overcurrent except by transferred tripping of a power-company circuit breaker using carrier or a pilot wire. BUS ARRANGEMENTS A bus is a junction of three or more incoming and outgoing circuits. The most common plant bus arrangement consists of one source or supply circuit and two or more feeder circuits. The numerous other arrangements and variations are mainly intended to improve the service reliability through the bus t o all or part of the load during expected maintenance, or in the event of equipment failure or source outage. Some very complicated bus arrangements have been used in trying to improve service reliability or continuity. Some of these arrangements are technically unsound and will not provide actual benefits. Other arrangements that do qualify from an engineering viewpoint are useful in meeting the rather typical requirements in the heavy industries that handle large amounts of power through main and subdistribution buses. These same bus arrangements will seldom prove acceptable for cost reasons in medium-size and small systems even when service continuity is considered to he unusually important. The highest quality of service reliability can often be obtained more economically for smaller plants, particularly for those with load-center systems, by over-all system arrangements that employ simpler and less costly bus arrangements. The double-bus arrangement shown in Fig. 13.5A is an example of the more complicated arrangements that is technically sound if good-quality equipment is used, but it is very costly for the usual sizes of feeder circuits. The arrangement is suitable for outdoor circuit breakers, station-type cubicles, or metal-clad construction. I n metal-clad equipments, some requirements can be met a t lower cost by employing two positions and one circuit breaker per circuit plus one spare removable circuit-breaker element, as illustrated for one of the several circuits. This variation still allows transferring any circuit or maintaining any circuit breaker without a feeder interruption. Figure 13.5.4 was intended t o indicate a preferred physical arrangement with companion circuit-breaker compartments in separate standard equipments facing each other across an operating aisle. A cable cannertion would usually join the circuit breakers. Occasionally a special

716


metal-clad equipment has been built with both buses running the lemgth of the equipment, placing companion circuit breakers in adjacent compartments. This is an inferior construrtiori because i t provides a possibility of involving both buses if there should he a serious switchgear failure. Most of the more complicated arrangements have in common the gen-

A

9

T 7 5 e

I

ICI

$Y

$Y

8Y

(Dl FIG. 13.5

Some typical bur arrangements used by industrial plants.


717

era1 charact,erist,icthat, individual lines can he connected to either of tivo buses (often without service interruption) with good maintertarice aci'ess t o most of the apparatus. Intermediat,e flexihility and reliability can be more economically obtairied for multiple-source bus arrangcment,s hy sectionalizing straight single buses. Figure 13.58 illustrates a typical two-source sectionalized-bus arrangement with a single cirriiit breaker per line. Figure 13.5R or some variation places lines and hreakers on the same basis of availability. Where metal-dad switchgear is used in the primary system, a feeder outage for circuit-breaker maiut,enance can he reduced to a matter of minutes with a spare removable circuit hreaker on hand. I n extending reliahility from a main bus t o a subhus in an important load area, parallel feeders may he used. In t,he load-center system as described in Chap. 11, each load-renter transformer has the same availahility as its primary feeder and supply breaker. Improvement in service reliability is secured by intercoiinect,ion at secondary voltage. When three or more sources are available at a main bus, Fig. 13.5f,"is a natural extension of Fig. 13.5R. However, Fig. 13.5D is more flexible and is usually preferred eveit when another circuit breaker is needed. This arrangement may be referred t o as a star bus, but it also is somet,imes called a synchronizing bus arrangement whether any of thc sources i s a generator or not. Particularly if reactors are needed t o parallel the sources, Fig. 13.5D will he preferable t o a straight bus (or a riug bus) with the current-limiting reactors installed hetween each tie circuit breaker and the common hus. The need for tie circuit breakers nil1 be obvious in some straight buses, but there will be other cases where the value may be in doubt. Experience shows they are too often omitted where a choire rail he made in t,he planning stage. The following remarks are intended t,o summarize thc various ways in which hus-tie circuit breakers may be useful initially and later. When two sources are used simultaneously hut must not he opcrat,ed in parallel, a normally open bus-tie circuit hreaker interlocked with the source rirruit breakers permits serving hot,h hus sections from one of the sources mheri the other is not available. Reasons for not paralleling t h e sources might he that they are not synchronized or h a r e a vokage phase difference. .4nother reason could be t o reduce the bus short-cirruit duty either initially or in t,he future if the duty might he increased beyond desired limits through additions t o the source capacity. For alternate (or preferred-emergency) or normally paralleled sources, a single straight bus may bc used. It is preferable to use a normally closed bus-tie circuit breaker so that one bus section can he kept availatile

718


when the other is out for maintenance or repair or t o permit additions during a plant expansion. For paralleled sources, relaying of the tie circuit breaker may be employed to split the system so that service continuity is retained on one bus if the other bus should feil or i t became necessary to back up a feeder circuit breaker on that bus. FEEDER ARRANGEMENTS

A feeder carries energy to a substation or bus or to several loads. The several feeder arrangements discussed below are illustrated together in Fig. 13.6. The primary feeders that supply load-center unit substations, as described in Chap. 11, are not included as further arrangements because they are fundamentally radial feeders. The main function of a tie feeder is t o connect two sources. It may connect two substation buses in parallel t o provide stiffness or service continuity for the load supplied from each bus. If either source has plant generation, then the tie feeder maintains the two parts of the system in synchronism and provides a circuit for transferring normal power and kilovars in either direction between the sources. A loop feeder also has its extremities connected to a source (usually a single source), but its main function is to supply two or more load points between. Each load point can be supplied from either direction; so it is possible to remove any section of the loop from service without causing an outage at any load point. A radial feeder connects between a source and a load point, and it may supply one or more additional load points between. If the connection to an intermediate load point is an “in-and-out’’ or “loop” Connection instead of a tap, the feeder does not, of course, assume loop characteristics inasmuch as each load point can be supplied from one direction only.

FIG. 13.6

Four primary-system feeder arrangements.

PRIMARY DISTRIBUTION SYSTWS

719

Radial feeders are the most widely used because they are simple, easy t o protect, and low in cost. They are simple because there is only one path for current to any given load point. They are easily protected by simple overcurrent relays a t the supply circuit breaker. The cost is low because there is no duplication of equipment. These comments apply to “single” radial feeders and not to “parallel” feeders which have characteristics resembling those of loop feeders. Parallel feeders consist of two or more feeders bused together both at the sending aud receiving ends. I t is sometimes more economical to design a heavy ctlhle circuit with two or more cables in parallel, but these cables cannot he considered as parallel feeders when single-circuit switching equipment is used. Parallel feeders as illustrated by Fig. 13.6 provide a high degree of service reliability, or continuity, if one of several methods of protecting parallel lines is employed. This protectiou is more complicated and expensive than the simple overcurrent protection ordinarily installed in single radial feeders. Separated circuits are desirable, and each will typically have the ability to handle the normal load with the other circuit out of service. Additional circuit breakers are needed; in general, four circuit breakers are required for parallel feeders supplying one load point, while only one circuit breaker is needed for a single radial feeder. There are obviously some heavy cost penalties against the parallel-feeder arrangement; so its excellent characterist,ias can he justified only for serving large loads such as suhdistribution buses or smaller loads with unusual service requirements. Several variations intended to improve upon the performance of single radial feeders have costs below those for parallel feeders. They all sacrifice service continuity as a characteristic. For particular requirements, one or another of these variations may be a preferred solution; hut t,hey teud in general to have questionable value. One rariatiou is t,o m e a conventional parallel-feeder scheme, but to use overcurrent protection only. This does not represent an important saving, arid a short, circuit in either line interrupts service. Moreover, the fault location is not indicated, and corisiderable time may be spent in finding which lioe is in trouble unless it is considered acceptable to take the chalice of closing again on t,he same fault. .I second variation is like the one above, except that one of the two circuits is operated normally open and is held as an alternate or reserve feeder. The equipment is the same, and the only advantage is that the fault location is indieated when service is interrupted. A third variation makes a further saving by using a total of four metalclad circuit-breaker positions but only two circuit breakers. The performance is similar to that of the second variation evcept that the outage

720


FIG. 13.7 Feeder-equipment compclriron for three alternative arrangements each rupplying the same three loads.

will he longer t o allow extra time t o move the circuit-breaker elements. Circuit-breaker maintenance rvit.hout service interruption requires that two spare circuit-breaker elemerits be available. Figure 13.7 is included in combination with Tahle 13.1 to lead into a discussion of loop feeders. Three load point,s are arbitrarily selected for compariug the number of cables and circuit breakers necessary for each of the three feeder arrangements illustrated. Tahle 13.1 then indicates how large the cost differencesmight he without considering any differences in protection costs. TABLE 13.1

Circuit Breakers and Typical Amounts of Cable Required for Supplying N Load Points by the Three Feeder Arrangements Shown in Fig. 13.7

......, . . . ......... .... ......, .....

Number of Circuit breakers.. Cable current rating.. . . Cable footage.. .

4N

2N

+2

Loop feeders perform much the same as parallel feeders, and the operating features can usually he ohtairied a t less cost for the two or more load points. But this is still a high-cost feeder system that will prove economically sound in about the same k i d s of situations where parallel feeders can be justified. Loop feeders look so attractive to many engineers that variations with lower costs are often considered and sometimes adopted. Most of these arrangements only resemble good loop systems and can he aualyzed into positions of lower value than radial arrangements of still lower costs.

72 1


Before esamiriirig any of these arrangements, there should be an understanding of the kind of equipment needed in a loop feeder t o produce the characteristics associated with it. h loop feeder is intended t,o provide service cont,inuity. It must therefore be operated normally closed and must have two sectiorializirig circuit, breakers at each load point. These breakers must be adequate for interrupting system short circuits, aud they must be a part of a complete protective system, including the source circuit breakers, so that any faulty section of t,he loop can be automatically removed from service mithout dropping any load. The preferred loop relaying is wire-pilot ditferent,ial. Directional overcurrent is slower but may be attempted t o reduce costs if there are only two or three load points. Figure 13.7 shows the switching equipment needed for the above performance in a loop feeder. One variation is shown in Fig. 13.8.4 where half the sectionalizing circuit breakers are omitted. Any feeder short circuit causes an outage at one load point, and an acceptable protection system is sometimes hard t o design, particularly for more than two or three load points v i t h any considerable distance between them. Figure 13.8B shows a further variation with no sectionalizing points. Scarcely any of the loop-feeder benefits remain. The service reliability is about the same as is provided by a single radial feeder, except that a feeder circuit breaker can be maintained without interrupting any load. Another variation of a good loop-feeder arrangement is the one illustrated by Fig. 13.8C. As a-ar true for Fig. 13.8B, a single fault drops all load. Service continuity for half the load can be obtained by operating the loop normally with a central sectionalizing switch open. Following

IEI

(A1 FIG. 13.8 Variations from the loop-feeder practice.

best

YY-Y (CI

722


a short circuit, service t o all the load can be reestablished by isolating the faulted section. The arrangement can serve the same complement of substations that can be served by a single radial feeder without needing individual protection of the substation transformers. Without further analysis this arrangement appears t o have attained fair service reliability with a moderate cost increase. However, most operating engineers will reject this system for cable circuits using metal-enclosed disconnects at the sectionalizing points because there are serious hazards involved in the switchingoperationslikely t o bc followed after a short circuit has occurred. Whether t,he loop was open or closed when the fault occurred, the problem remains t o find the section in trouble so that service can be restored to all loadpoints. While it is not the best procedure, a common routine involves energizing thc feeder from one end by repeated trial, after adding a section at a time, until the faulty section is located by immediate tripping of the source circuit breaker. These operations present repeated opportuuities t o make swit,ching mistakes. Even if the operator appreciates the hazard, there is still the temptation t o energize the sections one at a time by means of the manual disconnects. These switches are also hazardous to operate under normal conditions because every section of a closed loop usually carries current. A loop feeder carried on open overhead lines with structure or polemounted disconnects for sectionalizing is much safer. One difference is t h a t short circuits on overhead lines are more often temporary than permanent. h more importaut difference is that any switch failure is considerably less likely t o injure a n operator because of the remote operation. Several precautions should he observed in loop and parallel feeder lay-. outs. It is well t o avoid compound loops-those with more than two source circuit breakers or with more than one path from one of two source circuit breakers back t o the other. Loop systems such as these may develop in an attempt t o relieve overloaded circuit conductors. This is seldom an efficient solution compared t o starting a new single loop. The load division is difficult t o calculate, requiring close estimates of circuit impedance and assumed iixed loads at each load point. Normal load variations and load growth can cause unexpectedly wide changes in the load division, and compound loops therefore tend t o give only temporary relief at best. Other compound-loop systems simply grow without sound planning. They tend t o be associated with poor load division, poor protection, and compromised service. Particularly where disconnects are used instead of circuit breakers, operation is more complicated and therefore more hazardous. It was implied earlier that tie feeders do not supply load points between the two sources. This arrangement, combining the usual functions of


723

tie and loop feeders, is sometimes advantageous. At the same time, ccrtain complications and compromises of such a dual-purpose circuit should be recognized and weighed before finally deciding in favor of it. Listed beloiv are several possible shortcomings: 1. The metering of load transferred between sources is more complicated. 2. Additional protection at the load points may he needed to avoid opening the tie circuit on load-tap fault. 3. If synchronism is maintairied by t,he tie circuit, suitable precautious must he taken to prevent closing the tie by any sectionalizing circuit breaker because it will not he feasible to install synchronizing provisions at the several sectionalizing points. Another precaution applies equally to parallel and loop feeders. These arrangements may complete a closed circuit around a bus-tie circuit breaker or a current-limiting reactor when the source circuit breakers are connected t o different buses. Such connections are oftcu overlooked in system design work. I t is obviously a mistake to short-circuit currentlimiting reactors by a feeder arrangement, and it is at least questionable to have a bus-tie circuit breaker short-circuited by independent primary switching operations a t one or more remote points in the system. A review of the several reasons for using bus-tie circuit hreakers, as covered in another part of this chapter, will indicate how seriously the over-all system characteristics may be altered by au oversight of this kind. FEEDER INSTALLATION

Power lines are installed in a number of ways to suit a variety of local conditions. Even under similar conditions, practices vary according to the experience arid preferences of users. Outdoor Feeders. Outdoor circuit,s may he run overhead on poles with open-mire construction. This is a low-cost method, and it cau be credited with some further minor advantages for particular circumstances. However, it is being used less and less in industrial plants hecause of its disadvantages. The main objections to open-wire overhead construction are the hazard and outage possibilities from weather, pole breakage, and accidental contact with lines by mobile machinery such as cranes and shovels. There have also been many cases of direct contact by maintcnance personnel working on building roofs or at other places close to exposed lines. Open lines present further problems in contaminated atmospheres. For example, in some steel mills the dust problems may require overinsulatiou and insulator cleaning as often as once or twice a year to prevent flashovers. There are other types of plants where conducting dust or chemicals will similarly affect line insulation levels. Overhead lines on poles may use aerial cable, and this alternative

724


method is finding increasing use. Aerial cable employs insulated conductors compactly bound t o a mcssenger Ti-ith one or more circuits on the same poles. This construction cost,s more, but it offers better value for many installations by modifying the disadvantages of open-wire circuits. (1) The construction is compact and does not require the largc mechanical and electrical clearances of open wiring. The chance of arcidental contact is therefore reduced. (2) Thc hazard arid outage possibility from accidental contact are greatly reduced. (3) There is no insulator cleaning problem in adverse atmospheric conditions. I t is t,rue that insulation failure of aerial cable will cause a longer outage time, but this disadvantagc is usually an acceptable risk. A second altcrnative for outdoor circuits in industrial plants is underground cable. The excavation, duct vork, and perhaps manholes bring the cost of this method above t,hat of aerial cable. It is certainly desirable t o he able t,o eliminate overhead conductors and the supporting structures from congested industrial areas, and underground cable work has been the most popular between buildings and even t o outlying areas where the distances are relatively short. Direct burial is the less eupefisive (:onstruction even if it is slabbed for mechanical protection, and it is used t o a small extcih for single circuits where extended outage in the event of insulation failure can he tolerated. Outdoor underground construction provides somewhat less protection than might be assumed. One of the most common causes of failure is mechanical, from construction or maintenance work. It is also true that some soil condit,ions may at,tack a cable sheath and that water, oil, or chemicals in ducts or manholes may reach and deteriorate the insulation. Indoor Feeders. There is a definite trend away from underground conduit and duct work for the cable circuits inside buildings. Increasing preference is being given t o the use of interlocked-armor cable installed overhead for these reasons: 1 . The cost, is lower than for underfloor circuits. 2. Overhead circuits are accessible and can be more easily changed. 3. Their location will not interfere with the fnture installation of machine foundations. 4. Overhead installation protects the cable from possible damage by water, oil, or chemicals that may collect in underfloor conduits. I n some places it Tvill be desirable for appearance reasons t o conceal the primary-system wiring. Another objection t o the exposed circuits was t h a t they might he mistaken for secondary circuits or that, conduit might be mistaken for piping. Possible confusion of this kind is hardly a valid argument against a generally desirable installation method, and it seems t o have disappeared from the thinking of most plant engineers.


725

POWER GENERATING STATIONS

Chapter 16 should be consulted for general information about industrial power generation. The following material deals mairily with the generating-station bus arrangement and some problems that may be cncoui1tered in meeting the general requirement of integrat,ing the whole primary electric system, including any normally used tie with a utility. Generators are integral parts of turbine-generator sets; so generator details tend to be decided while dealing with the problems of prodnoing heat energy at a planning stage when no thought may have been given to the effect these characteristics can have on electrical problems. Coordinated planning will produce the best over-all solutions and may prevent unsatisfactory or costly compromises. Generator voltage rating and winding conriection are examples of characteristics that should be chosen entirely on the basis of what is best in the electric system. Generator kva rating is closely related t o the prime-mover size; so the selection of turbine-generator set size should recognize both heat and electrical objectives. The important effects of generator kva and voltage on the primary-syst,em smitchgcar mill be indicated in the following discussion. There mill he many eases vhere direct parallel connection of two or more sources on a generator hus will answer all the requirements for the initial inst,allation, but available switchgear will set a limit on the future size of the system. Even the initial station may be too large to permit direct parallel connection at generator voltage. The usual solution for industrial systems is a synchronizing bus, as illustrated in Fig. 13.9, which also s h o w the generally preferable method of connecting to purchased power through the synchronizing bus. Some curve data applying t o an elementary synchronizing bus pattern will help to make clear a few points t.hat are worth remembering. Figure 13.10 shows several identical generator-bus sections interconnected hy a synchronizing bus and identical reactors. The reactance values are

FIG. 13.9

Typicol generator synchronizing-bus arrangement with purchased-power tie.

716

P R M R Y DISTRIBUTION SYSTEMS

FIG. 13.10

Typical reactance values for

CI

generotor synchronizing-bus ponern.

reasonable and may be considered to be either those suitable for momentary- or interrupting-duty calculations. All reactanczs are per-unit values on the kva base of a single generator. Figure 13.11 shows how the generator-bus duty is limited by the synchronizing reactors. The total short-circuit duty includes a direct contribution of fifteen times normal from one generator and the motors supplied by one bus. The remaining cont,ribut,ion comes through one synchronizing reactor and varies with the numher of generators. The maximum contribution through a 10 per cent reactor is ten times normal for an infinite number of machines, makiug the total maximum load-bus duty equal to twenty-five times normal. The dashed line shows the direct variation of load-bus duty with machines directly paralleled and emphasizes how very effective 10 per cent synchronizing reactors are in limiting the duty. A further point of interest is the duty on the synchronizing bus. Even with reactance as high as 10 per cent, the synchronizing-bus duty cxceeds the load-bus duty for four or more machines. It is plain that the synchronizing bus may have to he braced for highcr short-circuit currents than thc load buses. It, is also apparent that the synchronizing bus should not be regarded as a load bus because the short-circuit duty may later exceed the ability of available or practical circuit-breaker equipment. The advisability of installing eit,her circuit breakers or disconnects t o isolate a faulty reactor from the synchronizing bus should also he questioned. The most important matter illustrated by Fig. 13.11 is that an early limit in size of system or numher of generators can be reached if the initial machine sizes are too large. Using the reactance values of Fig. 13.10, the load-bus duty for a single generator supplying an equal kva of motor load is fift,een times normal and the maximum generator rat,ing for 250mva switchgear would then be 250,000 divided by 15, or 16,667 kva. For an infinitely large system, the 250-mva ability mould be exceeded if the generator ratings were higher than 250,000 divided by 25, or 10,000 kva. In a similar way, the use of 5 per cent instead of 10 per cent synchronizing reactors would require limiting the machinc size to 250,000 divided by 35, or 7,140 kva, to he conserwhive in planning a large system using 250-mva


727

switchgear. I t is riot recommended that actual systems he planned using these assomed reartatices; the closest estimates that can be made for the particular apperet,us should be used instead. The proper valuc of synchronizing reactance is sometimes a perplexing question. The least reactance that will meet short-circuit objectives is desirable. The value of 10 per cent mentioned several times is actually

z

I-

6or

2 50 W

z c)

40

0 -

I

3

4

NO. O F GENERATORS ( O R LOAD BUSES)

__

FIG. 13.1 I Effect of synchronizing reoctori on short-circuit duties for the rynchroniringbur arrangement of Fig. 13.10.

728


fairly high, as \ \ i l l be shiiivti. First it should he remcmf,cred that the ohnrir value of r r w h t l r e rontrols short-circuit crtrretlt \\-bile the per rent reaibtwe 011 the amorlnt of load transferred is at1 iildex of voltage drop for that load. I t should also he rioted that synchronizing reactors are oftctl designed for handling only a f r a h o n of t,he generator kva and v i l l thcti oftrii be referred t o it1 per cent oil their own base. Figure 13.12 hclps to visiidizc the effect of varyitrg the amount, of synchroirizitig rrnrtatlw f o r the same generator and motor feartatlccs risrd beforc. .\I1 rra~~tatlres are again per-unit, values h s e d on the k v a rat,ing

FIG. 13.12

Decreoring effect of higher values of reactance in t h e synchronizing-bur

orrongement of Fig. 13.10.


729

of a single generator. It is evident that the first few per cent of synchronizing reactance are the most eRective in reducing short-circuit levels and that relatively little further henefit results from increasing the reactance from, say, 5 t o 10 per cent or higher. There are some minor objections t o larger reactors, such as higher costs, higher losses, and finally system instahilit,y. However, the most practical objection is the voltage drop accompanying load transfer, and this drop varies directly with the amount of reactance. For example, a 10 per cent reactor on a given base rauses twice the amount of voltage change caused by a 5 per cent reactor on the same base when transferring the same load. Figure 13.13 illustrates this point for three reactor values. Figure 13.13 illustrates another matter of interest. Reactors typically have relatively small values of resistance so that voltage drop a t a load power factor of unity is very low, but it increases sharply as the power factor departs from unity in the lagging direction. (The curves of Fig. 13.13 assume a reactor X / R of 50.) It is therefore important t o recognize probable operating power factor in contemplating the use of larger values of synchronizing reactance. Even where load power factor will be high, it should he remembered that low-power-factor demands such as those caused by starting motors may produce objectionably high temporary voltage drops. The curves of Fig. 13.13 show voltage drops from a point of constant voltage in the system. This point will he at active generator buses for

LOAD P F ( L A G G I N G ) FIG, 13.13

Voltage drop ot load bur when transferred through synchronizing reactors.

CI

load equal to generator rated h a is

730


actual systems and a t the synchronizing bus as well only if an infinite number of generators supply the synchronizing bus. Estimates of actual voltage drops therefore must use an “effective” value of synchronizing reactance. For example, if normal generator kva is obtained through a 10 per cent reactor from two active generators of a three-generator scheme, the total voltage drop includes the drop through one 10 per cent reactor to the synchronizing bus plus the drop through two more reactors effectively in parallel, and the total effective reactance is 15 per cent. Two bus arrangements with interconnection reactors produce operating results similar t o those from a synchronizing-bus scheme, but neither is preferred from an over-all characteristic standpoint. Both arrangements employ tie reactors between otherwise isolated generator buses, one in the farm of a ring bus and the other in the form of a straight bus with one less reactor. Both arrangements use more circuit breakers than a synchronizing-bus scheme and are inferior with respect to voltage drop for several machines. A minor advantage is that neither arrangement has a synchronizing bus where a failure could force isolated generator operation. A bus-tie reactor between two generator buses is basically the same as a synchronizing-bus scheme. If only two generators are installed initially, a single reactor should be used. At the time of installing the third machine, two additional reactors would be installed in the familiar synchronizing-bus configuration. Several other arrangements have been used in generating stations either t o limit short-circuit duty or to reduce the continuous-current requirement of generator switching equipment. Generator series reactors have sometimes been used to assist in controlling short-circuit duty in modernizing or expanding existing systems, but they can be regarded as generally undesirable. Duplex reactors offer a more efficient control of short-circuit level and also permit a reduction in continuous-current rating of the two breakers per generator. Double-winding generators produce similar results without needing external reactance. All these arrangements are somewhat specialized solutions. A very satisfactory arrangement is the unit system of directly transforming the generator output in a step-up transformer to eliminate all switching equipment a t generator voltage. An example of such an arrangement is shown in Fig. 10.26. This scheme has not often been used in industrial systems, which seldom have a transmission problem where the higher transformer output voltage is more suitable than the generator voltage level.

Chpter 14

by Donald S. Brereton

Power Systems for Commercial Buildinqs c

The subject of power systems for commercial buildings involves both the basic power-supply system consisting of the main switchboard or substations and principal secondary feeders and panel boards. I t also involves the many details of installation of branch circuits, out,lets, wall switches, lighting fixtures, and many other related types of equipment. This chapter will deal primarily with the discussion of the basic powersupply system and not with the details of installation of branch circuits and associated equipment. The range of commercial-building loads is from a few kva t o over 20,000-kva demand for a single building. Commercial buildings are considered to include office buildings, hospitals, schools, and practically any type of building except those occupied by manufacturing plants. They also include the officebuildings in industrial plants. Small buildings having less than a few kva demand and located outside the utility secondary network areas are almost invariably supplied secondary power by the utility, generally a t single phase, 120/240 volts. The power systems in these buildings involve service-entrance equipment and primary branch circuit equipment for their own use. Larger buildings, particularly those located in secondary network areas of the utilities, are generally supplied power a t three phase, 208Y/120 volts. The utilization equipment for larger buildings outside the network area in which the power component is relatively low and the main load is 120 volts may be supplied by 208Y/120-volt unit substations. A typical one-line diagram of a building supplied 208Y/123 volts by the utility is shown in Fig. 14.1, and a building supplied a t high voltage but having its own 208Y/120-volt substation may have a typical one-line diagram like that shown in Fig. 14.2. 731

732

POWER SYSTEMS FOR COMMERCIAL BUILDINGS

As buildings become larger, say 500 kva or more, there is an increasing tendency t o use higher voltages than 208Y/120 volts. These types of buildings have, in general, three kinds of distribution systems. The first is the building with considerable power load which may have 480 volts supplied either by the utility or by its r-,---UTILITY I 1 own local substations for the power II I either by the local utility or by its own II ?( 0: 0: I substations for the lighting load. A ? I typical one-line diagram of such a building when power is supplied by its own substation is shown in Fig. 14.3, and one when power is supplied by theutility network systems is shown in Fig. 14.4. I n other cases the entire building supply is 480 volts, either from a spot network of the utility or from the building's own substations. (This is discussed in detail later in this chapter FIG. 14.1 Typical one-line diagram of and is shown in Fig. 14.18.) The 480 CI commercial building supplied with volts is used directly in power-utiliza2 0 8 Y / 1 2 0 volts from the network system tion equipment and then transformed of the utility. from 480 to 120 volts for local floorutilization. Thistypeof distributionsysr----------1 tem is particularly applicable in old II -r -----_----_ UTILITY I I buildings where riser conduit sizes, etc., I $ I are such that increased capacity can be obtained economically only by using higher voltage risers, i.e., 480-volt instead of 208-volt risers through the building. It is also applicable t o larger buildings where thedistancesand loads are such that the savings in switchgear and circuit conductors are enough to more than offset the cost of the 480120-volt transformers for supplying the 120-volt load. Such a building might he a large department store where there is considerable incandescent lighting and where the outlet FIG. 14.2 Typical one-line diagram of load and showcase lighting load are commercial buildine. rupplied With .~ large medium voltage b y the utility.

____--__

L ,A ..;

,""V,

~

L * ~ A ,

I"

1

---


733

r----------1 UTILITY

I

-__

-

I I

FIG. 14.3

Typical one-line diagrom of o commercial building when power is supplied by its own substation.

FIG. 14.4 Typical one-line diagram of a commercial building when power i s supplied by the utility network.

734


Figure 14.5 illustrates a large modern commercial building receiving power from the utility a t 13.8 kv and utilizing 480 and 208 volts for lighting, 2400 volts for air-conditioning motor loads, and 480 volts for auxiliaries and elevators. The reason 208Y/120-volt services are shown for use in the lower floors of the building is to provide the services necessary for the large quantity of showcase lighting on these floors. It should be noted that additional 120-volt loads are required on other floors but that $UTILITY SERVICE

7TH FLOOR

FIG. 14.5 A typical one-line diagram of (I large commercial building receiving mediumvoltoge power from the utility and tranrforming to various utilization voltages.


735

it will prove more economical to sci'vc those by a dry-type t.ransformer stepping down from the 480-volt, system. The newest system is applicable to large buildings of several hundred kva or more demand xvhere there is not individual metering arid xhere t,he general-area lighting is by means of fluorescent lamps. The 480Yj277volt higher voltage combined power and lighting system is finding wider anti wider appliratiori lx!cause of its lower over-all cost for these t,ypes of installat,ioiis. I t is with this system that the rest of this chapter is particularly concerned. 480Yj277-VOLT COMBINED LIGHT AND POWER SYSTEMS FOR LARGE COMMERCIAL BUILDINGS This type of system is receiving increased reiognition because of the increased power load in commercial buildings and because the need for adequat,e lighting in commercial buildings has inarcased the light,ing load. Further, the trend has been almost entirely toward the use of fluorescent lighting for gcrieral-area light,irig, which can he readily adapted t o a higher volt,age supply. This is shoxn ii Pig. 14.6. Cntil the a d r c n t of the fluoresirnt, lamp, most distribution of p o w r for lighting loads was at 120 volts for t,he most e f f i c i e n t operation of incan-

FIG. 14.6

General-oreo fluorescent lighting

in a modern office building

736


descent lamps. Higher voltage incandescent lamps are fragile and generally not satisfactory. Modern lighting practices, however, include more arid more the use of fluorescent lamps either alone or in combination with other types. Since the fluorescent lamp has a ballast in series with it, the 120-volt limitation no longer applies. The required voltage at instant starting is always above 120 volts and is supplied by a transformer built into the ballast. Thus the use of fluoresccnt lamps has opened new possibilities for the use of higher voltage circuits to reduce the cost of the power system. Combined light and power systems have been used for a number of years in industrial p l a n k Combined light and power systems using 480Y/277 volts have been in service over 10 years (see references 2 and 3). The fluorescent-lamp ballasts are connected line-to-neutral and the motors connected line-to-line in the 4XOY/277-volt system. This system has resulted in a substantial reduction in cost of the power system for supplying these plants compared with separate power and lighting distribution systems. I n the industrial plants the control of these lamps has been accomplished by the use of fused combination motor starters to serve large blocks (up to 15 kva each) of fluorescent lamps, but this type of control is not suitable for office buildings. Examples of new large modern commercial huildiugs using the 480-volt system for both air-conditioning and lighting loads have been described in the tcchnical press (see references 10 t.o 16). Examples of large arid small huildirigs are given in A I E E papers (see references 4 to 9). The economies of a 4x0-volt system for supplying fluorescent ceiling lights and integral horsepower motors can be obtained in office buildings. This is brought about by the development of a small control relay which can be mounted in the higher voltage fluorescent lighting circuits for controlling the 277-volt distribution circuit in the lighting fixture. This relay is operated by coils whose nominal voltage rating is 24 volts. Since only 24 volts has to be brought t o the wall switches, this obviates the necessity for using 277 volts on standard 250-volt wall switches. Special 277-volt wall switches are available. HOW COMPONENTS OF THE HIGHER VOLTAGE SYSTEM ARE APPLIED

A typical building is shown in Fig. 14.7 in a cutaway view. The various components of the 480Y/277-volt system are shown. I n the basement, the unit substation (or main 480-volt switchboard in cases where secondary power is purchased) (1) steps the supply voltage down to the required 480Y/277 volts for distribution within the building. The busway risers (2) are used to distribute the 480Y/277-volt power to each floor in the building. Fusible plugs at each floor are used to connect from


737

the husway to the panel boards (3). The panel boards supply 277 volts line-to-neutral for the lights (4). The 24-volt remote-control switches (5) control the lights through relays (6) and a transformer (7). A drytype transformer (8) rated 480:120 volts, single phase, supplies the panel hoard (9) for the floor circuits (10). Elevator, fire-pump, and air-conditioning motors (11) operate on 480 volts, line-to-line. The application of these componcnts is also shown in the one-line diagram, Fig. 14.8.

ECONOMIC COMPARISON OF THE 480Y/277- AND THE 208Y/120-V0LT SYSTEMS Because of the 120-volt limitation imposed by incandescent lamps, the 208Y/120-volt three-phase four-wire system has heen commonly used in the past. Motors are supplied with the line-to-line voltage of 208 volts. The lighting load is distributed on the three phases, but connected from one line to neutral a t 120 volts. Similarly, the 480Y/277-volt system is a three-phase system with neutral readily availeb!e, 480 volts line-to-line and 277 volts line-to-neutral.


738

MOTOR FEEDER STbNDARD 3 CONDUCTOR CIRCUITS

~

''<

PANELBOARD BRANCH CIRCUIT

RELf

T480-1zov ^ _R ^^..r^

,...

Lm

12QV PANEL BOARD

I

120- 240 V -CONTROL TRANS

I FIG. 14.8

INTERLOCKED ARMOR CABLE BUSWAY CONDUIT a CABLE

-

LIGHTING FIXTURE

1

I

~

REMOTE CONTROL WITCH

I

REMOTE CGNTROL WIRE

One-line diagram of equipment for higher voltage lighting.

It is quite customary to supply motors and control a t 480 volts. This has been done in industrial plants for years to take advantage of the savings inherent with a higher voltage system. By extending the neutral circuit in the lighting feeders only, the fluorescent lighting load can be distributed between the threk phases and connected from one line to neutral a t 277 volts. The power load is fed from three-conductor three-phase circuits. It is not necessary to extend the neutral circuit in the feeders supplying only power load. A typical office building consisting of five floors and a basement is used as an example in comparing 208Y/120- vs. 480Y/277-volt, systems. Each floor, as shown in Fig. 14.9, has three wings each 75 by 160 ft divided into 24 areas, 20 by 25 ft. An illumination level of approximately 55 footcandles is assumed, representing a load of about 4 va per sq ft. In actual practice the load depends on the type of light source used and the utilization of the lighting equipment. With these figures as a basis, the total installed cost of the lighting distribution system for the two voltages can be compared. To get the over-all comparative costs, the necessary power distribution system including motors and control for air conditioning, elevators, and fire pumps must be included also.


739

The comparisons are made on the hasis of installed cost for specific equipments selected on the basis of sound system engineering. It should be noted, however, that loral codes, type of building construction, and other factors may alter either the engineering choice of equipment or the total installed cost. Therefore the figures shown below should be reconsidered in the light of local codes, maintenance problems, construction details, etc., for any specific building. One floor of a single wing (75 by 160 ft) is considered as a unit area. The entire building will include 15 such unit areas. The lighting load of 4 va per sq f t is equal to 48 kva per 75- by 160-ft area. I n addition, approximately 1 va per sq ft (12 kva per floor) is assumed a t 120 volts t o supply office machines, fans, etc. Higher 120-volt loads may result in office areas using many electric typewriters, calculators, tabulating machines, etc. Thus a total of 60 kva is required for each 75- by 160-ft unit area. The cost-comparison figures for the lighting equipment shown on the following pages are for a unit area. Total figures for the entire building will be fifteen times as great. The power load for such a building includes a large air-conditioning compressor motor (300 hp), several small air-conditioning motors, four 50-hp elevator motors, and a 100-hp fire-pump motor. With these assumptions as a basis, the two distribution systems, one operating a t 480Y/277 volts and one a t 208Y/120 volts, are compared as shown below. I t should he noted that it is difficult to divide the lighting circuits on the two systems so that comparable circuits will carry the same current. Therefore, multiplying factors have been applied to each system so that comparable current-carrying capacities are represented by each system.

75

z FIG. 14.9

75 I_' Floor plan of building-five

floors and basement.


740

480Y/277-VOLT SYSTEM

lighting. If every piece of equipment used in the lighting system were included, it would only confuse the main issue of distribution equipment. Lighting fixtures, ballasts, and lamps do not vary too much in cost as a function of distribution voltage. Fluorescent lamps of a given rating are the same regardless of the primary voltage of the ballast. Therefore, these items have heen omitted in this study. The most forrnidablc barrier t o the use of 277-volt distribution for lighting in officeshas been the 250-volt limit imposed on general-purpose switches for lighting circuits by the Underwriters’ Laboratories Standards for Snap Su,itches. The remote-control wiring system solves this problem by using small magnetic relays which, under UL Standards, are rated a t 277 volts, 10 amp. This system offers outstanding advantages for officebuildings since it is extremely flexible, convenient t o install, safe to operate, and makes possible a substantial saving in over-all system cost. Essentially this system includes a transformer, to supply control power a t 24 volts for use on the control sivitches, and relays and small switches for each circuit. Thus, only 24 volts is in reach of the building occggsnt,s. The remote-control equipment for each 75- by 160-ft unit area, including 48 relays, 48 switches, one transformer, and the necessary conduit and low-voltage wire, has an installed cost of $450. The total weight of copper in these components is approximately 16 Ib. There are several possible choices for branch-circuit wiring. Depending upon the construction of the building, the type of figure used, local building codes, etc., conduit 15 KVA -480/l20 VOLT

PANELBOARD FOR LIGHT!? \

D A S ~ DLINES INDICATE

120 VOLT FLOOR CIRCUITS 8 CIRCUITS - TOTAL OF I6Oo FT mou’T AND 3500 FT NO.12 WIRE

-

I

l -

LIGHTING CIRCUITS- I2 CIRCUITS TOTAL.1700 CONDUIT AND 3550 FT NO.12 WIRE

PANELBOARD FOR FLOOR CIRCUITS

FIG. 14.10 480Y/277-volt system-bronch for one unit area.

circuits for lights and 120-volt floor circuits


FIG, 14.1 1

741

Typical fused panel board designed for use with silver-rand cvrrent-limiting

fuses.

arid wire, caiilc, plug-in typp ~~rarri,li-(,iri,irit wiring or Q iloor may lie used. Each system has its nwri field of applitiiiion, and it, is iicyond tllc scope of this section to compare the various trms. 'I'hcrefort~this s t u d y l i u s included conduii and IT-ire for ihc braiirh circuits as shown iii Fig. 1-1.10. The total iristnllcd cost for eavh floor of each \x-ing is $11185, and it iiicloi1i.s 73 111 of copper. For specific in llaiioiis, the plug-in type of brarich circuit. may he more applicable for tho ficxiiiiliiy it iiif quite compnratjlc i n m s t , rspc lly if many outlets ar The l2O-volt branch wiring iern for offiw fails, oilice machines, etc., shown in Fig. 14.10 includes -kva transformcr to step ilie 480 volts down t o 120 volts and cnndiiit and wire to distriljutc this power t o floor outlets. The t,otaI installix1 cost of t,r:iiisrormer, nirc, aiid coiidiiit is $2111. The copper represeirted Iiy t,his system xeighs approximately 164 Ih. A single panel hoard (Fig. i . 1 1 ) is used t o supply the 277-volt circuits arid the 48O-volt, c;ircuit to the transformer for 120-volt power. 'Yo provide adequate short-cirvirit protection, each circuit of thi: panel l ~ o a r dransists of a discoirriect sn.it,r.h arid a silver-sand currcrit-Iimitiiig fuse. Twelve 277-volt circuits and one 480-voIt circuit arc required. To a l l o i ~ some spare circuits, an 18-circuit patic1 h a r d is d with :ui iiist'alled cost of 8'276 iricludirig fuses; ihis iiicliirles 15 11) of copper. The panel


742

board for the 120-volt floor circuits has eight circuits and costs $25 installed and has approximately 2 Ib of copper. The total cost of the two pane1 boards is $301 installed, and they include 17 Ib of copper. Power Source. The power for commercial buildings, hospitals, department stores, and other buildings of the type t o a h i c h this discussion is applicable may be served in a number of ways. I n nonnetwork areas, separate transformers may be used for each building. I n other arcas, 480Y/277-volt network power may be available. I n still others, only 208Y/120-volt network power is available. I n the basic cost comparison, the building has been assumed t o be supplied by a double-ended load-center uriit substation. This nill provide more accurate comparisons, ivhich is the objective to be attained here. Other power sources are discussed more fully later. For the building under consideration, a capacity of 750 kva a t each end of the unit substatiou is required, see Fig. 14.12. The installed cost of such a substation is approximately $36,000, and it includes approximately 6500 Ib of copper. SECONOARY FEEOERS

As in the case of the branch-circuit wiring there are severa1 possible choices for the feeder circuits t h a t carry the lighting load t o each floor iri each iving. Again depending upon such factors as local codes, type I

l13.8KV 750 KVA

750KvA

L

7” h 9

1200 AMP

Y 350 MCM IN IIT

FIRE PUMP 100 HP

200-

3:iOMCM

CAELE IN C(INDUIT

A

3 LIGHTING CIRCUITS BUSWAY SEE FIG.4

E LFVATOR

MOTORS 4-MHP

/

CENTER

c,,

300 HP

MOTORS

200F750 MCM CABLE IN CONDUIT

FIG. 14.12

480Y/277-volt ryrtem-unit

breokerr ore roted 50,000 omp.

wbrtation. (lhe tronrformer recondory and tie l h e feederr ore roted either 25,000 or 15,WO omp.)


PANEL BOARDS

400 AMP 4 WIRE 3 PHASE PLUG- IN BUS WAY TOTAL 404

100 AMP FUSIBLE PLUG

P

I I 5 T_T

FIG. 14.13

743

I

480Y/277-volt system-burway to each floor.

of building construction, etc., the logical choice might be conduit and cable, armored cable, or plug-in busway. Because of the flexibility offgred and the t,ype of building under consideration, plug-in busway has been selected for the secondary feeders in this study. One busway is used for each wing, as shown in Fig. 14.13. Fusible plug-in switches using current-limiting fuses and properly selected switches are used a t each floor to connect from the busway to the panel boards. Air circuit breakers with au interrupting rating of 25,000 amp cau be used in many installations aa long as their interrupting rating is not exceeded. On larger systems, with short-circuit duties above 25,000 amp, the high interruptingduty switch and silver-sand fuses should he used. The total installed cost of the busway system shown in Fig. 14.13 is $6200 for the entire building. The system includes approximately 2200 Ib of copper. On specific installations a conduit aud cable system may prove less expensive and more applicable. POWER SYSTEM FOR SUPPLYING MOTORS

The estimate for the air-conditioning equipment for this building includes one 300-hp motor and a group of smaller motors. The 300-hp synchronous motor, reduced voltage starter, and 200 ft of conduit and cable for this motor have an installed cost of about $11,550. The total weight of copper in these components is roughly 2525 Ib. The group of smaller air-conditioning motors are assumed to be broken down as follows: two 50 hp, three 25 hp, one 15 hp, and one 10 hp. These are all induction motors controlled from a motor-control center. The


744

total installed cost of motors, control, conduit, and cable (200 ft) is estimated at $5900. The estimated weight of copper in these components is 945 lh. The total air-conditioning equipment costs $17,450 installed and has 3470 Ib of copper. In addition to the :tir-ronditioning load, this building will include motors for the elevators and fire pumps. A total of four 50-hp elevator motors and one 100-hp fire-pump motor is assumed. Comhiiiation magnetic starters fdr these motors arid cable and conduit lengths of 200 f t aie used. The total estimated installed cost is $6470 for all this equipment with an estimated 1145 Ih of copper. 208Y/120-VOLT SYSTEM

Lighting. Since the area chosen for the study does not change, the remote-control system for the 208Y/120-volt system is the same as for the 480Y/277-volt system. Although the remote-control system is not essential on 120-volt circuits, it has many advantages over the conventional wiring system, and the additional cost is insignificant in the over-all cost. The branch-circuit miring, however, is different because the 20SY/lZOvolt system requires more circuits t o serve the same area. Again conduitand-wire type of branch-circuit distrihution is used, as shown in Fig. 14.14. Since more circuits are involved, the installed cost increases from $1985 to $2830, and the weight of copper increases from 73 to 102 Ib. The 120-volt circuits for office fans, business machines, and desk lights PANELBOARO FOR LIGHTS AND FLOOR CIRCUITS

75'

NOTE: 120 VOLT FLOOR CIRCUITS NOT SHOWN-SAME AS SHOWN IN FIG. 2

FIG. 14.14

LIGHTING CIRCUITS - 2 4 CIRCUITS TOTAL- 2420 FT CONDUIT A N D 5 0 0 0 L l NO.12WIRE

208Y/12O-voll system-branch circuits for lights in one unit area.


I

150H CABLE

I

13.8KV

4D0 MCM

IN

75OKVA

7 5 0 KVA

$3000 AMP

T

200

745

L L T

3000 AMP

v

4 'E)

fl

2-400 M C M CABLE IN CONDUIT ELEV MOTOR5 4-50HP

-

200 f l - 2 7 5 0 MCM CABLES IN

7 S M L L AlR W N D MOTOR5

CONDUIT

FIG. 14.15 208Y/120-volt ryrtem-unit rub9tation. (The trandormer recondary and lie breokerr ore roted 75,000 omp. The feederr ore rated 25,000 omp.)

are not shown in Fig. 14.14. Although the same number of circuits are used for this load as ivhen the 480Y/277-volt system is used, the cost decreases from $2111 t o $1875 and t,he rneighJ-of copmr fiamSi61 t o 72 lh since the 480-120-volt transformer is no longer required. A11 the circuits for the lights and for the floor plugs can he fed from.one panel hoard. There are 24 circuits for lights and eight floor circuits. A 42-circuit panel board with spare circuits is selected ivith an installed cost of 5330 and a total weight of copper of 10 lh. Power Source. Again for cost-comparison purposes, a double-erided unit substation is selected t o distribute power for all lighting and a11 motors. The substation shown in Fig. 14.15 has a n installed cost of 555,300 and includes 6700 lb of copper. This compares with $36,000 and 6500 lb of copper for the 480Y/277-volt system. SECONDARY FEEDERS

The feeders t o carry the lighting load t o each wing are shown in Fig. 14.16. Again, one busway is used for each wing with t a p boxes a t each floor t o carry the circuit t o each panel board. The panel board includes the necessary feeder circuit breaker. The total installed cost of this system for the entire biiiiding is $12,400 and includes roughly 5200 lh of copper, compared with $6200 and 2200 lb of copper for the 480Y/277-volt system.

746


POWER SYSTEM FOR SUPPLYING MOTORS

The 300-hp synchronous motor, reduced voltage starter, and 200 ft of cable and conduit have an installed cost of $15,500 and include 4200 lb of copper. The group of smaller motors together with their control center and cable have an installed cost of $9010 and include 2400 lb of copper, whereas with the 480Y/277-volt system the same equipment costs $6470 and includes 1145 lb of copper. TABLE 14.1

Comparative Installation Costs for Systems -4 and B 480Y/277-voR system

208Y/120-volt system

lnrtolled

Instolled

~

Equipment COIt

................... ....... ............. ...................... ................... 6. Buswoy risers...................... 7. Air-conditioning equipment. .......... 8. Elevator and flre-pump equipment.. ... Total.. ........................ Total per k v a . . ................. 1. Remote control. 2. Bronch-circuit wiring for lights. 3. Floor cirwits I120 volt). 4. Panel boordr 5. Unit substation.,

Weight of opper, Ib

cost

Noight of opper, Ib

-1

17,450 6,470

240 1.100 2,460 255 6.500 2,200 3.470 1,145

$6.750 42,450 28,125 4.950 55,300 12,400 23,370 9,010

$138,835 92.55

17,370 11.58

$182,355 121.57

$6,750 29.800 31,650 4,515 36,000

6,200

~~

240 1.530 1.080 I50 6,700 5,200 5,855 2,400 ~

23,155 15.44

-

Savings with 480-volt system: approximately $30 per kva, approximately 4 Ib per kva of copper.

POWER SYSTEMS FOR COMMERCIAL BUILDINGS COMPARATIVE COSTS

747

OF 480Y/277- VS. 208Y/12O-VOLT SYSTEMS

Table 14.1 shows the total installed cost for all items used in both systems. It should be noted that the figures given in the foregoing text for the lighting equipment are based on one unit area, that is, one floor of a single wing. Since there are three wings, each with five floors, the totals in Table 14.1 are fifteen times the figures given in the foregoing text. As shown in Table 14.1 there is a substantial saving both in money and in copper with the 480Y/277-volt system. As noted earlier, the fixtnres, lamps, and ballasts cost about the same for the two systems. A total saving of $30 per kva is shown in Table 14.1 for the 480Y/277-volt system compared with the 208Y/120-volt system. The items which show the greatest savings are the branch-circuit wiring, the unit substations, and the busway risers. The savings shown for the air-conditioning, fire-pump, and elevator eqnipments are also appreciable.

SAVINGS FROM SMALL OR LARGE BUILDINGS

Although the building chosen for this study is substantially large, similar savings can be shown for smaller buildings. For example, if a building is selected one-third the size (one wing) and it is assumed that the air-conditioning and other loads drop in proportion, a distinct saving can still be shown. Although the unit savings in dollars per kva would he less, they are still large enough t o be worthy of consideration. In general, the lighting portion (items 1 to 4, Table 14.1) will he less for the 480Y/277-volt system since the saving in branch circuits usually more than offsets the additional cost of the 120-volt floor circuits. And, unit substations even in the smallest rating of 225 kva cost less for a 480Y/277-volt system than for a 208Y/120-volt system. Since the cable or busway for the 480Y/277-volt power is smaller, there is an inherent saving with this system, regardless of size, as long as the fluorescent lamp load and power load combined are greater than the 120-volt load.

REASONABLY HIGHER PERCENTAGE O F 120-VOLT LOAD HAS NO SERIOUS EFFECT

Some buildings such as department stores still use a considerable number of incandescent lamps, which must he operated on 120 volts. Other offices use an abhormally large number of 120-volt office machines. However, even in such buildings, the 480Y/277-volt system can usually be shown t o be advantageous as long as the total 120-volt load is not greater than one-third to one-half of the total power load.


748

MODERNIZATION OF EXISTING BUILDINGS

The above study, of course, deals entirely with new construction and quite definitely shows the advantages of the 480Y/277-voIt system for new office buildings. The same type of cost study can he made for a modification or expansion of an existing building. I n planning an expansion of an old huilding, the 480Y/277-volt lighting system is worth consideling, for the existing cables will be capable of carrying a good deal more load a t 480 volts than they can a t 208 volts. The savings in cable can he appreciable, and a better system results. HOW TO APPLY 480/277-VOLT LIGHTING WHERE DIFFERENT POWER SERVICES ARE AVAILABLE A close examination of Table 14.1 shows that both the lighting system (items 1 to 4) and the power system (items 5 to 8) show a saving if the 480Y/277-volt system is used. Therefore, even if no substation is involved and power can be purchased a t either 480Y/277 volts or 280Y/120

-

I

I NTER RUPTING SWlTC H

7”

TRANSFORMER

FIG. 14.17 building.

ULU

T

Combined light and power system illustrating one unit substation inside a


749

NETWORK

FIG. 14.1 8

Two typical types of 48OY/277-volt spot network distribution ryrtemr.

volts, a saving in the lighting system alone can be shown by choosing t,he 480Y/277-volt system. Figures 14.17 t o 14.20 indicate four systems used t o supply power t,o office buildings. Figure 14.17 is the system used in this study, that is, with power supplied t o the building at a high voltage such as 13.8 kv. This would be typical of nonnetwork areas, and the transformers may be owned by the building ovner or the utility. Figure 14.18 shows the 480Yj277-volt network system for distribiitirig power t o buildings. With such a system, distinct sa\Tiiigs can still be shown for the 277-volt lighting system since a 120-volt lighting scheme would mean the addition of step-down transformers (at about $12 t o $15 per kva installed) for the entire lighting load. The savings represented by the 480Y/277-volt system in such a rase would be about $15 per kva. Figure 14.19 is a 208Y/120-volt network system v i t h autotransformers used within the building t o produce the 480Yj277 volts. Even with such a system, subst,antial savings can still be shown for the higher voltage system. Since autotransformers are relatively inexpensive (ahout, $5 t o $8 per kva inst,alled), the total savings shown by using such a 480Yj277volt system instead of the usual 208Yj120-volt system \r.ould he about $22 t o $25 per kva instead of the $30 per kva shown in Table 14.1. The possible savings of $30 per kva might be decreased by another system problem. If the utility supplying power t o the commercial building can provide only a 480-volt delta system, a means for establishing the system neutral must be provided. An example of how this problem


750 7 - -

I

I

UTILITY

---I-----

I

\AlAs

-1

( h 1 I. k ;

aJ-1,

W”Y\

TO STREET

I

INCOMING BREAKER

Bus TIE

4 B O Y / 2 7 7 VOLTS

FIG. 14.19 Transforming from the 208Y/l2O-volt to the 480Y/277-volt system of the utility by use of autotransformers.

is solved is given in reference 11. Jn this particular case the supply transformer mas connected Y-delta, and on the 480-volt delta secondary three single-phase transformers were connected Y-delta t o obtain the neutral of the 480-volt system. It must be remembered that the secondary of such a grounding bank must be connected delta to result in a low zerophase-sequence reactance for the grounding bank and that the singlephase transformers, when operating in this manner, are only a t 57.7 per cent of their rated voltage. Anot,her method t o ohtaiuthissystemneutral would be a grounding transformer. i n either case the fact that a system neutral is not available subtracts from the possible savings of the 480-volt lighting system. Of course, this problem is eliminated when a delta-Y transformer is used. One additiorial problem of the system that is grounded by a grounding or a Y-delta t,ransformer is the resultiug zero-phase-sequence reactance. Care must be taken to choose either an especially low reactance or a large transformer for grounding, to maintain an X , / X , ratio of about 1. This is done to prevent extreme overvoltages on the fluorescent lights during a ground fault 011 the system. This problem is also eliminated when the delta-Y transformer is used. When power is supplied a t high voltage, there is an additional advantage in using a romhioed light and power system. By using a douhleended unit substat,ion, a certain amount of reserve capacity can be realized quite simply; however, if a power substation at 480 volts is used v i t h a lighting suhstation a t 120 volts, Fig. 14.20, the secondary selective

752


interrupting rating a t least equal to the short-circuit duty a t the point of application of the fuse. Tests have shown that the 600-volt 20-amp fuse mentioned above will adequately protect the relay up to an available short-circuit current of 25,000 amp rms. This fuse has an interrupting rating of 100,000 amp. 2 . A fuse is riot a complete protective device but requires a suitable switch in series with it. Care should be taken in selecting this switch. 3. I t is therefore essential, to obtain adequate short-circuit protection. of the 480Y/277-volt system branch-circuit lighting in office buildings, that panel hoards with suitable switches and not larger than 20-amp current-limiting fuses be used. 4. Circuit-breaker panel boards are not adequate because circuit breakers are not current limiting, and in general the small panel-boardtype circuit breakers do not have high enough interrupting rating. 5. Standard NEC fuses are not in general adequate because they are not as current limiting as the CLF fuses and do not have established interrupting ratings arid have limited interrupting ability on a-c circuits. The required reliability of apparatus and devices for use in commercial buildings will eventually demand that all manufacturers of electrical equipment adhere t o Section 11 15 of the National Electrical Code. This section requires that the “devices intended to break current shall have an interrupting capacity sufficient for the voltage employed and for the current which must be interrupted.”

THE NATIONAL ELECTRICAL CODE PERMITS 480Y/277-VOLT LIGHTING The 1953 Kational Electrical Code revised the 1951 Code to permit the use of the 480Y/277-volt lighting system in commercial buildings. This permission is given under Article 210 entitled “Branch Circuits,” and specifically Section 2113 entitled “Voltage.” Section 2113 deals with restrictions placed on Voltage for lighting “ ( 1 ) in industrial establishments . . . ; ( 2 ) in railroad properties . . . ; (3) for infra-red industrial heating appliances , . . ’ I and “dwelling occupancies.” Because this section plays such an important part in the application of the higher voltage lighting system, it is appropriate that that portion of the 1953 Code dealing with lighting for commercial buildings be quoted in its entirety. “Branch circuits supplying lampholders, fixtures, or receptacles of the standard 15-ampere or less rating shall not exceed 150 volts to ground, except ( 1 ) in industrial establishments, office buildings, large schools and stores, the voltage of branch circuits vhich supply only the ballasts for electric discharge lamps in permanently installed fixtures mounted not less than eight feet from the floor, which do not have manual switch control as an integral part of the fixture, may exceed 150 volts t o ground, but


753

shall not exceed 300 volts to ground for branch circuits supplying lighting fixtures only that are equipped either with mogul-base screwshell lampholders or with lampholders of other types approved for the application, mounted not less than 8 feet from the floor, which do not have switch control as an integral part of the fixture . . . .” This section of the Code first appeared in the Yational Fire Protection Association report on the proposed Amendments to the 1951 Xational Electrical Code, NFPA No. 70 PR1, in August, 1952. It contained a “comment” concerning the revision of Section 2113 to permit higher voltage lighting for commercial buildings and is also quoted. “The proposed new wording extends the use of 277 volt to neutral supply, to additional occupancies only where the higher voltage circuits supply the ballasts of fluorescent lighting fixtures. Since power a t 480Y/277 volts is not generally available except from unit substations or transformation equipment installed a t the particular occupancy, the limitation to ‘office buildings’ and ‘large schools and stores’ seems to supply the necessary assurance that this system will not be encountered in small detached store buildings or in the ordinary primary school buildings.”

REMOTE-CONTROL LIGHTING IS FLEXIBLE

A large degree of flexibility is imparted t o our lighting system by the 24-volt remote-control system. Regardless of how well a lighting system has been planned, especially in office buildings, a rearrangement often becomes necessary or desirable. The low voltage of the remote-control system simplifies moving the lighting switches t o suit the rearrangement. Since conduit is seldom required, it is as easy t o rearrange control switches as it is to install or relocate a telephone. Rewiring is done in a matter of hours instead of weeks, as is often the case with conventional line systems. TO PANELBOARD

OFF COlL

ON COlL

TO LIGHTING FIXTURE

m IPO”

FIG. 14.21

Schematic diagram of the 24-volt remote-control circuit.

754

P O W E R SYSTEMS F O R C O M M E R C I A L BUILDINGS

In its simplest form, this control system consists of a 2Cvolt switching circuit, that, energizes a relay mounted on or near t h e lighting fixture. The relay i n turn controls the lighting circuit. A 120-24-volt transformer orl ollc of the floor circuits supplies 21-volt poirer for a larger number of switching circuits. Figure 14.21 shows such a Ion-voltage circuit. Commerrial-1,uildirrg application of the 480Y/277-volt, system is possible berausc the use of remote-control equipment permits the location of all higher voltage circuits at approved heights. These load circuits are controlled by the low-voltage control system- the only portion of the distrilrmtion syst,em nithin reach of t,lle building occupants. Idow-voltage remote cont,rol. making possible the use of inexpensive

FIG. 14.22 Multirelector remote-control switches providing individual and grouped con. trol of lighting and other facilities.


755

wire in the swit,ching circuits, permits a wide variety of special switching circuits at low cost. A few of the possible arrangements are indicated here. Many more will occur on reviewing specific conditions. Any one or all of a number of fixtures can he coutrolled from a single location by installing multiselertor switches. The same fixtures ran he controlled from several different locations either individually or collectively. In short, switching combinations are pract,ically unlimited. Figure 14.22 shows a n example of multiselector switching. Watchmen and cleaners can enter at one point, turn on the lights they need t o perform their duties, and turn off the lights a t another location when they leave. Lights in storerooms, closets, vaults, stairwells, and halls can all bc controlled from a central location, such as the huilding manager's office. Time switches can easily he incorporated into the system. Arid because of lower miring costs and the fact that each time switch can control more lowvoltage circuits, such installations are less expensive than for 120-volt lighting. REFERENCES 1. Lang, R. R., High-voltage Fluorcseent-lighting Power Systems for Office Build-

ings, General Electric Company Publication GET-2307A. 2. Krtufmann, R. I%.,and N. A. Kieh, Modern Electric Powrr Distribution Ideas As Applied in a Large War Plant, AIEE Technical Paper 45-78. 3. Beeman, D. L., Combined Light and Power Systems for Industrial Plants, AIEE Technical Paper 46-52. 4. Halberg, M. N., and D. L. Beeman, Power Supply for Air Conditioning Compressor Motors in Commercial Buildings, AIEE Conference Paper, November, 1947. 5. Brereton, D. S., Distribution a t 480Y/277 Volts for Power and Lights in Industrial Plants and Office Buildings, AIEE District Paper, April, 1852. 6. Saunders, C. C., The Philosophy of Electrical Design for the Office Building of Tomorrow, AIEE Conference Paper, June, 1952. 7. Brereton. D. S.. Recent Pronress in 480Y/277 Volt Systems in Office Buildings, AIEE District Paper April, 1953. 8. Kurt, H. D.. and D. L. Bceman. Pronrcss in Power System Engineering for Commercial Buildings, AIEE Technical Paper, A Z E E Speczal Publicalion 3-66, September, 1954. 9. Brereton, I). S., and H. J. Donnelly, 480 Wye/277-voIt Power System in Telephone Building a t Menands, N. Y., AIEE Technical Paper, A I E E Special Publication S-66, September, 1954. 10. Lang, R. R., Lighting for Modern Office Buildings, Load, December, 1951. 11. Sawyer, A. I., Voltage Up-Wiring Cost Down. Electrical West, December, 1952. 12. Beeman, D. L., and H. D. Kurt, Higher Network Voltages in Large Buildings, Elec. Wodd, Mar. 8, 1954. 13. Beeman, D. L., and H. D. Kurt, 480-volt System Makes Strong Bid for Office Buildings, Power, March, 1954.

7.56


14. Beernan, D. L., and H . D. Kurt, Commercial High-voltage Installations-Parts I and 11, Eleelrical Consl~uetionand Mainlenance, April, 1951, and May, 1951. 15. Henderson, IT. IT., and H. S. Hiekok, Why 277/48O-v Building Distribution? Elec. World, Oct. 18, 1954. 16. Kurt, H. D., Parameters Indicate Where 480Y/277 Volts Saves Wiring Dollars, Elec. World, S o v . 29, 1951.

Chapter 15

by Donald Beeman and L. G. Levoy, Jr.

Modernization and Expansion

of Existing Power Systems When modernizing or expanding a plant power distribution system, a thorough review should be made of the entire system regarding present and future requirements. A t this time it is possible t o obtain a more economical, flexible, and reliable power system. The desired system change can be then made easily and a t lower cost. There are two phases t o this problem: (1) Modernization, i.e., when some part of the system is t o be replaced because it has become unsafe or obsolete. There is often little or no expansion of capacity in this phase. (2) Plant expansion, i.e., where there is to be an increase in power-system capacity t o serve new buildings or increased load in an existing building. The principles used in modernization and expansion of existing power systems are the same as those outlined in the previous chapters. The technique of applying these principles is somewhat different for older systems than when starting with no existing system or facilities.

HOW TO PLAN A MODERNIZATION OR AN EXPANSION PROGRAM

A modernization or expansion program can usually best be planned by first laying out a complete, ideal power system of the type and voltage desired. Plan it as if the entire plant were t o be new. The second step is t o see what modifications of this plan are required to use existing equipment that is in good condition, with a reasonable service life left. Every step in modernization and expansion should he based on the ideal plail rather than on the existing system. Too often, good intentions in modernizing power systems are spoiled by letting the existing setup influence too greatly the plans for the future, particularly if the existing system is not the correct one for the plant. 757

758

MODERNIZATION AND EXPANSION OF EXISTING POWER SYSTEMS

MODERNIZING OR EXPANDING WITH THE LOAD-CENTER POWER SYSTEM AND HIGHER DISTRIBUTION VOLTAGE

The greatest savings can usually be made when changing to a higher voltage and a load-center power distribution system, for example, in (1) power systems of several thousand kva which now purchase or generate and distribute power at 600 volts or less, and (2) large power systems of several tens of thousands of kva which now purchase or generate and distribute power a t 2400 volts. A change in existing utilization-device voltage rating is not suggested, hut rather to purchase or generate and distribute power a t higher voltage with step-down substations a t the load centers t o supply the motors with the same voltage as formerly used. Looking a t this another way, the existing secondary and utilization voltage system is left as is and a higher voltage grid or backbone is added to the system merely for the purpose of transmitting power from the source to load centers. This backbone or grid feeds the existing system a t strategic points through step-down transformers. This approach of superimposing the higher voltage grid permits a stiffer system to which the source transformers and generators may be connected and permits far greater growth than would otherwise be practical a t a lower voltage. This approach has been widely used with many different voltage levels. For example, existing 480- or 600-volt systems have had 4160-volt grids superimposed t o carry the power a t 4160 volts from the source to the load centers where it is stepped down to 480 or 600 volts and fed into the existing 480- or 600-volt distribution system. Similarly, large 2400-volt systems have had 13.8-kv grids superimposed. I n this case transformation would be to 2400 volts in some cases and directly to 480 volts in other cases. I n any event, the existing 2400-volt and 480-volt distribution systems in the load area need not be materially altered. In larger plants such as steel mills 13.8-, 23-, 34.5-, or 69-kv grids have been superimposed on 6.9-kv systems. When the grid voltage is above 15 kv, all power is stepped down from, say, 69 kv to some voltage below 15 kv for local area distri6ution. Again the local area distribution is not materially altered. The following are some examples of plants that have adopted higher voltage when modernizing and expanding. 1. Plant A began operations at 480 volts and within a few years decided to go to 2400 volts. Now the plant is generating and distributing a considerable amonnt of power a t 13,800 volts. 2. Plant B, which began operations a t 600 volts, recently rewound the supply units to 4160 volts, using this for distribution voltage. 3. Plant C began operations a t 600 volts. It has rewound some supply units to 2400 volts and added new supply units a t 2400 volts.

MODERNlUITlON A N D EXPANSION OF EXISTING POWER SYSTEMS

759

4. Plant D (nine paper mills) originally had mills that were either 480 or 600 volts. However, the later mills have been 2400 volts, and a new mill to be built will be 4160 volts. A study is now being made with respect to raising the voltages in the older mills. 5. Plant E’s original plant distribution system was at 600 volts. Later distribution was a t 11,OOO volts. 6. Plant F’s original installation was a t 2400 volts. Recently 13,800 volts was installed. 7. Plant G’s original installation was a t 480 volts. Later 2400 volts was installed. Now 13,800 volts is being used. 8. Plant H’s original installation was a t 600 volts. Later it was a t 13,800 volts. 9. Plant 1’s initial generators were 2400 volts. Generators now are installed a t 13,800 volts. ADVANTAGE

OF CHANGING TO HIGHER VOLTAGE

The advantages of higher voltage distribution and the load-center system are aa follows: 1. Plant expansion can be made a t lower cost. 2. The bottleneck caused by excessive short-circuit currents in lower voltage systems is removed. 3. The voltage drop is reduced. 4. Losses are reduced. 5. Modernization and expansion can be done on a step-by-step basis. These advantages will be exemplified under the discussion of modernization and expansion which follows. However, to set down some basic principles, a few general cases are treated first. Lower Cost Expansion. Assume that a load of 1500 kva is to be carried 1000 f t . If power is generated or is available a t 480 volts, the cost will be about $42,500 for the circuit breakers and cable shown in Fig. 15.1A. If power is generated or available a t 4.16 kv, the cost will be only ahout $32,000 for the circuit breaker, cable, and the 1500-kva load-center unit substation shown in Fig. 15.1B. This is the most difficult case to prove at higher voltages because of the neressity of transformers in one case and not the other. Nevertheless there is a substantial saving in just transmitting power as far as possible a t high voltage as is done in the load-center power distribution system instead of all the way at low voltage. Removal of Short-circuit Bottlenecks. To illustrate the second advantage of the load-center system with higher voltage distribution, which is freedom from short-circuit-current bottlenecks, assume that a 3125-kva generator is to be added to an existing bus. The short-circuit current on the existing 480-volt bus is about 75,000

MODERNILATION AND EXPANSION OF EXISTING POWER SYSTEMS

760

amp. The addition of the new souice either from the utility or a new plant generator and new motor load associated with it will increase the short-circuit current hy about 60,000 amp, or t o a total of 135,000 amp. The latter value is 35,000 amp ahove the interrupting rating of the Iargest availahle lom-voltage air circuit breaker. Therefore, when the new source is added, some expensive step must be taken to remove the short-circuit hottleneck by limiting the short-circuit current t o 100,000 amp or less. This can he done by installing reactors, splitting up the system, etc. After taking these steps ali new switchgear may he required hecause of the increased short-circuit duty on the existing circuit breakers. With a 4160-volt system of this size having 100 mva or higher interrupting-rating circuit hreakers, the new source can he added without exceeding the circuit-hreaker interrupting rating. Therefore, there are no short-circuit-current hottleneeks in a system of this size a t 4160 volts. -1s

-

:IRCUIl BRELKER

'OLTS -4160 --

t

100.000 AMP

A

I.C.

A

150 Y V I CIRCUIT BREAKER

5- 3-CONOUCTOI

500 Y C M pBLEs

E 1 FEET

-CIRCUIT BREAKERS

TR4NSFORYER 1500 K V A

c

50.000 AHP. I.C. 4 0 0 AMP FEEOERS

4 P P R O X . COST $42.500

$ 2 0 P E R KVA

FIG. 15.1

APPROX. COST $32.000 $ 2 1 PER K V 4

E

A LOW-VOLTAGE

J

'400 AMP FEEDERB

SYSTEM

~810-CENTER SVSTEH

Comparison of cost of tronrmitting power o? 600 volts or lerr o1 higher voltage 01 i i done in the lood-center power dirtribution ryrtem.

VI.

tronrmitting i t

MODERNIZATION A N D EXPANSION OF EXISTING P O W E R SYSTEMS

761

Reduction in Voltage Drop. The higher the voltage used t o carry power to load renters, the less the voltage drop. With higher voltage distribution the performance of the motors, heating devices, electronic control, lamps, and other utilization equipment is improved. A typical 1 5 0 0 4 480-volt feeder (Fig. 15.2) to rarry a given load has a voltage drop of about 9 per cent, \\-hereas a typical high-voltage feeder to carry the same load consisting of 1400 f t a t 4160 volts and 100 f t at 480 volts has a voltage drop of only about 0.83 per cent. Transformer voltage drop is the same in either case, i.e., about 3 per cent. See Chap. 4 for a more detailed discussion of voltage-spread problems. Step-by-step Basis. When expanding and modernizing an existing system using the load-center system and higher distrihution and/or generation voltage, the entire existing system need not be changed all at once. Sinre a load-renter system is made up of small unit suhstationsi.e., building blocks-the system may be gradually modernized hy adding a few units each year. This can be done with a minimum disturbance to production schedules. The application of the step-by-step procedure will he noted through the examples. Idalmost every case the change to higher voltage can be made when expanding the plant without expanding, scrapping, or even changing the existing system. After the first step is completed, the old system can be modernized by adding one or more load-center unit substations per year. This eliminates the necessity for a large capital outlay a t any one time. The foregoing examples illustrate the disadvantages of large 600-volt class systems. In many cases, even 2400 or 6900 volts are too low, particularly where loads of approximately tens of thousands of kva are encountered. Here the 2400- or 6900-volt system brings about the same problems and costly expansion as 480 or 600 volts for loads of approximately a few thousand kva. The examples will not he repeated for a 2400- or 6900-volt system because the hasic principles are the same as for a 600-volt class system. WHEN TO CHANGE TO A LOAD-CENTER POWER SYSTEM WITH HIGHER VOLTAGE DISTRIBUTION

The best time to change to a load-center power distribution system using higher voltage distribution t o load-center substations is when modernizing or expanding a power system, i.e., when 1. Installing new switchgear 2. Installing new subst.ation capacity 3. Installing new cable 4. Installing new turbine generators 5. Adding substantial load to the system


763

Do not just add another feeder or transformer, or generator, or new switchgear to a low-voltage system. Look ahead to see if a change in voltage will not pay real dividends. MODERNIZING POWER SYSTEMS

Existing power systems are often modernized although there is no expansion involved at the moment. The switchgear may be obsolete, representing a real hazard to the operating personnel and property. The circuit breakers may also have inadequate interrupt,ing raparity. Substation equipment may he inadequate; transformcrs may be old and have high losses and weak insulation; or cables may he old and their insulation hrittle and subject to increasing frequenry of failure. When modernizing the system by replacing switchgear, substations, or cable, be sure to see (1) if the plant distribution voltage should be increased, or (2) if a different circuit arrangement should he used, or (3) if more provision should be made for future expansion, or (4) if the system should be made more reliable. Such changes can very often be made at this time for little or no more money than would be expended for just replacing the obsolete part of t,he system. Replacing Old Switchgear. When modernizing or replacing existing obsolete switchgear, there may be cases where voltage is at the correct

NEW SWG-

FIG. 15.3

Diagram showing how new metal-clod switchgear con be installed on a balcony above old switchgear to make transfer of circuits from old to new switchgear easy and fast.


764

600 VOLTS 5000-KVA TJRBINEGENERATOR

600 VOLTS 2500- KVA TURBiNE-fl GENERATOR

(600 VOLT BUS,

2 4 0 0 VOLTS 3133-KVAT U R E I N E - 0 GENERATOR

1

/1

AUTOTRANSFORMERS

600 TO‘ 2400

REDUCED-VOLTAGE STARTER

VOLTS 1500-KVA TRANSFORMER

1 5 0 - K W MOTOR-GENERATOR 75 - K W MOTOR- G E N ER ATOR

o 66

5 0 - K W MOTOR-GENERATOR E X C i T E R

FIG. 15.4

One-iine diagram rhowing 600-voll ryrtem ar it exirted in

iumber mill.

leve1 and all other components of the system are adequate for the plant. I n this case, the modern, safe, compact metal-clad switchgear should be used instead of open-type switchgear. There are many problems of a detailed nature in replacing existing switchgear. One method, which permits an orderly installation and change-over procedure with a minimum of service outage, is t o place the new switchgear on a balcony above the existing switchgear (Fig. 15.3). The new metal-clad switchgear may be installed and then the feeder cables transferred from the old to the new switchgear one a t a time. When replacing switchgear, be sure t o check the system voltage and layout. Perhaps you cab get a complete new system for the price of a new switchboard. For example, a sawmill generated power a t 600 volts (Fig. 15.4). The plant engineers were aware of the inadequacy of the existing oil circuit breakers used on the 600-volt system. One alternative was t o install new 600-volt switchgear, the cost of which was about 585,000. By rewinding one generator for 2400 volts and adopting a loadcenter distribution system, a complete new power system (Fig. 15.5), including new unit substations plus new windings in one generator, was obtained for approximately 820,000 more than the cost of a new 600-volt switchhoard. Replacing Old Substations. Many old power systems already have the most desirable primary and secondary voltages and may have rela-


765

tively modern primary switchgear. However, they may have old obsolete substations for stepping power down from primary voltage to utilization voltage. When modernizing such substations, find out what system arrangement is best, and get more for the modernization dollar. For example, the following system existed in a metal working factory in the East. The primary voltage supplied by the utility mas 13.2 kv. The 13.2-kv metal-clad switchgear was less than 10 years old. Irtilization voltage was 480 volts. However, the substations stepping down to 480 volts were very large-5000 to 10,000 kva each. They had inadequatc switchgear and were in poor condition generally. The area around the substations had become congested to such an extent t,hat the old oil-filled transformers represented a real fire hazard should they fail and catch fire. The voltage drop in the system was about 20 per cent mauimum, causing many operating difficulties with utilization equipment-such as failure of fluorescent lamps to light.

50- KW 7-EXClTE%

9

666

r 1

:&:

7 5 - K W MOTORGENERATOR

1 5 0 - K W MOTOR' GENERATOR L--J

LOAD-CENTER SUBSTATION IN THE M I L L FIG. 15.5 One-line diagram showing the modernized system in o lumber mill. original system is sho.:n in Fig. 15.4.

The

766

MODERNIZATION A N D EXPANSION

OF EXISTING POWER SYSTEMS

These substations could have heen replaced a t the same loeation with large unit substations of modern construction. Such large substations are costly and inflexihle, however. A review of this system showed two things: (i) there were many big motors which could hetter he served a t 2400 volts instead of 480 volts, and (2) small substations sround the factory area would cost less than the Iarger substations. They would improve voltage conditions and allow severa1 thousand feet of existing secondary feeder cable t o be sold as scrap. This woulù further increase the savings a i t h the load-center system. The system adopted (Fig. 15.6) eost about 10 per cent less than just replacing existing substations. The feeder voltage drop was decreased from 20 to about 2 per cent maximum. By adopting some substations with 2400-volt seeondaries, the larger motors (above 200 hp) which were purchased new were rated 2300 volts instead of 440 volts, as would have heen necessary with the old system. FROM U T I L I T V

TO OTHER

SUBSTATIONS

KVA

KVA

CIRCUIT SUBSTATION

66666666 2300 VOLT MOTORS AND STARTERS FOUR SIMILAR LDhD CENTER UNIT S U B S T A T I O N S A T LOAO C E N T E R S IN P L A N T

FIG. 15.6 One-line diogrom of modernired load-center power distribulion ryrtern where 480-volt load-centei unit rubstotionr and 2400-volt moster uni1 rubrtationr replace old Iarge low-voltage rubriotioni.

MODERNlZAllON AND EXPANSION OF EXISTING POWER SYSTEMS

767

This change-over was made on a step-by-step basis, installing a few new load-center unit substations each year. Thus, there was no large outlay of money during any one year for modernization of the power system. Replacing Old Cable. When cable insulation gets old, it usually gets brittle and mechanically weak. Cable failures have become so frequent in some old iristallations that something had to be done t o correct the situation. One corrective approach might be to replace the existing old cable. However, a better system might he obtained if a load-center power system with higher voltage distribution were selected. For example, one plant had a large (about 10,000 kva) 600-volt power system fed from one substation bus. The 600-volt feeder cables were all in underground ducts. Cable failures were reaching the point where somethitig bad to be done. About 25,000 ft of cable had to be replaced. I t was estimated that it mould cost $5.60 per f t , or $140,000, to replace cable. This did not include the necessary cost of tearing out the old cable. A modern load-center system (Fig. 15.7) cost about $190,000 installed. However, the cable copper losses were reduced in the loadcenter system because the power is carried most of the way at high voltage instead of all the way a t low voltage as with the old system. The capitalized savings in losses were $15,000 ($1500 annual saving in power bill capitalized at 10 per cent). This reduced the cost of the load-center system to only $175,000 ($190,000 - $15,000). The old 2500-kva transformers had a resale value of $17,500. The net extra cost of the loadceiiter system was only $17,500 more than just replacing the cable. The load-center system had several advantages: 1. Reduction in feeder voltage drop (from 10 to 1.8 per cent)-which improved performance of utilization equipment. 2. S e w secondary switchgear (600 volts) of adequate interrupting capacity as obtained with the new load-center power system. 3. Lower cost future expansion of the system. 4. More flexible system. A completely moderii system was obtained for only $17,.500, or $1.75 per kva, by rhangiiig to the load-center system inatead of just replacing cable. Again this change over utilized the building-block characteristics of a load-renter system. A few small substations have been added, and a few more \vill be added each year until the job is complete. EXPANDING A N EXISTING PLANT

When expanding an existing plant, there is an ideal opportunity to start a modernizatiou program of the power distribution system at little or no extra cost. Whenever a n expansion is planned, check the system layout to see if a load-rrnter system would be better, but most, impor-


769

tant-check t o see if a higher distribution voltage would not be more economical. Often it is said, “This is the last expansion we’ll ever make.” Once in a great while that may be true, but it. is not the history of the great majority of industrial plants. Even if no new buildings are built, the use of more electric power per unit of product plus greater utilizing of existing floor space will increase power demand in a given factory. This causes expansion of existing power systems. The most expensive single error often made is to design a power-system expansion on the basis that there will never be another expansion. Such a procedure makes it all the more expensive when the next expansion does come. The time to set up a proper power system is almost always a t the time of present expansion of an old system and not the next expansion. Here are several examples to illustrate the more important considerations when expanding a plant power system. Expanding a Small Plant. Supposing a plant (Fig. 15.8) wants to add 1500 kva of load in a new addition. The existing substation could he expanded as shown in Fig. 15.9. This would cost about $26,000, the voltage drop would be 9 per cent, and the cable copper losses would cost about $280 per year. Making this expansion by a load-center method (Fig. 15.10), high voltage would be taken t o a load-center unit substation a t the load center. The cost of the expansion by the load-center method would be about $26,000. The voltage drop would be less than 1 per cent, 13.2 KV LINES FROM UTILITY EXISTING SUBSTATION 3000 KVA

I

I

I

FIG. 15.8

NEW ADDITION REQUIRING ADDiT1oN 1500 KVA

Layout of plant requiring 1500 kvo additional substation capacity.

MOOERNIZATION A N D EXPANSION OF EXISTIdG POWER SYSTEMS

770

and feeder ropper losses xonld be negligible. Therefore, by expanding this system on the load-center system basis, the user saved $280 per year in losses at no increase in first cost. Also, the user started a system which will fit into complete modernization of the plant (Fig. 15.11) which can be done on a step-by-step basis and spread the expense over several years. For small or large plants, the same savings can be obtained by using the load-center system for plant expansion. In larger plants, several small substations are used rather than a few large ones. Use Higher Voltage and Eliminate Double Transformations. There are plants where double transformation is used, for example, from 13.8-kv

I

II THREE

I000 KVA

4

13.2 Kv' h
I

I

INCOMING LINE CIRCUIT BREAKERS

I

1

SINGLE U PHASE I TRANSFORMERS I

NSFORMEAS

I I I

I L

--_---___ EX IS T IN G SUBSTATION

NEW BUlLOlNG

FIG. 15.9 One-line diagram for expanding electrical system for plant shown in Fig. 15.8, using low-voltage distribution to supply power in new oddition.


771

1

I

I I

l

~

~

~

~

~

v

A

SINGLE PHASE

I TRANSFORMERS I 1

2 0 0 F E E T OF 13.2 KV LEAD-COVERED CABLE

CIRCUIT BREAKERS1

L-

-___---

A

E X I S T I N G SUBSTATION

1- - - - 1

I .

L-

I I I I

TRANSFORMER ' S

:! CIRCUIT BREAKER \

' I

NEW 1500 KVA LOAD CENTER

I I

I I I

=-' I/

-LI=.f I I L _ _ _ _ _ L- _ _ _ J I

NEW 0 U l L D l N G One-line diogram of expclnded electrical system for plmt shown in Fig. 15.8, using load-center distribution to supply power in new addition.

FIG. 15.10

class voltage (utility supply voltage) to the 2.4 kv and from 2.4 kv to 480 volts (Fig. 15.12). When expanding such a system, the double transformation can be eliminated and power taken to the load centers a t utility supply voltage and stepped down directly to 480 volts (Fig. 15.13). The National Electrical Code permits voltages up to 15 kv in buildings. Eliminating this double transformation will save about $4 to $15 per kva

772

MODERNIZATION AND EXPANSION OF EXISTING POWER SYSTEMS 13.2

ADDED DURING EXPANSION EQUIPMENT SHOWN B Y DASHED L I N E S

4L

6

I500 KVA

& 80 VOLTS

535 555 + & &

Y Y Y

+++ Y Y Y

I

I I I [-OLD

BUILDING

I

CONNECT TO EXISTING 4 8 0 VOLT FEEDER CABLES

i I

IOOOKVA

1946

SOLID LINES snow NEW EQUIPMENT FOR MODERNIZATION OF OLD PLANT. STEPS TAKEN 1 9 4 5 - NEW 1500 KVA SUBSTATION INSTALLED 1946- NEW DOUBLE ENDEO 2000 KVA SUBSTATION INSTALLED 1947- NEW 13.8 KV METAL CLAD INSTALLED PLANT NOW EXPhNDED AND MODERNIZED WITH LOAD CENTER SYSTEM ON A STEP BV STEP BASIS.

FIG. 15.11 One-line diagram of expanded and modernized electrical system for the plant shown in Fig. 15.8, using load-center distribution system installed on the step-by-step baris.

MODERNIZATION AND EXPANSION

OF

EXISTING POWER SYSTEMS

773

in the cost of the power system. The double transformation may.be necessary where the utility primary voltage is above 15 kv because voltages above 15 kv should not in general be taken into the buildings. I n some cases, such as a t pumping stations where a11 but a very small amount of power is used a t 2400 volts, double transformation to 480 volts may be justified. EXPANDING PLANT POWER GENERATING SYSTEMS

There are many large plants in which considerable process steam is used and which, therefore, generate much of their own pomer. Many such plants have large low-voltage (600 volts) systems or very Iarge 2400-volt systems. Expansion of such systems presents a real opportimity to ehange to a load-center system and to higher voltage generation and thus

I

12.5 KV INCOMING L I N E FROM U T I L I T Y

METERING

LI,

IA

1500 KVA E A C H

,T--

~

2 4 0 0 VOLTS

4 8 0 VOLT POWER

2 4 0 / 1 2 0 VOLTS LIGHTING

T 8 0 V O L T POWER

FIG. 15.12

An exirting power dirtribution ryrtem with unnecerrriry double trmrformation

of aII power from 12.5 to 2.4 k.

774

MODERNIZATION AND EXPANSION OF EXISTING POWER SYSTEMS I

!

12.5 K V INCOMING LINE FROM UTILITY

i I I I

L

~

f

3 9 + . L

,

I

4 7 VOLT8

Y - -Y_ - -Y_ ___

I _J

FIG. 15.13 b B d lines show expansion of Ihe system shown in Fig. 15.12, wing the k a d ceder power distribution ryrtem in which power i s stepped down directly from 12.5 kv to 480 volts.

to save money. Examples will be used to illustrate the points involved. The following also applies where power sources are from transformers instead of generators. Methods of Expanding at Higher Voltage. Expanding a lon-voltage system a t higher volt,age is a simple procedure. It merely requires the addition of a bus-tie transformer bet,meen the buses. Then expansion can he t o t,he higher voltage as if the lower voltage system did not exist.

MODERNIZATION ANO EXPANSION OF EXISTING POWER SYSTEMS

77s

The old low-voltage and new high-voltage systems can be interconnected in severa1 ways. Figure 15.14 shows one method of employing a hus-tie transformer t o

-

NEW EXISTING

O00

NEW M E T A L C L A D SWITCHGEAR

-/ $

OR 2 4 0 0 VOLTS

1 I

BUS TIE T R A N S F O ~ M E R 4160 VOLTS OR

13.8 KV

I

A

/

X

LOAD CENTER UNIT SUBSTATION

TRANSFORMERS

--

'-0RE

I

CIRCUIT AKERS

6 0 0 OR 2 4 0 0 VOLTS EXISTING

EXISTING SWITCHGEAR REACTORS +

-NEW METAL CLAD SWITCHGEAR

13.0 K V

BUS T I E TRANSFORMER

0

"

1,

T O LOAD

FIG. 15.14 One-line diagrom rhowing how
776


connect a new system a t higher voltage t o an old system operating a t lower voltage. There may not he circuit breakers ou both sides of the bus-tie transformer. In many cases the cireuit breaker is omitted from the low-voltage side because of the Iarge currents involved but is included on the high-voltage side. If so, it is assumed that the transformer is part of the low-voltage bus, and proper relaying must be used to protect the transformer. However, it is generally not desirahle to have it directly connected to the high-voltage bus without an intervening circuit breaker. Figure 15.14 also shows the application of the hus-tie transformer where a synchronizing bus is used on the low-voltage system. I n a few cases, where there has heen a very large expansion of a power system that necessitates the addition of a synchronizing-bus arrangement for the new high-voltage turbine-generators, the arrangement shown in Fig. 15.15 has heen used. The existing 2400-volt bus is sectionalized. One section of the 2400-volt bus is connected t o each of the new highvoltage buses through a transformer. This saves iiistalling a synchro-

O

NEW

i

O NEW METAL CLAD SWITCHQEAR

13.8 K V

SYNCHRONIZING BUS

EXIÇTING E 4 0 0 VOLT 8US SPLIT HERE

FIG. 15.15

One-line diagrem of (I ryrtem in which the exirting 2400-volt bur ir rectionalired and connected throuqh transformeri to a new 13.8-kv ryrtem. The 13.8-kv rystern acts 01 a rynchroniring bur ior the 2.4-kr ryrtem.


777

niaing bus on the existing lov-voltage system, and permits circuit breakers to be employed on both the high- and lorn-voltage sides of the bus-tie transformers. Hoivever, it might not be feasible to connect transformers to the synchronizing buses through circuit breakers in surh large highvoltage systems because the duty on the synchronizing buses may be beyond the rating of available circuit breakers. I n some cases it is desirable to establish a new higher voltage bus for supplying new load before a new higher voltage generator is added. The arrangement shown in Fig. 15.16 will enable this t o be done at minimum cost. The new higher voltage feeders are fed through a step-up bus-tie transformer from the existing low-voltage bus. A t a future date a new generator may be added to the higher voltage bus. The load-center system is employed for serving the new load.

OLD OIL CIRCUIT BREAKERS

TRANSFORMER

NEW BUS 4160 VOLTS

600 VOLTS OLD

BUS

GENERATORS

NEW METAL CLAD SWITCHGEAR

FEEDERS TO NEW LOAO CENTER UNIT SUBSTATIONS

‘.-,,

I

:“,‘,“,“6,,,

FIG. 15.16

Existing 600-volt bur extended a t higher voltage through bur-tie transformer lo supply new load through load-center unit substations before new higher voltage generator i s added.

LARGE 6OJ-VOLT CLASS SYSTEMS

Figure 15.17 shows the one-line diagram of a large existing system of the 600-volt class. This particular plant happens to be operating a t 480 volts. The addition of a new 3500-kw 0.8-power factor 4375-kva turbogenerator and 2000 kva of new load is planned. Also, dismantling the (two) 500-kva turbogenerators is contemplated because they are old and inefficient. The short-circuit current in the existing plant, without the new turbinegenerator, is about 115,000 amp. This is three t o five times the interrupting rating of the existing oil circuit breakers. If the new turbine is added to the existing bus as shown to the left in the one-line diagram of Fig. 15.17, the short-circuit current will be increased t o about 165,000 amp. This further accentuates the unsafe

778


conditions of this switchgear and materially increases the hazard already existing. Such a condition should not be tolerated in any plant. If, however, the new generator is added at 480 volts, any switchgear associated with this generator and the new feeder circuit breakers should have adequate interrupting rating for the service. This would require a switching arrangement such as shown in Fig. 15.18. While this arraugement provides for adequate switchgear for the new installation, the shortcircuit current on the existing switchgear will increase to some extent. The increase will not be so great as if the generator were added as shown in Fig. 15.17. The cost of the new switchgear, new reactors, and the bus connection between the generator and the switchgear shown t o the left in Fig. 15.18 is approximately $65,000. A significant part of this cost (about $8500) is for the bus between the generator and the switchgear. About 50 ft of 6000-amp bus is required. Distributing the power at 480 volts t o the new 2000-kva load will cost about $28 per kva, or $56,000. This is based on the arrangement shown in Fig. 15.1.4 and includes the feeder breakers at buses A and B , feeder cable, and new switchgear a t the load (dashed lines, Fig. 15.18). This makes a total cost of $121,000 ($65,000 plus $56,000) for the new switehgear and for the distribution system to care for the 2000 kva of new load. If the new generator were added at 4160 volts instead of 480 volts, the arrangement would he as shown in Fig. 15.19. The new 4160-volt bus would be connected to the existing 480-volt bus through two 1500/1875kva transformers. The two 1500-kva transformers are used rather than one 3000-kva transformer since the 1500-kva transformers may later be used as part of load-center unit substations. T h e interrupting duty on the existing 480-volt switchgear is increased by about the same amount for the arrangement shown in Fig. 15.19 as for the arrangement shown

KVA

2500 KVA

1075 K VA

500 YVA

500 K VA

Y 2 0 FEEDERS

FIG. 15.17 A large existing low-voltage power system with excessive short-circuit currents to which is to be added the new generator shown to the left.

I

I

I I I

Y

NEW

I

I

I

I

I

i

I I i

4 8 0 VOLTS

-

O L D OIL []+CIRCUIT d [ ] BREAKERS

ing a duplex reactor necessary to keep the short-circuit and load currents within the ratings of available air circuit breakers.

2 0 E X I S T I N G F E E D E R S AND SWITCHGEAR

F E E T , 6000 AMP BUS

FIG. 15.18 The lorge existing low-voltage system shown in Fig. 15.17 with the new generator and the required rwitchgeor inciud-

\

SYNCHONIZ ING

DUPLEX

r\

4 3 7 5 KVA

780


L-.

MODERNIUTION AND EXPANSION OF EXISTING POWER SYSTEMS

78 I

in Fig. 15.18. The cost of the new 4160-volt generator circuit breaker, the connections between this circuit breaker and the generator, the two 1 W k v a transformers, and the 4160-volt breaker serving the two 1500kva transformers is about $43,000. 1nst.allation costs of this equipment are assumed to be about the same.as for the equipment in Fig. 15.18. However, the arrangement of Fig. 15.19 will probably require less space in the powerhouse. To care for the new 2000-kva load will necessitate two 4160-volt feeders and a double-ended 2000-kva unit substation, shown by the dashed lines in Fig. 15.19. The cost of this portion of the distribution system is about $21 per kva (Fig. 15.1B). Therefore, this represents an investment in a distribution system of approximately $42,000. The total cost of the switchgear for the generator and connections to the 480-volt existing bus, plus the new distribution system, is $85,000 ($43,000 $42,000), compared with $121,000 for adding the new generator and serving the new load directly a t 480 volts. This saving is approximately 28 per cent on switchgear and distribution equipment cost. If the new generator is added a t 480 volts, modernizing the rest of the system a t 480 volts (Fig. 15.20) mould cost an additional $84,000 for switchgear and reactors for the 2500-kva and 1875-kva generators and twenty feeders. There would he no new distribution-system expense in this case. This system would still have high voltage drop and high feeder losses because of the long 480-volt feeders. Modernizing this system a t 4160 volts could be done on a step-by-step basis, as shown in Fig. 15.21. The first step would be to install new suhstation capacity t o care for 2500 kva of load from the 4160-volt bus. Load-center unit substations, up to 1500 kva each, could he used eeonomieally. After these substations are installed, the second step (dashed lines, Fig. 15.21) would be t o rewind the 2500-kva generator for 4160 volts and connect it to the 4160-volt bus. The third step would be to install new 480-volt sivitehgear for the 1875-kva generator, as shown to the right in Fig. 15.21. Credit is given for one 1500-kva transformer which mas formerly used for the bus tie between the low-voltage and high-voltage buses. The cost of modernizing the existing 480-volt system a t 4160 volts would be about $100,000. Of this, $31,000 would be for new paverhouse switchgear and reminding the 2500-kva generator and $52,500 for 2500 kva of distribution system capacity a t $21 per kva (Fig. 15.1B). This figure of $100,000 for modernization of the existing system at, 4160 volts compares with $84,000 for modernization of the existing system at 480 volts. The total cost of the modernized and expanded 450-volt system (Fig. 15.20) would be $205,000 (5121,000 for the system associated with new generator and $84,000 for modernization of the switchgear associated with existing generators).

+

MODERNIZATION A N D EXPANSION OF EXISTING POWER SYSTEMS

701

-G

a

(L

w

0

0

a 0 v)

N

a

> Y

n I-

n

*

W

;5 2: '5I3-

0-I

784

MODERNIZ4TION AND EXPANSION OF EXISTING POWER SYSTEMS

The complete cost for additions and modernization at 4160 volts (Fig. 15.21) would be $185,000 ($85,000 for the system associated with the new generator and $100,000 for modernization of the evisting 480-volt system a t 4160 volts). There is a saving of about $20,000 by using the high-voltage method. The other advantages are less voltage drop, therefore better performance of equipment, lower losses, and at the same time a completely new system is obtained. A big advantage of the 4160-volt system is that the next turbinegenerator may be added for an expenditure of about $12,500 for a circuit breaker to connect the generator to the 4160-volt bus. The future addition of a turbine-generator to the arrangement shown in Fig. 15.20 would require switchgear as shown for the 4375-kva turbine-generator (Fig: 15.20), which would cost approximately $49,000. This represents a saving of $36,500 ($49,000 - $12,500) for 4160 volts, and, in addition, there would be approximately $5.50 per kva saved in the distribution system using the load-center system and 4160 volts generation. limits of Sizes of 600-volt Sources. The limits of generator size for the system shown in Fig. 15.20 are 3,125 kva when the synchronizing bus reactor, as shown t o the right, is used and 6250 kva when the duplex reactor is used for systems rated 600 volts. For systems rated 480 volts, these figures are 2500 and 5000 kva, respectively. I n many cases transformers as well as generators may be used to supply a large part of the 600-volt system. The transformers may be treated just like generators, and the circuit arrangements shown in Figs. 15.18 to 15.21 may be employed. The limits given above for the system using one synchronizing bus reactor and generator per bus section enable keeping the load current helow 4000 amp per circuit breaker and the short-circuit currents below 100,000 amp. This will permit the use of 50,000-amp interrupting-rating feeder circuit breakers in cascade. These limits are chosen because 4000-amp circuit breakers are generally the largest size available in drawout construction. They will also satisfactorily cascade with 50,000-amp interrupting-capacity feeder circuit breakers. It is necessary t o limit the total short-circuit current t o 100,000 amp, as 100,000 amp is the largest air circuit breaker made. This also permits the use of the smaller and lower cost 50,000-amp interrupting-capacity feeder circuit breakers in cascade with 100,000-amp main circuit breakers. The limit chosen for the duplex reactor arrangement enables keeping the load current below 4000 amp per circuit breaker. Also, it will keep the short-circuit currents within 100,000 amp and not use excessively high reactance in the synchronizing bus reactors. Balancing Reactors to Eliminate Unbalanced Currents in l a r g e 600volt Generators. There are other serious problems involved where large low-voltage generators are used. Observations show that the three


705

phases of large generators may be carrying different currents, which cannot be eliminated by shifting or removing all single-phase loads. These unbalanced currents are the result of generator lead bus-bar unbalance, which is particularly noticeable in plants where the main load consists of motors operating directly from the generator bus without intermediate transformers. Unbalanced generator phase-current values of 5 to 10 per cent have been observed in such plants. This is particularly objectionable where non-salient-pole or wound-rotor turbine-driven units are used since current unbalance causes additional field heating inside the machine. This additional heating is usually less severe in salient-pole marhines. When a turbine-generator is carrying a balanced three-phase load, the magnetic flux in the rotor is constant except for minor ripples caused by tooth-and-slot effects. However, when the stator current is unbalanced in the three phases, the rotor flux tends to pulsate a t twice normal frequency. This generates relatively severe eddy currents in the rotor structure, causing added heating and, sometimes, burning of rotor wedges. I n severe cases, the rotor may he badly damaged from overheating even though the generator is carrying less than rated load. Generator loading must be reduced t o avoid excessive heating when this unbalanced current flows. In addition to the effect on the generator, the unbalanced current causes similar overheating in motors. It also reduces their availab!e peak torque and decreases their effirieucy. Consequently, in the interests of good service and good system operation, it is desirable to keep the phase currents as closely balanced as possible. Several cases of phase-current unbalance were investigated, and each revealed that trouble was caused by relatively long runs of bus bars mounted side by side in one plane. This flat, equal bus-bar spacing introduces a dissymmetry .in the circuit. The three phases of the bus do not have equal inductance and impedance because space between the two outer bus bars is twice that between the middle bar and either of the other two. Although the difference in impedance may be small, it causes an unbalance in voltage. These unbalanced voltages are composed of two balanced symmetrical components: (1) the ordinary positive-sequence voltage and (2) a symmetrical negative-sequence component, which is usually small compared with the positive-sequence voltage. It happens, however, that motors and generators have relatively low impedance to this negative-sequence voltage; therefore a small negative-sequence component of voltage produces a relatively large negative-sequence current. Consequently, the effect is magnified; so a small negative-sequence voltage appears as a larger percentage of unbalanced current than the percentage of voltage unbalance that exists. The effect is especially noticeable where motors


786

and generators are connected by unbalanced bus runs without intermediate transformers. When transformers are used hetween motors and generators, both the positive- and negative-sequence impedances of the circuit are increased by equal amounts; so the unbalance is not so pronounced as described above. I n static loads of heaters, for example, the positive- and negative-sequence impedances are equal and the negative-sequence current is not amplified. When severe unbalances of currents are found to exist because of long low-voltage bus runs with flat equal spacing, they are usually corrected most readily by adding some reactance in series with the middle leg of the bus, thereby balancing the reactance in three legs. A simple way of obtaining this reactance is by installing a set of laminated silicon-steel punchings around the middle leg of the bus, as shown in Fig. 15.22, with an air gap to prevent saturation. This trouble was corrected in one plant with a 3750-kva three-phase 60-cycle 600-volt turbine-generator used to drive a group of induction motors. The generator bus was 100 f t long, consisting of channel bus bars with flat horizontal spacing on 11-in. centers, Fig. 15.23. Feeder taps t o motors were made from a small cross bus run. The generator could carry only about two-thirds of rated load without overheating. PHASE A

9 1 COPPER BUS REACTOR ON CENTER BUS

1

SILICON - STEEL PUNCH~NGS/

AIR GAP

tY=Y

4 LlJ PHASE C

FIG. 15.22 Phase currents are balanced in large lowvoltage generator leads b y adding sheet silicon-steel punchingr to increore reactance of center leg of generator leads.

MOTOR FEEDERS

FIG. 15.23

Instollation of large low-voltage generator in which a bur layout caused unbalanced phase currents in the generator and system which increased the generator heating. Load unbalance corrected by oddition of silicon-steel punchingr to increase reactance of center lag of generator leads.

Currents in the three legs of the generator bus were checked and found to be 2400, 2600, and 2500 amp, respectively. The middle leg was carrying the highest current. A series reactor was added to the middle leg of the generator-bus assembly. The reactance was adjusted by changing the length of punching stack until currents in the three bus legs were equalized, and the generator could then be operated a t full load without overheating. Conductor spacing can be arranged during installation or design of low-


788

voltage systems to eliminate unbalanced reactance by one of four methods shown in Fig. 15.24: 1. Use of multiple three-conductor cables 2. Paired-phase bus arrangement 3. Triangular spacing 4. Bus conductors transposed to cancel unbalanced reactance

LARGE 2400-VOLT SYSTEMS

( a ) THREE- CONDUCTOR CABLES C A

A B

II

II

(b) PAIRED- PHASE SPACING

B

C

i

I (c)

TRIANGLE SPACING

A

(d)

TRANSPOSED SPACING

The same principles apply to large 2400-volt systems which are t o be expanded as apply to large 600-volt systems previously discussed. Whenever a higher voltage is indicated for expansion of an existing 2400volt system, the higher voltage should be, in general, of the 13.8kv class, rather than of the 4160or 6900-volt class. However, in some cases, like that shown in Fig. 15.25, an existing2400-volt system can he expanded more economically a t 4160 volts than a t 13.8 kv. I n this example, so that these conditions will apply to the general case, major turbine-generators should he readily reconnectiblefrom 2400 volts delta to4160 volts Y. All cables and lines should he insulated for 5 kv. The major portion of the loadshould he fed through transformer banks consistingof either single-phase transformers or reconnectihle three-phase transformers which can have the primaries readily reconnected from 2400 volts delta to 4160 FIG. 15.24

Conductor orrangemenh to minimize or eliminate unbalanced reactance in three-phase bus runs.

790


volts Y. I n most of the practical large 2400-volt systems as is the case with the one shown in Fig. 15.25, there may be some 2400-volt load such as main pumphouse motors and power-station auxiliaries. This will necessitate retaining some 2400-volt power or rewinding the motors for 4160 volts and possibly purchasing new starters. Rewinding motors for higher voltage may cause some reduction in horsepower rating. Therefore, it is desirable to maintain 2400 volts for these motors. I n the example shown in Fig. 15.25, it is desired to add a 9375-kva turbine-generator. The short-circuit duty on the existing bus is considerably in excess of the interrupting rating of the existing circuit breakers. Also, this switchgear is obsolete from the standpoint of safety and operating convenience. Therefore, when the new turbine-generator is added, all the powerhouse switchgear should be replaced. This could be done a t 2400 volts, as shown in Fig. 15.26. The cost of all new switchgear for the new 9375-kva generator and its feeders, and the other generators and their feeders, is approximately $235,000. The reactors and bus between the new generator and the switchgear are included. Existing generators and the primaries of the existing stepdown transformers, particularly banks of single-phase transformers, can generally be reconnected for 4160 volts Y from 2400 volts delta without major construction changes in the apparatus. Therefore, the system can be readily expanded a t 4160 volts, as shown in Fig. 15.27. Since there is some 2400volt load, such as in the powerhouse auxiliaries and at the pump house, the 2500-kva generator will be left a t 2400 volts and connected to the 4160-volt system through a 2500-kva transformer. New 50-mva switchgear ill be purchased for this bus and new 250-mva switchgear for all 4160-volt generators and feeder circuits. The approximate cost of switchgear and reactors i n d 2500-kva transformers is $207,000. It is assumed that reconnection of existing transformers will cost about $14,000. Adding this to the cost of the new switchgear, the total cost of the 4160-volt scheme (Fig. 15.27) is $221,000. This represents a saving of about $14,000 compared with making the expansion and modernization program a t 2400 volts. Even if the existing 2400-volt switchgear was not replaced, and only switchgear for the new 9375-kva generator plus synchronizing bus switchgear and reactors for the other generators were purchased, the cost would be approximately $179,000. This would compare with $193,000 for making the expansion and modernization program a t 4160 volts. This contemplates using existing switchgear for the 2400-volt bus in Fig. 15.27. There will be additional economic gains in the future because the next 9375-kva turbine-generator can be added a t 4160 volts, for about $50,500 (switchgear, synchronizing bus, reactors, etc.). The snitchgear, reactors, etc., required to connect the future turbine-generator t o the bus a t 2400

M O M R N l Z A T l O N A N D EXPANSION OF EXISTING POWER SYSTEMS

a

a

I

4

I 0 U

w

791

I.C.

CIRCUIT

7 5 0 0 KVA

2 5 0 0 KVA

R E M A I NS IENCGT I 2 O4 N0 0 VOLT

TRANSFORMER 2500 KVA

N E W 5 0 MVA 1.C.SWITCHGEAR

7500 KVA

FIG. 15.27 One-line diagram of the system shown in Fig. 15.25 modernized and expanded a t 4160 Volb with one generator remaining a t 2400 volts. Buses tied together through bur-tietronrforrner.

N E W 4160 V O L T SECTION

-REACTORS

M A I N GENERATOR CiRCUiT BREAKERS

9 3 7 5 KVA

3

z

Y m

<

Y)

m

m

s

B

0

?!

m

H

P

m

0

z

b

P

-z

zII

0

s

Y


793

volts would cost about $88,000 (same circuit as for 9375-kva generator, Fig. 15.26). There is also a saving in the distribution system because it costs less to transmit power a t 4160 volts than a t 2400 volts. Limits of Size of Sources in 2400-volt Systems. The limits of the size of generators that can be used in the system shown in Fig. 15.26 are 7500 kva when the single synchronizing bus reactor is used and 15,000 kva when the duplex reactor is used. The limits of generator sizes are based upon two factors: (1) not requiring circuit breakers larger than 2000 amp (the largest metal-clad circuit breaker available) and (2) not exceeding the interrupting rating of 150-mva circuit breakers (the largest metal-clad circuit breakers for 2400-volt service). In many cases, transformers as well as generators may be used to supply power to a large 2400-volt system. The transformers may be treated just like generators. The circuit arrangements of Figs. 15.26 and 15.27 may he used, substituting the supply transformers for the generators. The limits of transformer kva, in some cases, may be lower, because of lower reactance than generators; therefore, they may produce more short-circuit current per kva of rating. While the addition of reactance in a system lowers the interrupting duty on circuit breakers, it does not lower load currents. Therefore, the kva limits of sizes of generators or transformers per bus given in the foregoing are maximums primarily because of the load currents approaching the largest ampere ratings of the circuit breakers available. However, if more reactance is added to reduce the interrupting ratings, the cost of the circuit breakers required is less. While additional reactance does have a beneficial effect in reducing short-circuit currents, it also has a detrimental effect. It causes greater steady-state voltage drop, particularly under operating conditions where one generator or transformer is out of service. It causes greater voltage drop when starting large motors and may interfere with the transfer of kvar from one generator bus to another. Particularly where low power-factor loads (0.9 or lower) are encountered, careful study should be made before adding reactance in a system in excess of the following: 1 . Generator series reactance to increase generator circuit reactance above about 12 per cent for generators having 10 per cent subtransient reactance X y or less 2. Synchronizing bus reactance which limits the short-circuit current contribution to a given bus to be less than about 75 per cent of the shortcircuit-current contribution of the largest generator on a given bus It is beyond the scope of this book to outline in detail the reasons for these general limits. They are based on experience as to stability, transfer of kvar, and voltage drop during normalandemergency conditions, etc.

794


EXPANSION OF A 2400-VOLT SYSTEM AT 13.8 KV

Figure 15.28 shows an existing 2400-volt system typical of that which might be found in a paper mill: A new 7500-kw 0.8-power factor 9375-kva turbine-generator and 9000 kva of load are contemplated, 2000 f t from the powerhouse. The short-cirrriit current of the esisting system, without the new generator, is i n escess of the 150-mva interrupting-rating switchgear employed on the 2400-volt rircuits. The addition of the new generator would merely exaggerate this condition. To add the new generator a t 2400 volts and keep the short-circuit current within the rating of the 1.50-mva switchgear will require a circuit like that shown in Fig. 15.26. Regardless of voltage at which the generator is added, the switchgear for the esisting generators will have to be rcarranged as shown to the left in Fig. 15.20 to reduce short-circuit currents to within the circuit-breaker interrupting rating. Therefore, the cost of the rearrangement of this switchgear is omitted in the following discussion. The switrhgear and reactors for adding the new generator and 2400-volt feeder cable will cost about $206,000. The unit substations and 2300-volt motor starters are not included in the above comparison because they are required for either generator voltage considered here. If the gcnerator is added at 13.8 kv, the arrangement mould be as shown in Fig. 15.29. This arrangement, including the 10,000-kva bus-tie transformer, new 13.8-kv switchgear, 13.8-kv cable, two 2000-kva transformers for stepping down from 13.8 kv to 2400 volts for the 2300-volt motors, costs approximately $140,000. This represents a saving of 30 per cent in this case, compared with the expansion a l 2400 volts. Even if the load had been only a few feet from the ponerhouse, it would still have been 1875 KVA

5000KVA

LOAD

16250 K V A ) 17500 K V A )

f

AT 2400 VOLTS

7 2

AT 4 8 0 VOLTS

9375 KVA

FIG. 15.28 A large existing 2400-volt system in an industrial p l m t to which D new 9375-kvo source unit i s to be added.

5 0 0 0 KVA

U I

6 2 5 0 KVA

480 VOLTS

10,000 KVA

Q

7 5 0 0 KVA

480 VOLTS

E A C H CA0.E

500,000KVA.1 C

MOTORS 3000 K V A

2 3 0 0 VOLT

:fhT:ERS


6 6

Ir

I 3 8 KV

0

NEW 9 3 7 5 UVA

FIG. 15.29 One-line diagram of the system shown in Fig, 15.28 modernized and expanded with n e w source added a t 13.8 kv and existing 2400-volt bur section. alired and connected lo synchronizing bur to keep short-circuit currents within the ratings of the 2400-volt rwitchgeor.

I

OLD OIL +CIRCUITBREAKERS

1875 KVA

796

MODERNIZATION AND EXPANSION

OF EXISTING POWER SYSTEMS

slightly less expensive to make the new addition a t 13.8 kv. Assuming no saving, or even a slightly greater cost on the original 13.8-kv installation, because of the fact that on the first step of changing to 13.8 kv it is necessary t o purchase a bus-tie transformer, the next 9375-kva generator t o he added a t 13.8 kv will represent a real saving since it is merely necessary t o buy a generator circuit breaker at about $21,000 to connect to the existing 13.8-kv bus. If the next generator is added a t 2400 volts, a complicated duplex reactor arrangement for the 9375-kva generator shown to the left in Fig. 15.26 is required, costing about $88,000, or $67,000 more than the 13.8-kv arrangement. The advantages of the 13.8-kv system are (1) less costly expansion and (2) greater kva can be carried per circuit. Therefore, fewer circuits are required and the cost of transmitting power is reduced. There is no change in utilization voltage when 13.8 kv is adopted as a generating voltage. Motors are still operated a t 2400 or 480 volts.

OLD

NEW

FIG. 15.30 Addition lo existing primary substation made by sectionaliring 4.16-kv bur. Short-circuit duty on bur A is 145 mva.


797

EXPANSION O F MODERN SYSTEMS WITHOUT CHANGING VOLTAGE

Many systems now use the highest practical primary voltage and the load-center system. Expansion is generally easy, i.e., more primary feeders and more load-center substations are added t o care for the new load in the plant. A problem may arise, however, in expanding the main substation or generating station.

69 K V

I

I

!

EACH

I

NEW

I I

I

SPLIT BUS HERE

250 MVA SWITCHGEAR

/

/ 150 MVA SWITCHGEAR

/

/

/ f----+ FUTURE EXTENSION

FIG. 15.31

One-line diagram showing the system of Fig. 15.30 expanded with all tronsformerr in pclrdlel and using synchronizing-bur arrangement to keep short-circuit currents within the ratings of the existing and new rwitchgeor.

798


9

9

10,000 K W 0.8 PF EACH

9 9 I38 KV

DUTY 415 MVA

500 MVA SWITCHGEAR

\

\

I

'N

"I

õõTa fb 2 . 4 KV

FIG. 15.32 Ona-line diogmm of rin exirting power ryrtam to which penerator ii to bc added at 13.8 kv.

2.4 KV

(I

new 12,SW-lna

Take, for example, the case of an airplane factory where power was purchased a t 69 kv and stepped down in two 5000-kva transformer banks to 4.16 kv. The 4.16-kv switchgear installed had an interrupting rating of 150 mva. I t was desired to add 7500 kva t o the capacity of the pnmary substation supplying this plant. I t was a simple problem, all except staying within the interrupting rating of the existing switchgear. New 250-mva switchgear was selected for the new substation capacity. Two choices were possible. The first was to not parallel the new transformer


799

with the existing 5000-kva transformer banks, in other words, operate with a split 4.16-kv hus. This led t o operating problems particularly in respect t o balancing the loads on the two bus sections. Had that not been a problem, the sectionalized hus shown in Fig. 15.30 would have been the least expensive method. The second method was to use a synchronizing bns, as shown in Fig. 15.31. This permits interchange of load between buses; however, there are limitations to this arrangement as pointed out in Chaps. 10 and 13. The reactors selected were of sufficient size to allow for future growth. Another example illnstrating the use of the synchronizing bus for reducing short-circuit duty is the expansion of the system shown in Fig. 15.32 to that shown in Fig. 15.33. 12,500 KW

6 0 0 0 KW E A G

0

.

*---i'

TO FUTURE UNITS

--

SYN EUS

f

TRANSFERREO FROM BUS N0.I BUS N 0 . 2 TO BALANCE ÇYSTEM

U

FlG. 15.33 One-line diagrom of the ryrtem rhown in Fig. 15.32 with the new 12.500-kva gwierator added on m e new 13.8-kv bur (No. 21. Only one rynchronizing-bur ieoctor (No. 11 is added now. Provirion for rynchroniring-bur reactor (No. 21 i. made to allow for fuhire expanrion. When reoctor (No. 2) is added, D breaker will replaice durnrny breoker at A.

800

MODERNIUTION AND EXPANSION OF EXISTING POWER SYSTEMS

SUPERIMPOSING A HIGHER VOLTAGE “BACKBONE“ OR GRID

A s mentioned earlier in this chapter, it is sometimes desirable to superimpose a higher voltage grid for distribution of power to load-center areas. Some of the foregoing examples, such as those in Figs. 15.5, 15.7, and 15.11. show how higher voltage “backbone” distribution systems were superimposed on a large 600-volt system. These are to be contrasted with the examples of Figs. 15.15, 15.19, and 15.27, where an extension was made at higher voltage but the old part of the power system was left substantially as is. Following are three examples further to illustrate this principle and the problems involved. In the first case, the plant was supplied a t 13.8 kv and all power was stepped down to 2.4 kv a t one location. The total load was 15,000 kva, and most of the power was used in segregated areas, Fig. 15.34. Consideration was given to splitting the 2.4-kv bus. However, that would not reduce interrupting duty to within the interrupting rating of the 2.4-kv circuit breakers. More load was to be added in the various mill areas, which would still further aggravate the problem of load balance on the three bus sections if the bus were split. More supply transformer capacity would soon be needed. The 2.4-kv cables to the mill areas were already overloaded. It was decided after considerable study that the best solution would be to use 13.8-kv distribution as a “backbone” system to supply the existing 2.4-kv distribution system in the mill areas through transformers. The transformers belonged to the power company, and they could use them elsewhere. The new power distribution system is shown in Fig. 15.35. The system change mas made on a step-by-step basis. Because of the large number of large 2.3-kv motors, the power on the first step was all transformed to2.4 kv a t themill areas. Inlateradditions thedouble transformation will be eliminated; 480 volts mill be transformed directly from 13.8 kv for the 440-volt loads. The 2.4-kv distribution system in the mill areas was practically unchanged. The 2.4-kv distribution center in each mill area is now supplied by a local transformer instead of by long 2.4-kv feeders from the main substation. Had this plant been served a t some voltage above 15 kv, say 69 kv, new transformers 69 kv to 13.8 kv would have to have been added. In another case, Fig. 15.36, a chemical plant had one area served a t 4160 volts stepped down from a 33-kv utility system. An expansion of similar proportion was planned about 135 miles away. T o expand the existing substation meant complicated arrangements of reactors and switchgear to keep the short-circuit duty within the rating of the existing 4.16-kv switchgear. Furthermore, the cost of transmitting 7500 kva at 4.16 kv for 146 miles is high. Power was already purchased a t 33 kv, and the substation belonged to the industrial plant.


801

Study revealed that the most economical arrangement would be to use 33 kv as the “backbone” voltage for carrying power to the load areas of this plant. Hence, the new area was served a t 4160 volts fed from a new 13.8 KV FROM UTILITY

UL 2.4 KV

I/

T

3- 5000 KVA TTRA“B,NK,”MER

s,

MILL AREA ‘I

I

8”

MILL AREA

“c

BUS

MILL AREA I‘

D

I_

1 1 MILL AREA

FIG. 15.34 Schsmotiç me-line diaaram of an exirting power ryrtem in

‘*E

‘I

BUS

0

Iarge mill.

802


$ 4 f Y

Y

2 I 2.4 KV

MILL~EZT

NEW 13.8 KV METAL CLAD

sw ITCHGEAR RATED SOOMVA

$ 2 . 4 KV MILL AREA "B"

MILL AREtS " 8 " a "c CLOSE TOGETHER 12T;EKv

STATION

T

MILL AREA"D" SUBSTATION

v

$Y4 T

ILL AREA"E" SUBSTATION

FIG. 15.35 One-line diagram showing the power system of Fig. 15.34 with 13.8-kv (backbone) distribution system used to feed existing 2.4-kv mill areas through local 13.82.4-kv molter substations.


FIG. 15.36 Chemical plant power system with widely segregated lood areas. area to be added 136 miles from existing load area.

803

New lood

804


33 K V FROM UTILITY

kvk l 4 4

EXISTING AREA

NEW AREA

FIG. 15.37 Power system of Fig. 15.36 exponded using 33 kv voltage for carrying power to load areas.

(IS

Ibackbonel primary


805

substation a t the new area, as shown in Fig. 15.37. Future plans contemplate similar expansion using 33 kv as the primary or “backhone” system voltage. This technique applies to strip mines or any area where distances are great and sizable loads are used a t each location. Primary or “baokbone” voltages as high as 220 kv have been used in this manner for plants in this country. A third case is typical of many of the large steel mills. Here primary voltages are often 6.9 kv. As plants have grown and the areas expanded, a point is reached where 6.9 kv becomes too low a voltage for economical

UTILITY

COLD M I L L

MERCHANT MILL AREA FIG. 15.38

Schematic one-line diagram of a large steel mill

6.9-kv generation and dirlribution.

(IS

originally operated wifh

006


and good practice in distributing such large blocks of power over such wide areas. A typical case is that shown in Fig. 15.38. Here expansion added considerably more power. Several voltages were studied, i.e.,

NEW69KVBUS

p1 /

0

p I"yI

4, I

REGULATING AUTO TRANSFORMER

\

NEW BLAST FURNACE AREA SUBSTATIONS

NEW COLD MILL AREA SUBSTATION

NEW HOTMILL AREA SUBSTATION

FIG. 15.39 One-line diogrom of system of Fig. 15.38 with 69-kv [backbone) system used for supplying 6.9-kv power at individual mill oreas.

MODERNIZATION AND EXPANSION O f EXUTNG POWER SYSTEMS

807

expansion a t 6.9, 13.8, 33, and 69 kv. The latter three did not show a wide divergence in over-all cost, but all three, i.e., 13.8,33, or 69 kv, were materially less expensive than expanding the 6.9-kv system. Local preferences and the fact that the utility line was 69 kv dictated 69 kv as the main transmission voltage. At the main mill areas the power was stepped down to 6.9 kv. The essentials of the new system with increased generation and load are shown in Fig. 15.39. Theexistinggenerators Nos. 1 and 2 and the powerhouse switchgear were left as is to serve existing 6.9-kv local load near the powerhouse. The existing 69-6.9-kv transformer was left as is except that it was connected to the new 69-kv bus instead of directly to the utility line. Existing generator No. 3 was removed from the existing 6.9-kv bus to reduce short-circuit duty on the 6.9-kv switchgear. Number 3 and the new No. 4 generators were connected directly to the new 69-kv bus through transformers. A 69-kv loop was used t o transmit power from the new 69-kv powerhouse bus to the new mill substations as shown in Fig. 15.39.

Chapter 16

by William B. Wilson

Steam and Power Generation TRENDS IN THE USAGE AND SOURCE OF POWER FOR INDUSTRIAL PLANTS

During the past 10 years, the total use of electric power in industry almost doubled and the use per unit of industrial production increased more than 25 per cent. Except during a national emergency, such as World War 11, industrial plants used 55 to 60 per cent of the total power used in the United States. Both the industrial power plants and the utility plants utilize steam turbine-generators, combustion-gas turbine-generators, water-wheel (hydro) generators, and internal-combustion engines for electric power generation. I n 1952, generation of the 463 (399.2 by utility and 63.8 by industrial plants) billion kwhr of electric energy in the United States was divided as shown in Table 16.1. Installed capacity in 1952 for the different plants was divided as shown in Table 16.2. Based on information in Tables 16.1 and 16.2, the average “plant factor” for both utility and industrial generating plants is approximately 50 per cent. I n making economic studies, it is reasonable t o expect that the “plant factor” for new plants will be appreciably higher than the over-all average figures indicated in the tables. I t must he recognized that the figures for installed generating capacity used in Table 16.2 include not only the generating capacity in normal use but also obsolete equipment which is retained for some very infrequent emergency operating condition. Much of this obsolete equipment does not have sufficient service reliability or operating efficiency to justify its use for normal operation. 808

....

s'v8

"

L'Z b'OZ t' 19

I I

................................ 4 p o d m Bu!io~ouoB. polloisu! y n p (us, ,ad z.z8 ............................... ID101 ...................... s"!qrq "o!l.nquo, p"IDlYl ............................wgs"quO3 SiuDldoApAH .. .. f.. ..". 'IIuqd aY!qrq wDs#s

v.Oz L'bS

"

"

"

'

pi01

SDB

'

-1-1-

0'001

....

....

Z'98

.... ....

5' 1

L'EC 8'VL

"O!I

I

I

I

I

ID!4mp"l

I

N O U V M N 3 9 WMOd (INV WV31S

608

STEAM AND POWER GENERATION

810

realizing economies in initial cost, maintenance and operating costs by installing larger generating units. The average kw rating of turbinegenerator units being sold for industrial power plant applications today is approximately 7500 t o 10,000 kw-just about twice the average rating of units being sold 10 years ago. I n many areas, “outdoor” power plants are being used to reduce the initial power plant investment costs. The trend to higher initial steam conditions has increased the “ by-product” power which can be generated by plants supplying steam for process use. A large number of industrial turbines are being sold for initial steam pressures above 800 psig today, and many are utilizing initial pressures higher than 1000 psig. Steam turbines operating in our utility plants are designed for initial pressures up t o 2400 psig and initial temperatures up to 1100 F. Units now on order are being designed for initial pressures up t o 4500 psig and initial temperatures up to 1150 F. The average initial steam conditions of industrial power plants as well as utility power plants have increased rapidly in recent years, and it is expected that this trend will continue. A partial list of industrial turbines with high initial steam conditions and of the larger units installed in industrial generating stations is given in Table 16.3. TABLE 16.3 Turbine capability roling, h

110,000 110.000

44.000

43,750 41,000 31.250 18,750

Partial List of Industrial Turbine-Generators

Steom conditions

Steam conditions

1200 lb-725 F 1 in. 1200 lb-900 F-1 in.

850 lb-900 F-155 in. 1250 lb-900 F-235 Ib 725 lb-725 F-165 Ib 1250 16-925 F-i 35 in. 1250 lb-925 F-150 I6

The requirements of each application where power is to be generated must be carefully considered. Technical data included in this chapter should assist the designer in the selection of the proper prime movers for each industrial power plant application. Electric power is very essential t o the economical operation of all


81 1

industrial plants. In many plants, a power interruption for only a few minutes or even fractions of a second can cause many hours of lost production in addition to possible damage to process equipment. In many plants the question of purchased vs. privately generated power must be considered. What then are the factors that determine whether power should be generated or purchased? For larger plants, the availability of low-cost power from the utility or of low-cost fuel of suitable quality for a local generating station are important factors that should be considered prior to the purchase of the plant site. PURCHASED VS. GENERATED POWER

Shall all electric energy be purchased, or shall generating facilities be provided so that a part or all of the electric energy can be generated within the plant? It is usually the use of process heat and not the magnitude of the electric power load that is the most important factor favoring industrial power generation. Reliability, provision for future plant expansion, predictions of future changes in plant processes, personal preferences, etc., all enter into making the decision, but like many such decisions, it is usually a simple matter of economics. Why Generate? There are many conditions where local power generation in the industrial plant is economically sound. These conditions include the following: 1. Industries using large quantities of heat energy, usually process steam, sometimes find it economical to generate their own by-product electric power either by steam or gas-turbine-driven generators. These facilities supply process steam and elcctric power simultaneously and a t high plant cycle efficiency. It is important to note that in these cases it is not the use of electric power but the use of heat energy, usually in the form of process steam, that determines the economics of power generation within the industrial plant. Typical of such industrial plants are paper mills, petroleum refineries, chemical plants, etc. Electric power generated by prime movers supplying heat for process use is usually called “by-product” power. Over-all power-plant efficiency of such plants can be as high as 70 to 80 per cent-this is two to three times the efficiency obtainable in the average power plant operated for the generation of electric power only where all steam must be condensed and heat is not required in plant processes. 2. Where large quantities of low-cost waste or by-product fuel are available, the industrial plant can sometimes generate low-cost electric power, particularly where large, modern, efficient prime movers are selected. Typical of such industries are large integrated steel mills which burn the excess blast-furnace gas for the generation of electric power.

812


3. Some plants such as mines and mine products processing plants are often located in undeveloped isolated areas of the country. I n such areas electric utility systems may not be developed t o care for such remote loads, and it may be less expensive to install generating facilities a t the mine or processing site than t o develop and extend for such long distances the electric utility system t o care for such a load. Often there is a speculative nature about such ventures or only limited quantities of raw material are available which may not justify a large expenditure on the part of the utility because such an investment may be lost should conditions change or the venture not develop as planned. 4. I n some cases where loads are very large (50,000 k w or more), load factors are high, and low-cost fuel of the proper type is available, power generation has been employed by the industrial plants when power was not readily available a t low cost from the electric utility. Typical of such cases are the aluminum-reduction plants, chemical plants, steel mills, and others requiring large quantities of electric power. I n some industrial plants where the consequences of a power outage are very serious and where electric power lines from an outside source are subject to frequent disturbance by lightning or other causes, the industrial customer may sometimes consider the generation of a part of the electric power required. This might be done to minimize the possibility of an interruption of power t o essential circuits even though the cost of generating poxx-er was greater than would he the cost of purchased power. In most areas, the reliability of power supplied by the utility company is adequate for Any industrial requirement. Where any of the above conditions exist, it is well t o study the cost of generating a part or all of the power requirements within the industrial plant. The most economical arrangement is often t o generate only such power as can be made as by-product power from the process heat or as can be generated from by-product fuel and to purchase the remainder. Electric power utilitieshave been very progressive in their planning of new generating facilities. They have installed large generating units of modern design, keeping step with metallurgical developments which permit higher steam temperatures and pressures resulting in improved cycle efficiency. Many large steam-turbine generating units are designed for steam reheat t o further improve cycle efficiency. Some industrial plants are or can be located adjacent t o a utility power station. I n such instances, it may be found that the requirements of the industrial plant for both electric energy and steam can be supplied from the utility plant most economically. If by-product or waste fuel is available from the industrial process, that fuel may be sold by the industrial t o the utility for use in the nearby utility power and steam generating


813

facilities. There are a few cases where electric utilities are supplying both process steam and electric energy t o industrial plants, and there will undoubtedly be more. These combinations have been very successful from the standpoint of both the industrial and the utility. I n effect, such an arrangement makes it possible for the utility t o make some by-product power to further improve its plant-cycle efficiency. Because of the efficient operation of the electric utility systems, their rates are usually very attractive. Small industrial plants having little or no requirement for process steam usually find that the electric utility can generate, transmit, and sell electric power for less than the cost of generation in the industrial plant using steam turbines, gas turbines, diesels, or similar types of prime movers generating electric power only.

FACTORS TO CONSIDER WHEN PURCHASING POWER

It is not intended t o discuss rates or the contractual phases of purchasing power. Where power is t o be purchased from the localutility, there are certain phases of an engineering nature which should be considered before the plant site is bought and before cost estimates of the electrical system are made for appropriation purposes. SIZE OF LOAD VS. POWER AVAILABLE

Among the points to be considered with the utility are the amount of power available or that can be made available. Is there enough available power in the area to support the plant which is intended to be built? Where only a few hundred or a few thousand kilowatts are required, power supply is generally no problem in picking the plant site as such quantities of power are readily available almost everywhere that industrial plants of these small sizes are generally located. However, when electrical demands are large, i.e., several thousands or tens of thousands of kilowatts, the exact plant location may be influenced by availability of electric power. When plants have large electrical demands, extra facilities may often have to be added to the electric utility system, and therefore there are many problems of mutual interest between the industrial planner and the local utility. ALWAYS GIVE THE UTILITY THE FULL STORY EARLY

Particularly if the plant load is sizable, give the electric utility the full story early in the consideration of a plant site. The primary objective of

814


the electric utility is to serve their customers and serve them well. They all strive to do this, but they can do this only if given the fullest cooperation, particularly in the early planning stages of selecting a plant location. The local utility, t o serve the plant best and on time, should know what the expected demands are, what the load swings are apt to be, and what kind of load factor and what type of load the plant is going to impose upon their system. Modern electric utility power systems are so adequate and so flexible and so thoroughly cover the country that they can meet most of the demands of even the larger new plants. However, it is always well to check with the electric utility first t o be sure that your plant loc&tion is proper from an electric-power-supply standpoint. POWER SUPPLY FOR CRITICAL PROCESSES

When considering the plant site for a critical continuous-process plant, one should study the utility-system connections t o be sure that there is or can he made a reliable source of power. Remote parts of some electric utility systems are fed by single lines, which may not he reliable enough for many manufacturing processes. Therefore, if the plant is t o be located in such a remote part of the system, a check is in order. Also eheck the substation and transmission interconnections in the immediate area t o see that there is a reasonable degree of assurance of a continuous supply of electric power to the plant during normal or emergency outages of a part of the utility power-supply system. VOLTAGE REGULATION

Sometimes starting large motors becomes a problem when a plant is located a t the end of a very long light line. There may be so much voltage drop that the larger motors cannot be started. In other cases there may be so much voltage drop associated with the starting of the larger motors that other motors already running and the lighting of the plant will not perform properly during the starting period of the larger motor. Voltage regulation or spread due t o normal steady-state load can usually be taken care of by feeder voltage regulators or load-ratio control in the transformers. I n other cases the normal voltage regulation or spread of the utility supply line may he, because of requirements for efficient operation of the over-all utility system, such as to require voltageregulation facilities in the industrial plant. I n any event, a check should be made of the voltage spread due t o normal load in the plant or due to utility-system operating practices to see if it would be necessary to install


815

voltage-regulating facilities. If such facilities are required, be sure that they are included in the equipment when electrical specifications are drawn up. If there is any question as t o whether or not they are needed, it is better that the voltage-regulating facilities be included as manufacturing processes are constantly requiring closer and closer voltage spread t o perform the manufacturing functions effectively and efficiently. MASTER-UNIT SUBSTATIONS

Check t o see what voltages are available from the utility. If the voltage at mhirh the plant will be served is above 15 kv, a master-unit suhstation may be required. If the voltage availahle from the utility is 15 kv or below, that voltage may be used directly for distribution of power within the plant (see Chap. 10). The voltage at which a given plant will be served in the average case depends upon the local practices of the electric utility and upon the size of the plant contemplated. If power is served to the plant at volbages above 15 kv and a masterunit substation is required, there is the problem of who should own the mast,er-unit substation. I n some areas the utility policies are t o supply such substations, or at least the transformer section of the substations. In ot,her areas the supplying of the snhstations is optional, and in still other areas it is required that the industrial plant supply the master substation complete. If a master substation is to be installed for serving the plant and no other circuits on the utility system, then it is generally advantageous, from the standpoint of flexibility of operation of the industrial plant, for the industrial plant t o own that master substation. The plant operators then have control over its rating, design, and its operation and can integrate it into the plant power distribution system more easily than where substations may be owned by someone else. SEPARATE UTILITY A N D INDUSTRIAL PLANT OPERATIONS

It is highly desirable t,hat rircuit breakers he employed between the connertion t,o the utility system and the industrial power system and that the circuit arrangements be such that the operation of the utility system from a switching standpoint is independent of the operation of t,he industrial plant. In this way there is a maximum of flexibilit,y obtained for the utility and t,be industrial plant, as earh may operate their respective systems with the greatest degree of freedom without interfering with each other from a circuit,-svitching standpoint. Typical one-line diagrams of utility-industrial interconnections are shown in Fig. 16.1. See Chap. 13 for further details on primary-system design.

816

STEAM AND POWER

GENERATION

1 1.

UTILITY SUPPLY LINES ONE OR MORE

Y

OR FUSES

Y

x $Y

$Y I

J

I

PLANTFEEDERS

PLANT FEEDERS

,-$ 4 I

6 4 (b)

(0)

UTlLiTY SUPPLY LINES

*

I

UTILITY SUPPLY LINES

BREAKERS OR FUSES

,

,

PLANT FEEDERS

PLANT FEEDERS (C ) FIG. 16.1

(d 1 Typical utility-indurtriol interconnections.

PARALLEL OPERATION OF GENERATING FACILITIES

When the industrial plant generates part of its own electric power and purchases the rest from the utility system, it is always highly desirable to


817

operate the generating facilities in parallel t o obta.in the most flexible operation for both the utility and the industrial plant. When it is desired to generate only part of the power and purchase the rest, it is important in the early stage of the negotiations to check to see whether or not the utility will permit parallel operat,ion of the industrial generating facilities nith the utility system. There are generally no serious terhnical problems involved in such parallel operation. However, there may be policy problems which are extremely important. After it has been derided that there will be parallel operation of the facilities, there are certain technical problems which should be considered early in the negotiations and particularly early in the design stages of the power system. In areas of severe lightning outages, the reliability of the industrial power system may be materially affected by outages on the utility supply line because of lightning. I n such cases, particularly where continuous processes are involved, methods of quick disconnection of the two systems during such disturbances should he devised. I n any case, when the two systems are separated, adequate facilities should be provided for resynchroniaing, and these points of resynchronization should be limited to a very minimum. There are always problems of control of the division of kvar between the local plant generators and the utility systems. There may he powercontract clauses that would make it desirable t o limit the kvar imposed on the ut,ility system by the industrial plant to a very minimum. It is highly desirable t o permit the greatest degree of freedom of operation in so far as kvar control is concerned. To obtain the maximum degree of freedom, always include voltage-regulating facilities iu the tie t o the utility system. This may be in the form of load-ratio control in the step-down transformers or in feeder voltage regulators in the utility supply line. Such facilities are generally furnished by the industrial plant. I n other cases power-limit clauses may stipulate a mavimum kw demand that should he imposed on the utility system. I n such cases, turbine-load regulating equipment is necessary. RELAY COORDINATION

While not an essential for consideration in negotiation stages, it is certainly important in the early stages of system design to study any requirements of the utility with regard t o relaying. These requirements must he coordinated with the relaying system in the industrial plant t o assure selective operation and proper clearing of any faults that may occur in the industrial system. For further discussion of the relaying problems, refer to Chap. 9.

818


COMPARISON OF INDUSTRIAL POWER PLANT CYCLES

I n all thermal generating stations today, even including proposed electric generating plants using atomic energy, the cycles are similar. I n the first step, the energy of the fuel is converted into heat energy by a chemical process. I n the case of our more common fuels, this is the process of combustion. In the case of atomic fuel, it is a reaction which takes place in an “atomic pile.” This heat energy may expand the gases in the cylinders of an internalcombustion engine t o produce mechanical work direct. It may also be transferred in boilers or combustion chambers t o fluids such as water (steam) or air a t some elevated pressure. This energy can then be converted into mechanical work by expansion t o a lower pressure through a turbine, engine, or similar prime mover. The mechanical energy is then converted into electrical energy by an electric generator. PLANT THERMAL EFFICIENCY

Of primary concern is the over-all plant-cycle thermal efficiency. There are definite limitations to the efficiency of different plant cycles. T o consider the cost of electric energy or heat energy separately is not sufficient-it is the over-all fuel and operating dollars that are important. For plants using the same type of fuel and generating all heat and electric energy requirements within the plant, fuel dollars are directly proportional t o over-all plant thermal efficiency. Over-all plant thermal efficiency is very useful for evaluating various plant cycles and is defined as the Total available and useful output Total actual input This concept of plant thermal efficiency clearly recognizes both heat energy and electric energy; the useful electric energy can be converted into an equivalent heat basis. High thermal efficiency is a very important factor in over-all ope!ating costs. Power-plant-cycle thermal efficiency for different types of power plants is covered in more detail later in this chapter. It will be noted, in plants requiring both heat energy and electric power, that each type of prime mover can operate a t optimum power-plant-cycle efficiency if the ratio between heat energy and electric power requirements matches the characteristics of the prime mover. This ratio differs widely for different prime movers as indicated by the curves in Fig. 16.2. The requirements for heat energy and electric power vary widely in different plants, and the different types of prime movers operate a t


819

optimum efficiency for different, definite ratios between plant requirements for beat energy and electric power. This being the case and because of the importance of the particular plant requirements for both heat energy and electric energy, application of the proper prime mover for each job considered is of prime importance. The curves of Fig. 16.2 show the by-product electric power which can be generated by steam turbines and gas turbines when supplying 100,000 lb of steam per hour for process a t different process pressures. These curves are based on the following: 1. Noncondensing or automatic extraction-type steam turbines extracting and/or exhausting all steam to process. 2. Gas turbines which are used t o supply steam to process by exhaustheat recovery boilers utilizing heat from the gas-turbine exhaust gases. 3. For certain applications, such as drying heavy fiberboards or the kiln drying of lumber, the gas-turbine exhaust gases may he used direct. For these applications, temperature and humidity of the drying gas can

PRESSURE OF STEAM SUPPLIED TO PROCESS PSlG FIG. 16.2 Electric power which can be generated by combustion gar turbines and steam turbines when supplying 100,ooO Ib of steam per hour for process.

820


be controlled by t,he admission of atmospheric air and moisture as may be required for the particular application. These units all operate a t their maximum cycle efficiency when the prime mover can supply all electric-energy requirements and a t the same time supply the heat energy required for process and building heating. The generation of power in this manner is often referred to as “by-product” power, and the over-all power-plant-cycle efficiency ranges as high as 70 to 80 per cent. Based on generat,ion of “by-product” power, these curves show that steam turbines or, as explained on page 867, gas turbines with supplementary fired exhaust-heat recovery boilers may he considered when the ratio of electric power to process steam is lorn-say, 2500 to 5000 km per 100,000 Ib per hour steam flow which is 20 to 40 Ib of process steam per kilomatthour. Gas turbines with straight exhaust-heat recovery boilers may be considered when 8000 to 10,000 km is required per 100,000 lb per hr steam flow which is 10 t o 12 Ib of proress steam per kilomatthour. I n the intermediate range of process-steam requirements, the condensing extraction turbine or the gas turbine utilizing exhaust-heat Fecovery boilers with some supplementary fuel firing might be selected. I n fact, some plant,s have found it desirable t o consider a combination of types of prime movers to give the highest over-all plant-cycle efficiency and greatest flexibility in operation. This r h o of process steam per kilowatthour varies widely in different plants, ranging from plants having no requirements for heat energy to plants requiring an equivalent of 40 Ib or more of steam per kilowatthour. Today, the average is perhaps no more than 20 or 25 Ib of steam per kilowatthour, and with the rapid increase in t,he use of electricity, the ratio is likely to decrease still more in the future. In many new plants and particularly in.t,heexpansion of existing plants, the ratio of process steam to electric power requirements may be quite low. The trend in industrial power plants is to higher initial steam conditions. This trend reduces the amount of steam per kilowatthour available when generating by-product power with steam turbines and tends to match more nearly the changing power and process-steam requirements.

STEAM POWER PLANT CYCLES-STEAM TURBINES First., consider the efficiencies, theoretical and actual, of a steamturbine power plant. TURBINE-GENERATOR EFFICIENCIES

The over-all efficiency of steam turbine-generator units-that is, the turbine-generator unit itself and not the turbine cycle-ranges from approxi-

FIG. 16.3 Mollier chart with methods of opproximating turbine expansion lines indicated.

I

822

L W i th

"m

1


o u Pro'iiss steam Requirements

t

e W i t h Process Steam Requirements

I

4

100

FIG. 16.4 Relative power-plont thermal efficiencies for various steam-turbine powerplant cycler.

mately 60 per cent for a 500-kw unit, 75 per cent for a 7500-kw unit, up to 80 per cent or more for units 15,000 kw and above. Steam conditions, type of turbines (condensing, noncondensing, automatic-extraction), and various other items also have minor effects on the efficiencies. For the small units (up to 1500 kW) an efficiency of 65 per cent may be used for preliminary estimating purposes. Except in isolated cases, the performance of turbine-generators in ratings 2500 to 15,000 km should have an error of not more than 10 per cent if an over-all turbine-generator efficiency of 75 per cent is assumed for preliminary estimating purposes. This does not mean that the turbine can take steam having a heat energy of, say, 1420.8 Btu per Ib (the enthalpy, beat energy, of steam a t 600 psig-825 F) and convert 60 to 80 per rent of this heat energy into electric energy. Because of the limitations of the steam cycle, if the above steam is expanded to 2 in. Hg abs (28 in. of mercury vacuum referred to a 30-in. barometer), through a condensing-steam turbine or any other steam-using prime mover, the energy available for use is only 1420.8 minus 916.5 (see points A and B , Fig. 16.3), or 504.3 Btu per Ib. The theoretical expansion line of a prime mover having 100 per cent


823

efficiency and this available energy is shown as line .4B on the Mollier chart, Fig. 16.3. This 504.3 Btu per Ib is only ahout 35.5 per cent of the total energy in the steam supply. More will be said about the Mollier chart later. Actual relative power-plant-cycle efficiencies for different operating requirements and various types of steam-turhine plants are shown in Fig. 16.4. THEORETICAL STEAM CYCLE

The cycle of a theoretically perfect turbine, using a surface condenser to condense the steam to supply boiler feed water for reuse in the cycle, is shown in Fig. 16.5. In this theoretical cycle, 847 Btu per Ib of steam condensed is rejected to tbe condenser circulating water a t a temperature helow 101 F (saturation temperature of the steam condensed a t 2 in. Hg abs). Only 69 Btu per Ib remains in the condensate ready to repeat its path through the cycle. As can be seen in this diagram, 59.7 per cent of TURBINE

100 %

600 PSIG-825F 1420.8 BTU PER L 100 %

EFFICIENT

*

AVAILABLE O!"

I

rC"

35.5 %

i

BOILER 100 7.

*

LO

I

2 IN HG ABS 101F OJC F -7,I DCO _I.v._. , LI> I B 64.5 O/. X

I

!I I

LtItNU:

-STEAM

-_--

WATER

NOTE: I - THEORETICAL CYCLE EFFICIENCY

2

- THEORETICAL

1420.8-316.5

= 1420,8.69

= 37.2%

STEAM RATE R)R TUREINE GENERATOR =

3413 BTu PER KW HR = 6 . 7 6 LB STEAM PER KW HR 1420.8 - 916.5 X FIGURES ARE PERCENT OF ENERGY IN STEAM SUPPLIEO TO TURBINE

FIG. 16.5 A rchernatic diograrn rhowing theoretical cycle of stroight condensing steclrnturbine plant operoting with steorn conditionr Mx) prig, 825 F, 2 in. Hg abr.

824


the heat errergyin t,he steam supplied to the turbine is lost in the condenser circulating vater. Assuming that a theoretical steam plant with steam conditions of 600 psig-825 F-2 in. Hg could be built which had no auxiliary losses and wit,h hoilers, turbines, and generators having 100 per cent efficiency to convert all this available energy into useful output, the cycle efficiency would be Total available and useful output Total a d u a l input or, for this theoretical case (see Figs. 16.3 and 16.5),

The heat energy available in the cycle The heat added in the boiler which is 1420.8 - 916.5 504.3 _ _1331.8 _ 1C20.8 - 69.0

or 37.2 per cent

For a theoret,ical plant operating hetween initial steam conditions of 600 psig- 825 F- 1 in. Hg abs, t,he theoret,ical cycle efficiency would be 39.2 per cent. For steam conditions of 2400 psig-1000 F-1 in. Hg abs, the t,heoretical cycle efficiency mould be 45 per cent. For the same theoretical plant operating with initial steam conditions of 400 psig-750 F, the theoretical rycle efficiency would be 37 per cent with turbine exhausting at, 1 in. IIg ahs and 35 per cent with turbine evhausting a t 2 in. Hg ahs. These efficiencies are based on the assumption that the heat energy (847 Btu per Ih for the GOO psig-825 F-2 in. Hg abs turbine) in the turbine exhaust steam, rejected to the condenser circulating water a t the relatively low t e m p e d u r e of 101 F, is not used. If only two-thirds of this 847 Rtu could be used, then the theoretical cycle efficiency of the 600 psig-8825 F-2 in. Hg plant would be 79 per cent, or more than doubled. If all this heat energy in the exhaust could be used, then, of course, the effiriency of this theoretical plant would be 100 per cent. The chart, Fig. 16.6, illustrates the ut,ilization of heat energy in steam supplied t o the steam turbine for different cycle arrangements. THE MOLLIER CHART (ALSO CALLED "STEAM' CHART)

Vsing the turbine-generator efficiencies given on page 820, the actual performance (pounds of steam per kwhr) of different types of steam turbines for various steam conditions can be estimated very easily. The steam conditions a t the turbine exhaust or any selected extraction pres-

S T E M AND POWER GENERATION

nis s'ozw

825

In. Hg a br

~

~

1

200

1

1

400

1

400

0 5 10

Pdg

14.33 15.24

9.10 10.13 10.88

14.42

13.35

12.09

13.63 12.63 11.53

9.68 10.91 11.84

8.67 8.82 8.96

10.30 10.77

8.23 9.11 9.76

7.06 7.99 8.11

9.92 10.35

8.01 8.83 9.43

7.67 7.79 7.91

11.71 12.49 13.20

11.29

8.72 9.62 10.28

8.34 8.48 0.60

15.99 17.13

10.35 11.72 12.74

9.18 9.36 9.52

20 25

~~

111.2

I5

10.70 12.19 13.34

9.80 10.00 10.18

600

1 600

I

1

10.40 11.05 11.64

10.04 9,64

8.63 9.19

7.86

7.53 7.65 7.76

7.09 7.26 7.40

6.888

6.633

6.254

261.2

I 850

1

8.71 9.03

-

7.24 7.88 8.34

6.961 7.06 7.15

6.582 6.728 6.849

5.862 6.190 6.410

-

9.82 10.42 10.96'

-

9.31 9.83 0.29

- -

9.13 9.49

-

7.48 8.20 8.71

7.17 7.28 7.39

6.7& 6.92: 7.05

5.98< 6.341 6.571

297.7

- -

336.2

- -

Initial superheal

11.59 13.37 14.76

1.5

I

1

____

5 10

7.20 7.38 7.53

600

preuure. psig

Initial temperature, deg F

I

10.8910.10 11.14 10.31 11.37 10.51

7.37 7.55 7.71

280

500

3.5 4.0 4.5

8.09 8.31 8.50

I

1

7.83 8.02 8.19

8.52 8.76 8.98

194 115t.91 301.9

250

9.98 10.32 10.62

I

1

2.0 2.5 3.0

9.07 9.35 9.59

94

250

.o

9.32 9.61 9.87

84.2 1112.2/

150

1niti.l

1 850

8.81 9.28 9.71

8.25 8.55

6.981 7.49 7.91

6.726 6.811

6.636

6.284 6.419 6.532

5.611 5.917 6.122

I

1

8.50 8.94 9.34

7.97 0.25

6.676 7.24 7.64

6.431 6.520 6.600

6.097 6.223 6.332

5.454 5.747 5.944

372.7 1422.7

850

TABLE 16.4 Condensed Table of Theoretical Steam Rates*

6.872 7.31 7.60

0.5

-

~

houd pro,. *we

EX.

-

1

I 1200

!

1

I

7.67 7.92

1 1

6.493 7.01 7.37

6.269 6.350 6.425

5.958 6.078 6.179

5.633 5.017

5.350

7.31 7.52

6.75C 7.06

6.30C

6.10i 6.17: 6.24C

5.828 5.933 6.023

5.286 5.535 5.70C

287

2400

~

~

Ex-

Pig

20 25

I5

10

5

3.5 4.0 4.5

2.0 2.5 3.0

0.5 1.0 1.5

obs

In. Hg

I",*

p,*,-

haw

8.44 8.86 9.23

8.14 0.53 8.87

5 10

7.70 8.02

8.30

0

______ -

/I-

7.95 8.21

6.709 7.26 7.63

6.638

6.474 6.560

6.148 6.274 6.380

5.519 5.807 5.999

331.2 1381.2

1200

-

550 600 650

500

350 400 450

200 250 300

125 I50 175

..... ...., ....,

rom

:enan and E yes,

...., ...., ...., ...., ----

I

..... ..... ..... ..... ..... ..... ,

26.94 32.9 40.6

21.52 24.77 28.49

20.03 23.04 26.52

19.17 22.06 25.38

15.34 16.57 17.84

,

:heoretical S &rn Rate Tat

1

L-

-

...... 66.1 55.0 51.2 32.8 30.5 29.23 ...... ..... ...... 67.0 38.0 35.4 33.9 ...... ..... ...... ..... 44.4 41.4 39.6 ...... ..... 52.6 49.0 47.0 ...... ...., ...... .... 63.4 59.0 56.5 ...... . ... ...... ..... ..... ..... 70.1 .... . . .... ...... ..... ..... ..... .....

28.99 35.3 43.6

16.03 17.31 18.64

13.65 14.13

..... ..... I ::::: ..... ..... ..... ..... ..... ..... ..... ..... ..... .....

34.7 42.4 52.4

17.22 18.60 20.02

14.27 14.77

12.19 12.68 13.17

12.74 13.25 13.76

13.66 14.22 14.77 15.32 15.87

11.43 11.69

11.94 12.21

12.77 13.07

33.9 43.5 57.7

43.5

20.02 22.12 24.42

17.29 18.05

10.62 10.90 11.17

9.69 10.02 10.33

11.07 11.37 11.66

10.43 10.76

10.08

12.13 12.45

11.80

10.71 11.10 11.46

..... 62.3

51.1

21.56 23.83 26.27

20.11 21.12

...... 25.51 ..... ..... ..... ..... ..... 26.75 28.31 17.9 37.5 31.9 ..... ..... ..... ...... 31.4

..... ,.... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... ..... .... ..... ...., ..... ...., ...., ....,

,....

15.3

i8.1

18.62 19.44

II .2

il . 3

21.84 22.88

16.5 10.0

26.44 22.51 28.08 23.86

,.... ,....

90 I00

15.05 15.79 16.54

16.19 17.00 17.81

18.78 19.80 20.81

17.20 18.15 19.12

!7.76 21.88 18.75 10.4 23.35 19.96 24.86 21.21 13.3

11.1 14.1 17.5

16.2 10.5 15.5

19.4 i7.9 i8.6

60 70 80

13.93 14.31

14.94 15.36

17.23 17.76

15.78 16.26

!6.82 !4.17 19.69 16.98 !8.20 !5.35 20.42 17.57

10.5 12.3

19.3 12.4

45 50

12.75 13.15 13.54

15.60 16.16 16.70

14.29 14.79 15.29

!5.42 !2.94 !0.81 17.52 15.23 !7.03 !4.20 !1 .PI 18.25 15.82 !8.70 !5.48 !3.03 18.96 16.41

11.3 13.8 16.4

30 35 40

13.62 14.07 14.51

13.85 14.46 15.04

12.68 13.24 13.78

15.24 13.38 16.02 14.02 16.78 14.64

7.56 8.65 9.74

!0.81 9.18 !2.32 !0.45 !3.86 !I .71

!4.55 !6.69 !8.94

I5 20 25

11.29 11.68 12.05

1 I .80

10.34 21.42 22.55

33.3 37.1 41.6 35.0 39.0 43.6

- -

19.31

16.36 17.32 18.31

13.45 14.44 15.41

150

150 100 150

100

150

100

150

100

100 250

175

125

1.83

90 00

60 70 80

45 50

30 35 40

I5 20 25

2.39 12.93

29.99

22.21 24.52 27.09

16.30 18.13 20.10

13.66 14.53 15.41

0.99 1.24

31.5

23.32 25.74 28.42

17.09 19.03 21.08

14.33 15.24 16.16

13.02 13.40

12.41 12.78

9.71 9.87

10.66 10.88

11.14 11.37

0.17 0.45 0.73

9.18 9.36 9.54

9.98 10.22 10.44

10.40 10.66 10.90

12.22 12.63

8.55 8.77 8.98

9.18 9.46 9.73

9.56 9.86 10.14

828

S T E M AND POWER GENERATION

sure can also be easily estimated. All that is needed is a Mollier chart, which is reproduced in engineering handbooks available in most engineering offices and schools. The use of theoretical steam rate tables as published by the ASME makes it even easier to estimate turbine steam rates. A condensed table of theoretical steam rates is included as Table 16.4. These theoretical steam rate tables are not normally so universally available as the Mollier chart, and therefore a brief explanation of the use of each will be given. A part of a Mollier chart is included as Fig. 16.3. Mollier charts range in size from a small sheet of paper to those used by designers, which may be as long as your automobile, 15 to 20 ft. I n spite of the many lines on such a chart, you will find it very simple to use. T h e Mollier chart is divided into two sections by what is called the saluration line. The saturation line is a point at which the water has all been evaporated into a vapor and is a t a temperature corresponding t o the boiling point of water a t the selected pressure. Along this saturation line, the vapor is often referred to as “dry and saturated” steam. At all points on the Mollier chart below this saturation line, the steam is mixed with water, more often called moisture and is sometimes referred to as “ n e t steam.” Lines parallel to, and below the saturation line, labeled 2, 4, 6, 8, etc., are lines of constant moisture in per cent. Above the saturation line, no moisture exists in the steam, and it is normally referred to as “superheated steam.” Lines parallel to and above the saturation line, labeled 20, 40, 60, 80, etc., are lines of constant superheat in degrees Fahrenheit. I n addition to the lines of constant moisture and constant superheat, there are also lines of constant pressure in pounds per square inch absolute (not pounds per square inch gauge), which extend upward to the right diagonally across the Mollier chart. Other lines starting a t the saturation line, labeled 200, 240, 280, 320, etc., and curving upward to the right are lines of constant temperature in degrees Fahrenheit. Horizontal lines on the chart show energy, normally called enthalpy, in Btu per pound for steam at various conditions. Enthalpy is of interest to the application engineer and the user, as well as t o the designer. Vertical lines on the chart are lines of constant entropy. Except to remember that steam would expand along a constant entropy line in a perfect turbine, the term entropy has little practical use in the application of steam turbines. By use of the Mollier chart, i t is easy t o find the energy available for power for any range of initial and exhaust steam conditions selected. Furthermore, it is possible t o determine the conditions (degrees superheat, temperature, enthalpy, per cent moisture, etc.) of steam exhausted or


829

extracted from a turbine a t any desired pressure for use in process or to supply other turbines designed for the lower initial steam conditions. Examples Explaining Use of the Mollier Chart. As will be noted from the chart, the lines of constant pressure are in pounds per square inch absolute (psia). Pressures normally quoted for steam turbines and steam boilers are in pounds per square inch gauge (psig), and, a t sea-level conditions pounds per square inch gauge are converted into pounds per square inch absolute by simply adding 14.7 (normal atmospheric pressure). For estimating purposes, 14.7 is usually rounded out t o 15 to simplify the arithmetic. Now t o check values used in Fig. 16.5 for the 60OpSig-825 F-2 in. Hg abs steam-turbine plant. Initial steam conditions are 615 psia-825 F. This is plotted as point A on the Mollier chart, Fig. 16.3. As stated earlier, if the turbines were 100 per cent efficient, the steam would expand down along a line of constant entropy on the Mollier chart. For this particular example, then, steam would expand from the initial conditions of 615 psia-825 F, as plotted a t point A , along a vertical line until it intersects the 2-in. Hg abs line, which is located a t point B on the chart. These theoretical expansion lines are of importance. For instance, data such as these, from the Mollier chart, were used in calculating the theoretical steam rates shown in Table 16.4. These theoretical expansion lines are also useful in estimating actual steam-turbine expansion lines, which are used in determining the steam conditions (temperature, degrees superheat, per cent moisture, etc.) a t any extract.ion or exhaust steam pressure. Now to establish an approximate turbine expansion line for the unit operating a t 600 psig-825 F-2 in. Hg abs-first determine the available energy between the points A and B ; a t point A on the Mollier chart, read an enthalpy of 1420.8 Btu per Ib, and a t point B read an enthalpy of 916.5 Btu per Ib, or a difference of 504.3 Btu per lb. For an average value of over-all turbine-generator efficiency, it was suggested that a value of 75 per cent be used. This was discussed earlier in this chapter. Neglecting turbine-generator mechanical and electrical losses, this means that only 75 per cent of this 504.3 Btu, or 378.2 Btu, will actually be used by the turbine. By subtracting this 378.2 Btu from the initial enthalpy of 1420.8 Rtu per Ib, the actual turbine exhaust enthalpy is 1042.6 Btu per Ib. This now establishes the turbine exhaust enthalpy based on the assumed efficiency and the selected exhaust pressure of 2 in. of Hg abs. The approximate turbine exhaust conditions are now plotted a t the intersection of the horizontal 1042.6 Btu per Ib enthalpy line and the diagonal 2 in. Hg abs line; this is shown as point C on the Mollier chart, Fig. 16.3.

830

STEAM AND POWER GENEFATION

When considering full-load operation of this straight condensing turbine, a straight line drawn between point A and point C on the Mollier chart will give a turbine expansion line (sometimes called a “state line”) of

sufficient accuracy for most estimating purposes. Actually, the slightly curved line drawn between point A and point C on Fig. 16.3 approsimates the actual turbine expansion line a little more closely than a straight line would. Having determined the theoretical expansion line and, from this, established an approximate actual turbine expansion line, what use can be made of such an actual expansion line? As shown on Fig. 16.5, 59.7 per cent of the heat in the steam supplied t o the turbine is given up to the condenser circulating water. I n industries such as the pulp and paper industry, the chemical industry, the petroleum industry, the textile industry and others, large quantities of heat energy are required for cooking, washing, heating, drying, and for other process uses. From a practical consideration, temperatures higher than 101 F (the temperature of saturated steam a t 2 in. Hg abs) are required t o do this work. For that reason, most of the turbines used in industrial plants are arranged to exhaust and/or extract steam a t a higher pressure and temperature as required t o satisfy these various process beat requirements. I n the esample, in the section on Theoretical Steam Cycle, it was shown that the theoretical steam-turbine-cycle efficiency could be more than doubled by recovering only two-thirds of the heat in the exhaust steam. More will be said about the type of turbines normally used in industry later, but now for a look a t the turbine expansion line AC t o see what information is readily available from such a line. Referring t o this expansion line AC, it is seen that the turbine exhaust has 6 per cent moisture and is a t a temperature of 101 F. I t is impossible to establish that temperature, on this small incomplete chart, but it is found by following the constaut-pressure line diagonally across to the right to the iutersection of the saturation line and then interpolating between lines of constant temperature. Now, by assuming that steam is to be extracted from this turbine a t 50 psig (65.psia), it is seen that this extracted steam will have an euthalpy of 1260 Btu per lb and a total temperature of approuimately 455 F, which is 157 F superheat. Of course, depending upon the usage of 50 psig steam, the turbine might be designed as a uoncondensing turbine exhausting a t 50 psig or as an automatic-extraction condensing turbine exhausting a t 2 in. Hg with automatic extraction a t 50 psig. In the’case of a noncondensing turbine plant, with all heat in the exhaust used for process, actual steam power-plant-cycle efficiency would approach 75 per cent. Another example using the Mollier chart is included on page 837 under Selection of Steam Pressures and Temperatures.

S T W AND POWER GENERATION

831

THEORETICAL STEAM RATE TABLES

Reference t o theoretical steam rate tables was made earlier in this section, and in Table 16.4 theoretical steam rates are calculated with data from the Mollier chart, Fig. 16.3. By drawing the theoretical (100 per cent efficient) expansion line, the energy available between any selected initial steam conditions and final exhaust pressure can be determined. This is 504.3 Btu per Ib for the example shown on Fig. 16.3. Since there are 3413 Btu per kwhr, then the theoretical steam rate (pounds steam per kwhr) for a steam turbine-generator is simply 3413/theoretical energy available. For the example shown on Fig. 16.3, the theoretical steam rate was 6.766 Ib per kwhr. For a quick check on the “by-product power” which can be generated with a given total quantity of steam required for process, the actual steam rate (pounds steam per kwhr) can be quickly estimated. It is only necessary t o divide the theoretical steam rate by an assumed over-all turbinegenerator efficiency to get the actual steam rate. I n the example of Fig. 16.3 for a straight condensing turbine, and an assumed turbine-generator efficiency of 75 per cent (refer t o Turbinegenerator Efficiencies, page 820),the actual steam rate was 9.02 Ib steam per kwhr. For a turbine steam flow of 100,000 lb per hr, the approximate turbine-generator output would be 100,000 Ib per hr/9.02 Ib per kwhr, or 11,000 kw. The output of a noncondensing turbine generating “by-product” power would be determined in the same way. The above methods are based on the use of the Mollier chart or the theoretical steam rate tables and an assumed efficiency. More accurate performance of straight noncondensing and straight condensing turbine-generators can be readily determined from Figs. 16.18 and 16.19 referred to under Performance of Steam Turbines, page 857. STEAM BOILERS

Modern fuel-fired power boilers are normally designed for efficiencies ranging from 80 t o 88 per cent, depending to some extent on the type of fuels used and the unit cost of fuel. As an estimating figure, for modern power boilers, a value of 80 t o 85 per cent for boiler efficiency should be satisfactory for most conditions. AUXILIARIES

Powerhouse auxiliaries such as pumps, fans, coal pulverizers, etc., consume 4 t o 8 per cent of the power generated in steam-turbine power plants. The amount of auxiliary power required depends upon the type

832


of fuel, boiler pressure, and boiler flow. Induced-draft fans, boiler feed pumps, and coal pulverizers (when used) are the large users of auxiliary power. An average value of 6 per cent should be sufficiently accurate for estimating purposes. Figure 16.7 gives estimating data on boiler auxiliaries. ACTUAL EFFICIENCY OF STEAM PLANTS FOR GENERATION O F ELECTRIC POWER ONLY

The theoretical cycle efficiency of straight condensing steam power plants (plants for electric power generation only) depends upon the initial steam conditions and the exhaust conditions, as was indicated by the


833

plants discussed earlier in this section. I n these plants, the theoretical cycle efficiency ranged from 37 t o 45 per cent. If the various average full-load efficiencies of the turbine (75 per cent), boilers (80 to 85 per cent), and auxiliary losses (6 per cent) are factored in, actual full-load efficiencies are 75 per cent X 82 per cent X 94 per cent = 57.8 per cent of the theoretical cycle efficiency. The actual over-all cycle efficiency of straight condensing steam turbine plants without reheat and/or regenerative feedwater heating is then 57.8 per cent X theoretical cycle efficiency, or 21.4 t o 26 per cent. Regenerative feedwater heating, reheat cycles, and other cycle refinenieuts can be used t,o improve the condensing steam-turbine cycle efficiency. A few modern utility plants of large capacity have an actual power-plant cycle efficiency of about 37 per cent. The greatest gain comes from reducing exhaust flow t o the condenser, and all these refinements are “directed t o reducing the flow to the condenser. ACTUAL EFFICIENCY OF STEAM PLANTS TO SUPPLY BOTH ELECTRIC POWER AND PROCESS HEAT

For steam power plants supplying both electric power and process heat, such as process steam (the common arrangement in industrial plants), the actual over-all power-plant-cycle efficiency can be as high as 70 to 8Oper cent compared with the 20 to 35 or 40 per cent now attainable with a steam plant used for power generation only (see efficiency of typical plants in Fig. 16.4). This higher efficiency is realized because heat in the exhaust steam is required for process work. With a perfect “heat balance” there would be zero flow to the condenser and all heat energy (enthalpy) in the turbine exhaust or extraction steam would be used in process. a explanation of the 70 to 80 per cent over-all power-plantNow f o r ~ brief cycle efficiency-since all heat in the turbine exhaust or extraction steam is utilized, internal (steam-path) efficiency of the turbine is not of importance as long as all electric power requirements can be generated with the steam required in process. Actually, in most plants, high turbine efficiency is important because all electric power requirements cannot he generated with the steam required in process. I n any event, with a perfect “heat balance” (all power generated from steam required in process), it is necessary t o consider only the mechanical and electrical losses of the turbine-generator, whereas steam-path losses also had t o be included when heat energy in the exhaust was not used. Mechanical and electrical losses for turbine-generators used in industrial plants are approximately 4 or 5 per cent. Factoring in the various full-load efficiencies for the steam turbine plant supplying heat to process, say, 95 per cent for the

034

S T U M AND POWER GENERATION

Percent of Boiler Flow to Main Condensers FIG. 16.8

Per cent over-all steam-power-plont efficiency

VI.

per cent flow to main

condensers.

tnrbine-generator (mechanical and electrical efficiency), 82 per cent for the boilers, and 94 per cent for auxiliaries (6 per cent auxiliary loss) gives an actual over-all full-load steam-power-plant efficiency of 95 per cent X 82 per cent X 94 per cent = 73.3 per cent. This is equivalent to an actual net heat rate of 3413/73.3 per cent, or 4650 Btu per kwhr. The actual full-load efficiency of industrial steam turbine plants then ranges from 15 to 35 per cent (22,700 to 9750 Btu per kwhr) for straight condensing plants up as high as 75 t o 80 per cent (4550 to 4260 Btu per kwhr) for noncondensing plants supplying steam t o process. The e 8 ciency of condensing extraction or a combination of condensing and noncondensing turbine plantsgenerating electric power and supplying steam for process will range from 15 to 35 per cent to as high as 75 per cent. The actual efficiency will depend upon the per cent of boiler flow passed to the condenser, assuming that the remaining boiler flow is exhausted or extracted from the turbines for use in process. This variation in approximate power-plant efficiency is shown on Fig. 16.8. Much useful information of interest t o engineers concerned with the specification, selection, application, and operation of steam turbines can be found in other texts.*

* L. E. Newman, “Modern Turbines,” John Wiley & Sons, h e . , New York, 194% Gustaf A. Gsffert, “Steam Power Stations,” 4th ed., McGraw-HiU Book Company, Inc., New York, 1952; J. Kenneth Salisbury, “Steam Turbines and Their Cycles,” John Wiley & Sons, Inc., New.York, 1950.


835

SELECTION OF STEAM PRESSURES AND TEMPERATURES

Steam turbines in the utility stations today are being designed for initial pressures up t o 2400 psig and initial temperatures up to 1100 F. Operating steam conditions of 4500 psig and 1150 F should be a reality by 1956. The levels of pressure and temperature in industrial plants are increasing, as they are in utility plants. The average steam conditions of industrial turbines designed today are approximately 600 psig-750 F. I n 1930, this average was approximately 250 psig--500 F. There are several industrial plants today utilizing initial pressures higher than 1000 psig. See the partial list of industrial generating units in Table 16.3 on page 810. As discussed under Theoretical Steam Cycle, page 823, the theoretical cycle efficiency of a straight condensing steam turbine plant increases from 37 per cent to 45 per cent when the initial steam conditions are increased from 400 psig-750 F to 2400 psig-1000 F. If all heat in the exhaust steam can be utilized in process, the theoretical cycle efficiency would of course be 100 per cent for either case. Those factors indicate three ways of increasing the cycle efficiency of the steam-turbine cycle: (1) increased initial steam conditions, (2) decreased steam flow t o the condenser, and (3) maximum utilization of heat in the turbine exhaust steam. OPTIMUM INITIAL STEAM CONDITIONS

What is the optimum initial steam pressure and temperature? Obviously there is no single answer to that question. Within the practical limitations of first cost, operating and maintenance problems, etc., the initial pressure and temperahre selected for an industrial turbine will depend upon the requirements of the system. These would be the requirements for exhaust and/or extraction steam and electric energy in the case of noncondensing and automatic-extraction type turbines and the requirements for electric energy in the case of straight condensing steam turbines. The first step then is to establish daily load curves for electric power and for process and beating steam. These curves should cover both winter and summer operation. Having established t,hese load curves, the ratio of process steam flow and electric power, as well as the process steam pressure (or pressures), will be important factors in establishing the initial steam conditions for turbine-generators. There are many other factors, such as fuel costs and the probability of plant expa.nsion or a change in process, t o be considered

836


in the selection of initial steam conditions. Engineers familiar with turbine-generator applications can assist in selecting the proper steam conditions and apparatus to meet each particular plant requirement. The exhaust or extraction pressures can be designed to meet the requirements for process heat or even high enough to supply other steam turbines in the system. One characteristic of steam which makes it such a desirable source of process heat is the fact that heat is given up a t constant temperature, except in the superheated region, and is dependent upon the pressure. For instance, if a paper mill desires a steam temperature of 300 F a t the rolls, process steam pressure selected would he 68 psia (53 psig) a t the rolls. If steam a t 366 F is desired for cooking pulp, then 165 psia (150 p i g ) would he selected as the process steam pressure. These pressures are easily determined by referring to a Mollier chart surh as illustrated in Fig. 16.3. The required pressnre can he read at the intersection of the saturation line and the line of desired constant temperature. For most processes it is desirable that saturated steam be supplied t o the process. I n plants where the distance from the power plant t o the process area is great, it is usually desirable to supply steam from the powerhouse with some superheat to compensate for radiation losses in the lines to process.

EFFECT OF INITIAL STEAM CONDITIONS ON PLANT FUEL CONSUMPTION

The example shown in Table 16.5 indicates the importance of initial steam conditions on the fuel consumption of a power plant. For this example, power-plant requirements were assumed to be 7500 kw of electric power and 125,000 lb per hr of process steam a t 50 psig-variations in superheat of the process steam were not considered. This tabulation shows that the plant f u d consumption increases almost 30 per cent as initial steam conditions are reduced from 1200 psig-825 F t o 200 psig600 F. With the higher initial steam conditions, a straight noncondensing turbine can generate the required electric power without any “condensing” kilowatts. At the lower initial pressure, more than half the electric power had t o be generated by steam flow to the condenser with the attendant losses t o the condenser circulating water. For a case of this kind, initial plant cost, quality of available feedwater, expected maintenance, and other factors must he weighed against the savings in fuel cost to determine the optimum initial steam conditions. It would he desirable to have complete turbine-performance data for the various types of turbines and for the various steam c3nditions for each method of supplying the plant process steam and electric power require ments. However, from a practical standpoint, that is too laborious when

837


many alternatives are considered. Although the data referred t o in the following examples are for full-load machine rating only, they are applicable for the various types of turbines. From these data, the most attractive operating cycles can be quickly determined; the final selection can be arrived a t by using more complete performance data and making calculations for the few alternates remaining. To illustrate the use of the data, several examples are worked out in detail. A summary of various methods of providing process steam and electric power is shown in Table 16.5. The comparison of the over-all plant thermal efficiency and relative fuel consumption for the various methods indicates the importance of such studies t o establish the optimum cycle. Although these data pertain only to the over-all fuel usage, due allowance should be made for the initial and operating costs and existing mill pressures; the most efficient cycle may not necessarily be the most economical one. TABLE 16.5

Comporison of Efficiency and Fuel Consumption of Different Plants

Power-plant requirement.: 7500 kw electric energy; 125,000 Ib per hr of &ccss superheat of eitrocted steam not Critical

Power in kilo

Turbinegenerator*

eroted t, by

.~ Turbine d e i g n steam conditions

Number required

'On-

Rating, kw

lensing ,team

itemn at 50 p i g ;

Per cent

oYer-.ll powerplant thermal effid*"
~~

~~

1

7500

1200 1b-825 F-50 Ib

7500

0

75

1 .oo

1

7500

600 lb-750 F-2 in. Hg AE'ot 50 Ib

5820

1680

69

1.09

1

6000

I

I500

600 lb-750 F-50 Ib 600 lb-750 F-2 in. Hg

I

6000

1500

68

1.10

I

7500

400 lb-650 F-2 in. Hg AE at 50 Ib

4700

2800

64

1.17

I

7500

200 lb-600 F-2 in. Hg AE (It 50 Ib

3070

4430

58

1.29

* AE

=

I

automatic extraction.

EXAMPLES INDICATING EFFECT OF INITIAL STEAM CONDITIONS

Example 1. Now to use a Mollier chart such as illustrated in Fig. 16.3 for the approximate solution of some specific problems. Assume that a power plant is operated to supply electric power and process steam for an industrial plant. The existing turbines are 6000-kw double automaticextraction condensing units designed for steam conditions of 600 psig750 F--2 in. Hg with automatic extraction a t 140 psig and a t 70 psig.

838


A t rated operating conditions, it has been found that the enthalpy of steam extracted from the 6000-kw turbines is 1223 Btu (70 FS degrees superheat Fahrenheit) a t the 70-psig opening and 1262 Btu (117 FS) at the 140-psig opening. This is a turbine steam-path efficiency of 78 per cent down to the extraction openings. The owner is contemplating installation of new 10,000-kw double automatic-extraction condensing turbines and has arbitrarily selected an initial steam temperature of 825 F. Steam-path efficiency for the new 10,000-kw turbines a t rated operating conditions is 80 per cent down to the extraction openings. What initial steam pressure should be selected (assuming 825 F initial temperature) so that the enthalpy and superheat of the extracted steam from the proposed 10,000-kw turbines will be approximately the same as the steam extracted from the existing 6000-kw units (assume initial pressures in steps of 50 Ib, i.e., 800, 850, 900, etc.)? This is a trial-and-error solution. As a first start it would be well t o plot the enthalpy of extraction points for the 6000-kw units and also the initial steam conditions on the Mollier diagram. These are all given in the general data preceding the problem. The problem also stated that steam-path efficiency for the proposed 10,000-kw units would be slightly higher (80 per cent vs. 78 per cent) than for the existing 6000-kw units. Therefore the expansion line on the Mollier diagram will be slightly steeper for the more efficient 10,000-kw unit. To assist in selecting the initial pressure, this estimated expansion line can be sketched in t o give approximately the same steam conditions a t the extraction pressures. 1. As a first assumption, assume an initial pressure of 800 psig-from the Mollier chart: At 800 psig-825 F, the-enthalpy is.. . . . . . . . . . . . . 1412.0 Btu per lb Theoretical extraction a t 140 psig, the enthalpy is. . 1228.0 Btu per lb Theoretical energy available.. . . . . . . . . . . . . . . . . . . . 184.0 Btu per Ib Energy used a t 80% eff. = 80% X 184 = . . . . . . . . 147.0 Btu per Ib :.Actual enthalpy a t 140-psig extraction opening is: 1412 - 147 = . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12G5.0 Btu per lb Using the Mollier chart, the extraction temperature is 484 F or 123 FS. Using 800 psig-825 F init,ial pressure, now check steam conditions at the 70-psig extraction opening: . . . 1412.0 Btu per Ib At 800 psig-825 F the enthalpy is Theoretical extraction a t 70 psig, the enthalpy is. . 1176.5 Btu per Ib Theoretical energy available . . . . . . . . . . . . . . . . 235.5 Btu per Ib Energy used a t 80% eff. = 80% X 235.5 = . . . . . . . 188.0 Btu per lb

STEAM AND POWER GE?4ERATION

a39

:.Actual enthalpy a t 70-psig extraction opening is 1224.0 Btu per Ib 1412 - 188 = _. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Using the Mollier chart, the extraction temperature is 387 F or 71 FS. 2. As a second assumption, assume an initial pressure of 850 psig: A t 850 psig-825 F, the enthalpy is.. . . . . . . . . . . . . 1410.8 Btu per Ib Theoretical extraction a t 140 p i g , the enthalpy is. . 1222.0 Btu per lb 188.8 Btu per Ih Theoretical energy availahl 151.0 Btu per Ib Energy used a t 80% eff. = :.Actual enthalpy a t 140-psig extraction opening is 1259.8 Btu per Ih 1410.8 - 151 = . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Using the Mollier chart, the extraction temperature is 475 F or 114 FS. Using 850 pig-825 F initial pressure, now check steam conditions at the 70-psig extraction opening. . . . . . 1410.8 Rtu per Ib At 850 pig-825 F, the enthalpy is Theoretical extraction a t 70 p i g , the enthalpy is. . 1170.0 Btu per Ib Theoretical energy availabl . . . . . 240.8 Btu per lb 192.0 Btu per lb Energy used a t 80% eff. = :.Actual enthalpy a t 70-psig extraction opening is 1410.8 - 192 = . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1218.8 Btu per lh Using the Mollier chart, the extraction temperature is 380 F or 64 FS. For this particular ease, either 800- or 850-psig initial pressure should be satisfactory from the standpoint of temperature of extraction steam. From the above trial-and-error solution, you will note that initial steam conditions can be estimated quite accurately by simply extending the expansion line of the existing turbine upward to the left. Any initial steam condition selected along the extension of this line will give the desired extraction conditions. For turbines having the same efficiency, linci drawn parallel t o expansion line for the existing turbine can he used in selecting initial steam conditions required t o supply different extraction or exhaust steam requirements. Example 2. Based on the information in Example 1 and neglecting mechanical and electrical efficiency of the turbine-generator (which is about 95 per cent), what by-product power will be generated with the new 10,000-kw units for each 100,000 Ib per hr steam extracted from a. The 140-psig opening? b. The 70-psig opening? Neglecting mechanical and electrical efficiency,the kwhr of " by-product" power are equal to


840

The Btu used 3413 or

Steam flow in Ih per hr X Btu used per Ib of steam 3413 Btu per kwhr For the 850 psig-825 Example 1:

F initial steam conditions assumed as part 2,

For the 140-psig extraction,

For the 70 psig extraction,

or

4430

+ 5630

=

10,060 kwhr total

For the assumed efficiencies for double automatic-extraction condensing turbine units and with assumed extraction flows of 100,000 Ib per hr from each opening (140 psig and 70 psig), note the small variation in temperature of the extracted steam for all the assumed initial steam conditions. Also note the great difference in by-product power as shown in Table 16.6. TABLE 16.6 Variation in Temperature of Process Steam and By-product Power from Steam Turbines with Different Initial Steam Conditions (Based on 100,000 Ib per hr of steam Row to each process pressure) Temporawre of

extracted rteom By-product

lnitiai deem conditions

600 prig-750 800 psig-825 850 pdg-825 1200 pdg-900

F F F F

478 F 484 F 475 F 467 F

386 387 380 375

F F F F

8.000 9,830 10,060 1 1,930

Example 3. This example is based on a plant requiring 125,000 lb per hr process steam at 50 psig and 7500 kw of electric power. Results of


84 1

several examples to determine efficiency and relative fuel consumption of various plants to meet the process steam and electric power requirements of the plant are given in Table 16.5. For the second plant listed in Table 16.5, assume that the mill requirements are to be supplied by a 7500-kw condensing-extraction turbine designed for steam conditions of 600 psig-750 F (261 FS)-2 in. Hg ahs with automatic extraction a t 50 psig. A condensing-extraction turbine is in effect a combined noncondensing turbine and condensing turbine which has the extraction steam flowing in parallel with additional steam that passes through the turbine and exhausts t o the condenser. The approximate performance of an extraction-type machine can be determined for estimating purposes from data included herein: 1. Determine the output from the extraction steam flow by finding the actual steam rate (ASR) of a straight noncondensing turbine having an exhaust pressure equal to the extraction pressure. Then correct the output, (kilowatts) so obtained by the multiplying factor 0.97.* 2. Find the deficit of kilowatts that must be produced by the condensing power by subtracting the corrected output found in 1 above from the total required electric output. 3. Find the r.team rate (ASR) for a straight condensing turbine corresponding to the initial pressure and temperature and the exhaust vacuum. Correct for a condensing-extraction unit by using the multiplying factor of 1.03.* Now to continue with this example and to find the required boiler output, the steam flow t o the condenser, and the over-all power-plant thermal efficiency. Step 1. Power generated by 125,000 Ih per hr extraction steam flow: Noncondensing steam rate for a 7500-kw unit (Fig. 16.18) a t 600 psig750 F-50 psig is 20.8 lh per kwhr. Noncondensing output: 125,000 lh per hr extraction flow = 6000 kw 20.8 Ib per kwhr steam rate

* The actual efficiency of automatic-extraction-type steam turbines is slightly lower than for straight condensing or noncondensing turbines having equal rating. This is due to additional control valves, longer turbine span, etc. These multiplying factors have been arbitrarily chosen because the actual decrease in efficiency will vary for different design conditions. This slight decrease in turbine efficiency is insignificant when the improvements in plant cycle efficiency made possible by the use of automatic extraction turbines are considered. Flow to the condenser will then be the corrected steam rate multiplied by the kildwstts generated by condensing power.

842


Corrected output from extraction flow for a 7500-kw automaticextraction unit: 6000 kw X 0.97 (factor from 1 above)

=

5820 kw

Step 2. Power to be generated by condensing steam flow: 7500 kw

- 5820 kw

=

1680 kw

S k p 3. Over-all (inlet-to-exhaust) steam rate: Actual steam rate of a 7500-kw condensing unit from Fig. 16.19 is 9.4 lb per kwhr. Corrected steam rate for 7500-kw automatic-extraction condensing unit: 9.4 Ib per kwhr X 1.03 (factor from 3 above)

=

9.7 lb per kwhr

Steam flow t o the condenser: 1680 kw (step 2) X 9.7 Ib per kwhr

=

16,300 Ib per hr

Total boiler or throttle flow: Slrom Conditions

0,

N ~ l ~ - F l ~ On u * Curr.1 l 400PSIG-650F

400PSIG-ll)OF

Turbine Throllle

Denom Exhaust Prsssurc ,PSIG 800PSIG-650F

600 PSIG-750F

SO0 400

300 200

100

w n r o t o r ovtpvtin ~ h o u s a n d sof Kilowollr

FIG. 16.9

Exhaust-steam conditions for 5000-kw 36CQ-rpm noncondenring fvrbinegenerator units operating at various initial and exhaust steam conditions.


125,000

+ 16,300

=

843

141,300 Ib per hr

Per cent of boiler output passed to main condenser: 16,300 Ih per hr condenser flow 141,300 Ib per hr boiler flow

=

0,115 = 11,5%

Over-all power-plant thermal efficiency:69% as read from Fig. 16.8. Approximate exhaust or extraction temperature and enthalpy of steam from noncondensing or automatic extraction turbine-generator units: Neglecting difference in efficiencybetween 7500-kw and 5000-kw units. Read from Fig. 16.9 a t full load (5000 kw) for the 5000-kw unit with steam conditions 600 psig-750 F-50 psig. Temperature, 325 F; enthalpy, 1200 Btu per Ih Turbine-generator required : One 7500-kw automatic-extraction condensing unit designed for 600 psig-750 F-2 in. Hg abs with automatic extraction a t 50 psig. Sl.om

Gondilionr 01 Turbine T h r o t t l e

600 Exhaust Tamp F

600

400 200

25

FIG. 16.10 Exhaust-steam conditions for 54XQ-h 3MX)-rpm noncondenring hlrbinegenerator unih operating a t various initial and exhaust steam conditions.


844

SELECTION

OF STEAM TURBINES-TYPE, GOVERNING, AND RATING

The various types of steam turbines can all have approximately the same turbine efficiency, This being the case, the important consideration is to select the proper type of turbine and the proper initial and exhaust steam conditions for each particular application. The important thing is to pick a turbine or combination of turbines which can supply all plant requirements for process and heating steam and electric power with a minimum flow of steam to the condenser. Once the requirements of the plant are established, the Mollier diagram can he used t o select the optimum initial steam conditions from the standpoint of power-plant-cycle efficiency. Whether a combination of condensing and noncondensing turbines is selected or automatic-extraction turbines are selected will depend upon the ratings involved and the balance between electric power and process and heating steam. TYPES OF TURBINES

Steam turbines are normally classified as: 1. Straight condensing 2. Straight noncondensing 3. Autoextraction condensing 4. Autoextraction noncondensing 5. Autoextraction-admission Condensing or noncondensing The actual engine efficiency (used energy + theoretical energy available for the actual steam conditions) of these different types of turbines used for power generation in industrial plants is about the same regardless of type, hut the power-plantcycle efficiency varies widely, depending upon the type of turbine and'the use of heat in the extraction and exhaust steam. These various types of steam turbines are indicated schemahically on Fig. 16.11. Using this schematic diagram as a reference, the various types of steam turbines will be briefly described below: Straight condensing turbines are designed to take steam a t boiler pressure and exhaust t o a steam condenser operating below atmospheric pressure. Normal exhaust pressures for a condensing turbine used in industrial plants range from 1% t o 3 in. Hg abs (27 to 28% in. Hg vacuum referred to a 30-in. Hg barometer), with 2 in. Hg being most often selected. Straight condensing turbines are widely used in our utility plants hut account for less than 10 per cent of the steam turbines being sold for power generation in the industrial plants.

846

STEAM AND POWER GENERATIOM

Straight noncondensing turbines are designed t o take steam at boiler pressure and exhaust a t atmospheric pressure or above. For equal power outputs, noncondensing turbines may require two to five times the steam flow required by the condensing turbine. Even so, when properly applied, the noncondensing turbine power-plant-cycle efficiency is usually two to three times that of the condensing turbine plant. Straight noncondensing turbines are widely used in industrial plants and account for approximately 30 per cent of the steam turbines being sold for power generation in industrial plants. When noncondensing turbines are applied, heat energy in the exhaust steam is usually used for heating, drying, cooking, and various other process uses. Many noncondensing turbines are also used, in both industrial and utility plants, as “topping” turbines, with the exhaust steam being utilized t o drive other turbines, which are designed for lower initial steam pressures. Noncondensing-turbine exhaust pressures may range up t o 600 psig or more and exhaust temperatures up to 600 F. Automatic-extraction condensing turbines are similar to straight condensing turbines except that necessary valves and controls are provided so that steam may be automatically extracted (taken out) a t one or more* automatically controlled pressures lower than the turbine initial pressure but higher than the turbine exhaust pressure. The autoextraction turbine can supply varying demands for extracted steam and electric energy, whether these demands vary individually or there is a variation in all demands simultaneously. The pressure of extracted steam is automatically controlled by the turbine governing mechanism and can be used for the same purposes as the exhaust steam from the noncondensing type of turbine described above. A cross-sectioned diagram showing the relation between various sections of a double automatic-extraction steam turbine with the location of control valves indicated is shown in Fig. 16.12. Autoextraction condensing turbines are almost exclusively used in industry. Of all the types of automatic-extraction turbines, these find the widest field of application. Because of their great flexibility in supplying variable demands for process steam and kilowatts, they account for approximately 45 per cent of the steam turbines being sold for power generation in industrial plants. * Turbines having one controlled extraction opening are called single autamatieextraction turbines, those having two controlled extraction openings are called double automatic-extraction turbines, etc. Single and double autamatie-ertraction turbines are extensively used in industry. The first triple automatic-extraction turbine is installed in a. paper mill of the Union Bag Company in Savannah, Ga. In mast plants, more than one or two automatic extraction openings &renot warranted for a single turbine.

FIG. 16.12 Cross-sectional drawing showing the variow sections and control ~(IIYBIof automatic-extraction steam turbine.

(I

double

848


Automatic-extraction noncondensing turbines are similar to autoe x t r a h o n condensing turbines except that the noncondensing turbine exhausts to a pressure equal to or greater than atmospheric and the condensing turbine exhausts to a pressure below atmospheric. The same number of controlled extraction openings can be provided. Like autoextraction condensing, autoextraction noncondensing turbines are almost exclusively used in industry and account for approximately 15 per cent of the steam turbines being sold for power generation in industrial plants. Automatic-extraction-odmission condensing turbines are similar to automatic-extraction condensing turbines except that control valves are arranged so that steam can enter, as well as leave, the turbine at the automatically controlled extraction-admission opening. Such turbines, acting as autoextraction units, supply steam to the process line vhen the use of this process steam exceeds the supply from other sources. When the supply of steam to this same process line, from other sources, exreeds the use of steam for process, then the Lurbiue governing mechanism acts, automatically, to admit t,he excess steam through the extractioti-admission opening of the turbine, so that it csn expand to exhaust pressure, or to a lower pressure extraction opening. During admission operation, the excess steam in the process line is used to generate power hy expansion in the turbine instead of possibly being blown to the at,mosphere t,hrough a relief valve. In some applirations, this type of turbine may he applied to fully utilize the capability of boiler plants which supply steam at two different pressures. For instance, it may sometimes be desirable t o purchase a turbine-generator designed for an initial steam pressure of 1200 psig and an extraction-admission pressure of 600 psig. Such a unit might he designcd t,o utilize 1200 psig steam from a new boiler plant plus 600 psig steam from an existing boiler plant. Depending upon the application requirements and expected normal operating conditions, this turbine unit could be designed to drive the generator at any per cent of rated output when supplied with 1200 psig steam only, when supplied with 600 psig steam only, or when supplied with 1200 psig steam only and extracting steam at 600 psig. A unit of this type might he installed for initial operation with 600 psig stcam before the 1200-psig boilers are in service, for instance, in a plant where management recognized the economics of an industrial power plant designed for higher initial steam conditions and was planning for this in the near future. With a turbine of this type, it mould not be necessary to shut the turbine-generator down during a forced or planned outage of either the GOO- or 1200-psig boiler plant. Automat,ic-extraction-admission turbines are usually applied in indus


849

try. They account for perhaps less than 5 per cent of the units sold to industry for power generation. SPEED GOVERNING

I n addition to these various types of steam turbines, there are many refinements and variat,ions which can be incorporated into the governing and control mechanisms. Take, for instance, t,he st,raight condensing turbine-this type of turbine is usually used for straight paver generation and as such is speed and load governed. Its normal speed governor regulation mould be sot for about 4 per rent for rated load change. In other words, if it were operating as an isolated unit, the speed and frequency would increase 4 per cent (2.4 cycles on a 60-cycle system) if full load were gradually taken off the unit without any change in position of the synchronizing device. If closer control of the frequency is desired, the turbine could be equipped with what is called a “speed corrector.” On turbines equipped with a speed corrector, the settled speed change to go through 4/4 gradual load change would be 0.1 per cent, or less than one-tenth cycle. The instantaneous variation from 60-cycle frequency would be greater than this 0.1 per cent, depending primarily upon the rate of load change and the system inertia. LOAD REGULATORS

If instead of having other turbines which can take the swingsin kilowatt load, it is desired t o take all load swings on one turbine, special turbinegoverning mechanisms can be furnished or the proper setting of standard governors may provide the necessary control for many applications. As one simple example of what can be done with the standard turbinegoverning mechanism, consider a typical plant, utilizing a “ topping” turbine, with steam-power diagram as shown in Fig. 16.13. The total electric load is 45,000 kw, with a 250-psig process steam load of 580,000 Ib per hr. Note that steam exhausted from the new topping turbine (unit No. 3) amounts t o 550,000 lb per hr. This is the entire steam flow required by the two existing condensing steam turbines plus 210,000 Ib per hr u,hich passes through the pressure-reducing valve (PRV) into the 250-psig header to help supply the steam required for general plant uses. Turbine-governing mechanisms are so arranged that, during periods of reduced plant electric load, the load on the existing condensing turbine generators will be reduced before the load on the topping unit is reduced. The curves of Fig. 16.14 show a division of steady-state plant load between the existing turbine-generators and the new topping tnrbine-

2 NEW 80 ILE RS

-

9 5 0 PSlG -900 F

I

If

LB/HR

170,000

0 LB/HR

--A

PR V

+c7

550,000 LB/HR

ir

lf

280 PSlG

170,000 LB/HR

210.000 LB/HR

r

I

i-1

LB/HR

2 5 0 PSlG

580,000 LWHR

PRV

370,000

LOW PRESS BOILERS

OLD

250PSIG-550F

ez

5m

50 B

y


851

generators. This division of load is accomplished automatically by the turbine-governing mechanisms without attention from the operator. By adjustment of the turbine-governing mechanism on the various turbinegenerator units the division of load between units can be varied over a wide range. The method of operation shown in Fig. 16.14 is desirable because fuel consumption, in Btu per kwhr, for the topping turbine (when credited with heat in its exhaust steam) is less than one-third the fuel consumption of the low-pressure condensing turbine cycle. In other cases, the contract with the utility may make it very desirable to keep purchased-power (kw) demand constant. Purchased-power demand can be held within the required limits by varying the power generation in the industrial plant. Depending upon the type of industrial load, it may be necessary to modify the action of the normal turbine (or other prime mover) governor with load-regulating equipment. Load regulators are wailable which act on the turbine-governing system

LOAD ON EACH TURBINE GENERATOR THOUSANDS OF K W FIG. 16.14 Diagram showing the division of load between topping and condensing units under turbine-governor control.

852


(usually the synchronizing motor), changing turbine-generator output to follow load changes within the industrial plant and minimize the load swings on the utility. “High-speed” load regulators are available which can act to change the load on the regulated turbine (or turbines) a t the rate of 4/4 load in, say, 4 see. “1,owspeed” regulators can act to change the load on the regulated turbine a t the ra.te of 4/4 load in 20 to 40 see or a t any slower rate desired. For most applications where manual load adjustment is not sat,isfactory, the lowspccd rcgulat,ors are adequate. There are applications, such as in the steel industry, where the high-speed regulator is desired, in an attempt to follow the rapidly fluctuating mill load. The controls discussed above for the straight condensing turbine can likewise he applied on the other types of turbines described earlier. INITIAL-PRESSURE GOVERNING

If a turbine is operating in parallel with other units which can supply variations in kiloivatt load, the turbine can be equipped with an initialpressure governor, The initial-pressure governor is used in cases where waste fuel, in varying quantities, is available for use in the steam boilers. This governor acts to close turbine valves (reduce steam flow and generator load) t o maintain constant boiler pressure. EXHAUST-PRESSURE GOVERNING

Noncondensing turbines are often supplied with exhaust- or backpressure governors. When operating on exhaust-pressure governor control, the noncondensing t,urbine must be t,ied in electrically with a system xrhich will maintain system frequency. The kilowatt load of the exhaustpressure governed t,urbine will then vary with the demand for steam in the exhaust line. EXTRACTION-PRESSURE GOVERNING

I n automatic-extraction turbines, the steam pressure in the extraction line is also cotitrolled by an extraction-pressure governor. The normal speed governor can control the load and frequency of these units by varying the flow to the exhaust, even though the extraction-pressure governor is act.iog t o supply varying amounts of steam to the extraction steam system. As mentioned earlier, extraction-admission turbines are very similar t o the autoextraction turbines except that the governor is so arranged that steam will be extracted frcm thc turbine to supply sieam to the process line when there is a deficit of steam in this line and will take steam


i

853

854


a t the admission opening for generation of power when there is an excess of steam in the process line which may be supplied by waste fuel-fired boilers, waste-heat boilers, or some other apparatus. GOVERNING PERFORMANCE CHARTS

The charts which are reproduced in Fig. 16.15 indicate how the turbinegoverning mechanism can control extraction pressures when the demand (flow) for process steam supplied by the turbine varies widely. The charts, Fig. 16.16, show how the turbinegoverning mechanism controls frequency with varying kilowatt load. These charts are typical for General Electric Company double automatic-extraction steam turbines.

FIG. 16.16 Typicol turbine kw load VI. time and plant system frequency VS. time On 10,000-kw double automatic-extraction General Electric Company turbine with extractionpressure governors and speed corrector in service.

TURBINE SIZE NONEXTRACTION TURBINES

The physical size of steam turbines depends more on steam conditions and the various requirements for process steam than it does on the rated power output. For example, the relative physical size of a “topping”


855

or noncondensing steam turbine may be 20 t o 30 per cent of the size of a condensing steam turbine having the same rated output. I n spite of the relatively small physical size of the topping turbine, the steam flow, in pounds per hour, may be two t o five or more times the flow required for equal output from the condensing turbine. AUTOMATIC-EXTRACTION AND/OR ADMISSION TURBINES

Automatic-extraction and admission turbines are in effect two or more turbines operating in tandem, with the proper governing and control, and physically located in a single turbine casing. The size of the various sections of the turbine will depend upon the application requirements and the different steam pressures. I n the application of automatic-extraction turbines, i t is important t o visualize the turbine as a series of turbines or turbine sections and realize that the total turbine output is the summation of the outputs of the various sections. Each section will have a specific flowoutput curve. I n selecting the proper turbine for the application, the application engineer must be very careful t o select a turbine with sufficient flexibility t o meet the range of operating conditions likely t o he encountered in normal operation. On the other band, i t must be remembered that each section of the turbine operates at maximum efficiency for a relatively narrow range of steam flows. I n Fig. 16.17, the division of load between the high-pressure and Ionpressure sections of a single automatic-extraction turbine is indicated for various load conditions. Note that flow-output curves for the two sections of the turbine are shown a t the left, and the flow capacity of the two sections of the turbine is indicated at the right. As can he seen, the turbine used for illustration xvas designed t o deliver rated output with extraction steam alone, and both sections of the turbine are never used t o capacity simultaneously. TURBINE-GENERATOR RATINGS

Manufacturers have standard turbine-generator ratings ranging from 200 km to more than 200,000 kw. The manufacturers are thus able t o meet the needs of industry. Turbine-generators rat,ed 500 t o 15,000 kw account for the greatest, percentage of turbine-gcncrators applied in industry. I n 1940, the avcrage rating of turbine-generators used in industry was about 4000 kw. I n 1954, that average rat,ing approached 10,000 kw. Some of the automobile steel, aluminum, and chemical plants have units rated up t o 100,000 k w or more.

856


g3

t2

100 PERCENT LOAD WITH NO EXTRACTION INLET FLOW

1100)

1300

:b

HIGH PRESS. SECT OUTPIJT IN % O F TURBINE RATING

200

SECTION

EXTRACTION EXHAUST FLOW (0) FLOW (1001

LOW PRESS. SECT OUTPUT IN % O F TURBINE RATING

50 PERCENT LOAD WITH NO EXTRACTION

(01 100 PERCENT LOAC WITH MAX EXTRACTION

(250)

50 PERCENT LOAD WITH MAX EXTRACTION (150)

FIG. 16.17

Section flow output curves rhowing division of load between high-pressure and low-pressure sections of the same single wtomotic-extroclion turbine operating under different conditions.


857

PERFORMANCE OF STEAM TURBINES NONEXTRACTION TURBINES

The data plotted on Fig. 16.18 show the steam consumed by noncondensing steam turbines rated 2000 to 15,000 km a t various load conditions. Figure 16.19 gives similar data for condensing steam turbines. Approximate data on the condition of steam exhausted from steam turbines a t different pressures are shown in Figs. 16.9 and 16.10. Average figures, for estimating purposes, on the efficiency of different turbine-generators are given under Turbine-generator Efficiencies on page 820. These average values of efficiency (65 per cent for ratings to 1500 kw and 75 per cent for units rated 2000 to 15,000 kw) are very useful for estimating performance of different types of steam turbines operating with different steam conditions. I t is only necessary t o divide the theoretical steam rate, normally ralled TSR, by the efficiency. Theoretical steam rate tables are published in booklet form by the American Society of Mechanical Engineers. A condensed table is shown in Table 16.4. AUTOMATIC EXTRACTION TURBINES

Example 3 in the section on Selection of Steam Pressures and Temperatures, page 840, is a step-by-step procedure for estimating the performance of automatic-extraction-type turbine-generators. It is estimated that performance calculated by the above methods will be accurate within 5 t o 10 per cent-adequate for most preliminary estimates. These data on the steam consumption and output of turbinegenerators do not indicate possible gains from any special conditions of operation that exist in most industries. When checking performance of a specific unit, such as an automaticextraction unit, a t various operating conditions, turbine-performance curves, such as indicated in Fig. 16.20 and based on detailed efficiency data, will he found very helpful. COMBUSTION-GAS-TURBINE POWER-PLANT CYCLES STATUS OF COMBUSTION GAS TURBINES FOR POWER GENERATION

The combustion gas turbine has now taken its place as EL prime mover, like the steam turbine and other prime movers, to generate electric power and supply heat energy simultaneously at high cycle efficiency.

858


0

z

.

'LBSAPPROX~MAT; STEAM RATE , IN LBS PER KW HR PER K W HR

FIG. 16.19 Approximate steam rates of condensing steam turbine-generator units.

HERE STEAM RATE:IO.l

D4SH-LINE ILLUSTRPTION FOR THE FOLLOWING CONDITIONS 4000KW 310 P S l G ~ 7 5 O P F T T - 2 " H G4 B S 3 / 4 LOAD

THROTTLE PRESSURE, I N P S l G

2 0 0 0 K W RlTlNG

11.52 3 4

EXHAUST PRESSURE IN.HG A B S

860


*Ii I I I 4 I I I

I


861

The first commercial, nonmilitary application of a combustion gas turbine-generator in the United States was in 1949 in the Belle Isle station of the Oklahoma Gas and Electric Company. Many other units have been placed in service since. The combustion gas turbines for commercial land application are designed with safety factors and stresses comparable to those used in steam-turbine design, to attain long life and low maintenance. WHAT IS A COMBUSTION GAS TURBINE?

The combustion gas turbine is a very simple self-contained prime mover. A simple open-cycle gas turbine is shown diagrammatically in Fig. 16.21. It includes the compressor, combustion equipment for the burning of fuel, and the turbine. Atmospheric air is compressed and passes into the combustion chambers. Fuel is burned with air required for combustion and mixed with excess air t o maintain required gas temperatures. These combustion gases are expanded through a turbine and are exhausted to atmosphere either directly or through various types of exhaust-heat recovery apparatus. Cooling-water requirements are relatively low-approximately 300 to 350 gpm for 5000-kw simple and regenerative cycle units. CHARACTERISTICS OF COMBUSTION GAS TURBINES DEVELOPED FOR POWER GENERATION

The efficiency of present-day combustion-gas-turbine-generator power plants a t full load ranges from approximately 18 per cent for the simple cycle unit, without intercoolers or regenerators, up to 25 per cent or more for the regenerat,ive cycle units utiliziug regenerators. When used in an industrial plant requiring both electric energy and process heat, heat energy in the gas-turbine exhaust gases can be recov-

p-1and comes ere

Fuel goes En here. and is burned to heat the air

The hot gases expand through the turbine +

LOR0

1 1 1

U

Air goes

and exhaust

L.)to the stack

FIG. 16.21

Schematic diagram of a simple-cycle gar-turbine unit.

0b2


ered. When heat in the exhaust gases can be utilized in process, either directly or to generate steam in exhaust-heat recovery boilers, actual gasturbine power-plant-cycle efficiency can run as high as 70 per cent-comparable to the high efficiency obtainable in other types of industrial power plants. One interesting characteristic of the gas turbine is the effect of ambient temperature on its output; the lower the temperature, the greater the output. For a decrease in ambient temperature of 40 to 50 F, the turbine output is increased approximately 25 per cent. On the other hand, temperatures above 80 F decrease the output. Much can be gained by surface or evaporative cooling t o 80 F or below. GAS-TURBINE POWER-PLANT-CYCLE EFFICIENCY

Heat in the exhaust gases, such as heat in the exhaust from a steam turbine, represents a great loss unless it can be recovered and used in the plant cycle. Efficiency of the simple-cycle turbine-generator unit alone, operating without heat-recovery equipment, is approximately 17 per cent. The gas-turhine power-plant-cycle efficiency ranges as high as 60 to 80 per cent when heat is recovered from the turbine exhaust gases for various process or power uses. The basic cycle efficiency was improved t o about 65 per cent plane cycle efficiency in one installation, where heat recovered from the exhaust gases was utilized to heat feed water for an existing steam plant. To illustrate use of the term plant-cycle thermal efiiency, the actual gasturbine plant-cycle thermal efficiency for this installation will he calculated, based on the following information: Fuel fired: 80,400,000 Btu per hr (higher heating value) Generator output: 3500 kw Feedwater heater flow: 350,000 Ib water per hr Feedwater in: 185 F (154 Btu per lb) Feedwater out: 295 F (264.5 Btu per Ib) Available and useful output a. Electric energy: 3500 kw X 3413 Btu per kwhr = 11,940,000 Btu per br b. Heat added to feedwater: 350,000 Ib per hr X (264.5 - 153) Btu per Ib = 39,025,000 Btu per hr b, or 50,965,000 Btu per hr Total available and useful output = a Input Fuel: 80,400,000 Btu per hr Air: Considered free and hence no charge for Btu Total actual input: 80,400,000 Btu per hr Thermal efficiency of gas turbine and its heat-recovery equipment = 50,965,000/80,400,000 = 63.4 per cent

+


u 4 0 J

n

c

863

864


Average actual operat,ing experience for the first four years has been better than original calculat,ions on which the instalhtion was based. Heat recovery from the exhaust gases of a 5000-kw unit of this type, in reasonably sized feedwater heaters or exhaust-heat recovery boilers, is approximately 45 million Btu per hr. When using an exhaust-heat recovery boiler, this is equivalent to 35,000 to 45,000 Ib of steam per hour a t pressures u p to 200 psig. Gas-turbine Cycles. Many different cycle arrangements have been considered for the gas turbine. There is the open cycle, in which atmospheric air is used once in its passage through the compressor and turbine, the closed cycle where the same atmospheric air is used over and over and is heated from an external source, and combinations of the two in varying degrees. The efficiency of any of these cycles will depend upon the extent t o which intercooling, regeneration, and other refinements are employed. When all factors are considered, the open-cycle arrangement has many very desirable features. The Simple Cycle. Figure 16.22 is a cross section of the simple opencycle unit indicated schematically in Fig. 16.21. The simple-cycle gas turbine is suitable for applications where a unit is reouired for stand-by service, such as to supplement a hydro system or to take peak loads, There fuel costs are low, or where there is a use for heat energy which can he recovered from the exhaust gases. The Regenerative Cycle. As can be seen in Fig. 16.23, a regenerator is an int,egral part of this gas-turbine cycle, recovering a portion of the heat in the exhaust gases and resulting in a gas turbinogenerator efficiency of approximately 22.5 per cent compared to 17 per cent without the regenerator. A cross section of this unit is shown in Fig. 16.24. When used xvith a regenerator, temperature of the exhaust stack gases is 566 F. The heat which can be recovered for feedwater heating or produring steam in exhaust-heat recovery boilers is somewhat lower than for the simplocycle unit. On the other hand, if an application requires

_. ... CHAMBER I.

COMPRESSOR

r

TURBINE

3500 K Y

FIG. 16.23

Schematic diogrom of regenerotive-cycle gar-turbine unit.

866


greater quantities of exhaust heat, the regenerator can be omitted and the low-pressure-turbine exhaust gases can be supplied directly t o exhaustheat recovery boilers or other exhaust-heat recovery equipment. APPLICATIONS

Where continuous-process operations are a must, the gas turbine with its quick-starting characteristics can be used for emergency or for peak load service in plants where the normal daily plant load is supplied from other sources. Now for a brief summary of some of the ways these gas-turbine units can be applied t o help supply the normal requirements of industrial plants for steam and electric power. Exhaust-heat Recovery Boilers-No Supplementary Firing. Figure 16.2 shows the relation between the gas turbine-generator kilowatt output and the steam output from the exhaust-heat recovery boilers and immediately suggests an application for steam generation. The arrangement for one mill requiring both process steam and electric power is shown in Fig. 16.25. Heat in the gasturbine exhaust gases can be recovered for the generation of steam in two boilers a t two different pressures-175 and 40 psig. From the last boiler, the exhaust gases pass through an economizer t o heat feed water for the boilers before exhausting to the stack. This arrangement would provide a total of 39,000 lb of steam per hour in addition to 3500 kw of electric energy. This is a ratio of approximstely 11 Ib of process steam for each kilowatthour. Over-all gas-turbine plant-cycle efficiency for this steam and power generation is approximately 60 per cent. These requirements could also be supplied by a double automaticextraction condensing steam-turbine plant. Over-all plant thermal efficiency of the steam-turbine cycle would be approximately 52 per cent.

lurbim o d m l

FIG. 16.25

Schematic diagram of a simple-cycle gar-turbine unit with exhaust-heal recovery boilers and economizer to recover heat from turbine exhaust gases.


867

This is lower than the gas-turbine-plant efficiency for this particular appliration because the ratio of process steam to electric power is low and a high percentage of the electric energy must be produced by the condensing section of the steam turbine. If the process-steam requirements were appreciably greater, then the efficiencies of the two cycles n.ould be comparable. FUEL

3500 K W

STACK STEAM TO PROCESS

I

t I

r

~

SUPPLEMENTARY FUEL NOZZLE

FIG. 16.26

t

BOILER

t

80 F

Gas turbine and exhouit-heat recovery boiler with supplementary fuel firing.

Exhaust-heat Recovery Boilers with Supplementary Fuel Firing. When required, steam flow from the exhaust-heat recovery boilers can be increased by supplementary fuel firing. I n this case of supplementary firing, heat recovered from the gasturbine exhaust gases is supplemented by heat from additional fuel burned in the turbine exhaust gases between the turbine and the exhaust-heat recovery boiler. The additional cost of providing for this supplementary fuel firing is low compared with the greatly increased flexibility of operation it affords. A schematic diagram for such an arrangement is shown in Fig. 16.26. This shows the exhaust-heat recovery boiler with the fuel nozzles for supplementary firing located just ahead of the boiler. Because of the large amount of excess air (approximately 500 per cent) used in the gas-turbine cycle, no additional air is required for combustion of this supplementary fuel. This further simplifies the installation and operation because forced draft fans and their attendent controls are not required. Depending on the balance between electric power and steam requirements, supplementary firing might be used continuously t o increase the

F

1

-

8OF

AIR IN

60

AIR I N _

COMPRESSOR

TURBINE

EXHAUST G A S W E

7260 KW

Ill

GENERATOR

3610 KW

GENERATOR

3620 KW

WbSTE

HEAT BOILER

FIG. 16.27 Schematic diagram showing gas turbine with exhaust-heat recovery boiler to "top" steam-turbine plant.

STEAM

FEEDWATER.

TURBINE

LXHAUST &45.,880E

I

STACK

STEAM

410 550' F

m D

OD

$TEAM AND POWER GENERATION

869

output of the exhaust-heat recovery boilers t o relieve the demand on conventional fuel-fired boilers. Supplementary firing might be used only t o supply occasional peaks in steam load. I n still other applications, supplementary firing might he used during an outage of conventional fuel-fired boilers. The output of the simple, lorn-cost exhaust-heat recovery boilers can easily be doubled or tripledsay, from 40,000 to 80,000 or 120,000 lh per hr steam flow-by supplementary firing. This being the case, the capacity of fuel-fired boilers normally maintained as stand-by can he materially reduced, with a corresponding reduction in first cost and fixed charges. Exhaust-heat Recovery Boilers Supplying Steom for “Topping” a Steam-turbine Plant. I n plants requiring additional power generation,

the gas turbine, with exhaust-heat recovery boilers, can be used to top low-pressure steam turbines. In plants where it is necessary to generate electric power but where there are no requirements for process steam, a plant-cycle arrangement such as shown in Fig. 16.27 will operate at an efficiency which is apprecia.hly higher than could be realized with a straight 400-psig steam-turbine plant. This is in effect a “binary ”-cycle power plant. Such a “topping” arrangement might he particularly attractive in plants where additional electric power is required, where the existing steam turbines are in good condition hut where the existing boiler plant must be supplemented or replaced. Turbine Exhaust Gases Con Be Used a s Combustion Air for Fuel-fired Boilers. High gas-turbine-cycle efficiencies can also be realized in plants

where the combustion-gas-turbine exhaust gases can be used as comhustion air for fuel-fired boilers. Because of the large amount of excess air used in the gas-turbine cycle to limit temperatures, the turhine exhaust contains approximately 17 per cent oxygen compared with 21 per cent for free air. Since so much oxygen is left in the exhaust gases, they can be used as preheated combustion air for fuel-fired hoilers which are not equipped with air heaters. One 5000-kw gas turbine can supply the combustion air required for a fuel-fired boiler having an output of approximately 250,000 Ib of steam per hour. This cycle arrangement is shown in Fig. 16.28. When the gas-turbine exhaust gases can be utilized as combustion air for fuel-fired boilers equipped with economizers, the gas-turbine plantcycle efficiency can be as high as 80 per cent when credited with heat in the exhaust gases. For the same percentage excess air in the fuel-fired boilers, boiler stack losses will be slightly higher when gas-turbine exhaust gases are used for combustion air. Exhaust Heat To Heat Feedwater for a Steam Plant. I n the first installat,ion for power generation, Fig. 16.29, additional heating of feed

070


water for an existing steam plant was required. Gas-turbine powerplant-cycle efficiency for this installation is approximately 65 per cent when credited with heat recovered from the exhaust gases (see Gas-turbine Power-plant-cycle Efficiency, page 862). Heat recovery from the exhaust gases of a 5000-kw unit of this type, in reasonably sized feedwater heaters or exhaust-heat recovery boilers, is approximately 40 million to 45 million Btu per hr. Exhaust Heat for Direct Drying of Various Materials. Still another application is the use of the gas-turbine exhaust gases for drying fiherboard, gypsum board, the kiln drying of lumber and other such material. Because of the high efficiency of fuel combustion in the gas turbine, turbine exhaust gases can he used for the direct drying of many products without danger of contamination. Although exhaust-gas temperature, as it leaves the turbine, may he too high for most of these applications, the proper drying temperature can be

FIG. 16.28 Schematic diagram showing fuel-fired steom boilers.

(I

g a r turbine supplying combustion air for


FIG. 16.29 Two 4000-kw gas-tvrbine units installed for exhourt being utilized for feedwoter heating.

87 I

power generation, with heat in

automatically maintained by mixing atmospheric air wit,h t,he t,urbine exhaust gases for temperature control as required. Humic1it.y of the drying air can also be easily controlled by spraying water into the hot exhaust gases when required. Ikpending upon the physical arrangement of the equipment, within llle plant, the gases leaving the exhaust-heat recovery boilers could also be used for drying. FUELS

Much effort is being espended at the present time to develop means of using a wide variety of fuels for the combustion gas turbine. Based on experience t o date, natural gas or distillate oils are excellent, fuels. With some mirior design changes, the gas-turbine units now in production can be adapted to burn 500 Btu per cu f t of gas or even blast-furnace gas having a heating value of approximately 90 Btu per cu it of gas.

872


At the present time, the steam-turbine plarit can utilize a much wider variety of fuels. This is a handicap for the gas turbine i n areas where the cost of suitable fuel is higher thaii that of fuels for other power plants, This handicap is partially overcome by the expected greater simplicity and lower installed cost for the gas-turbine plant or the low requirements of the gas-turbine plant for cooling water and high-quality boiler feed water. SELECTION OF PROPER TURBINE- STEAM OR G A S

When the gas turbine is credited with the heat in it,s eshaust gases which cau bc recovered for various uses throughout the plant, its cycle i as high as 70 per cent or more. This is comparable to efficiency c a ~ he the cycle efficiency of a good noiicondensing steam-turbine plarit exhausting steam for process use and generating by-product electric power. How then can we select the proper cycle:’ Refer again t o Fig. 1G.2. These curves show that steam turbines or gas turbines with supplemeutary fired exhaust-heat recovery boilers may be considered when the ratio of electric power to process steam is lon- say, 2500 to 5000 k w per 100,000 Ib per hr steam flow, which is 20 t o 40 Ih of process steam per kvhr. Gas turhines should be consi-lered when thc ratio of electric power to process steam is high^-. say, 8000 to 10,000 kiv per 100,000 lb per hr steam flow, which is 10 to 12 Ib of process steam per kwhr I n the intermediate range of process-steam requirements, the condensing-extract,ion turbine, or the gas turbine ut,ilizirig erhaust-heat recovery boilers with supplementary firing, or a combination of the two may be sclected. The requirement,s of process steam per kilowatthour vary n-idely in differcut mills. The over-all average is perhaps no more t,han 25 Ih per kwhr, and 11-ith the rapid increase in the use of elcctric power, the ratio is likely t o decrease still more in the future. In some new mills, and part.icularly in the expansion oC misting mills, the ratio of process-steam t o electric-power requiremeiits may be quite low A Ion. ratio favors the gas t,urbine. GAS TURBINE VS. STEAM TURBINE

Let us make a few comparisons between the gas-turbine and the steamturbine plant. Initial Cost. From available data, it, is estimated t,hat the cost of a nev gas-turbine plant compares favorably with steam-turbine plants utilizing units of comparable kiv rating.

STEAM AN0 POWER GENERATION

073

However, each case should be carefully studied by competent engineers so that accurate and complete installation and operating costs are estab-

lished. Wit,h such information the proper selection of the source of power can he made whether this is purchased power or local generation or a combination of the two. Control. The gas-turbine plant has one simple control panel. The simplicity of this plant makes complete automatic control possible. Efficiency. When properly applied, both steam and gas turbines operate with high cycle efficiencies. Reliability and Maintenance. A total of over 500,000 hr of operation has been accumulated on the gas turbines now in service Even though this is equivalent to more than 50 years of operation, it is realized that an accurate prediction of reliability and maintenance cannot be based on this relatively short period of operation. However, the designers are confident that the gas-turbine plant will compare favorably with the steamturbine plant. It might be well to point out here that the combustion gas turbines for commercial laud applications are designed with safety factors comparable t o those used in steam-turbine design so far as long life and low maintenance are concerned. DIESEL-ENGINE POWER-PLANT CYCLE

FOR STRAIGHT POWER GENERATION

The cycle efficiency for diesel-engine plants varies with such things as the compression ratio, the weight of air used per pound of fuel, etc. Quoted full-load efficienciesfor power plants designed for the generation of electric power only range from perhaps 25 up to 35 or 40 per cent. An efficiency of 35 or 40 per cent is very good for small power-plant units and is higher than realized in small steam power plants used for the generation of electric power only. Even though the diesel plant may have higher efficiency, the cost (per million Rtu) of fuel suitable for the diesel plant may be two or more times the cost of fuels suitable for the steam plant. This is one of several factors which minimizes the use of diesel plants. However, the diesel plant is often selected for very small generating stations. TO SUPPLY HEAT FOR PROCESS

Heat recoverable, for normal process use, is relatively low in the diesel power plant. This is true because, in addition to the exhaust loss, the loss to jacket water is high, percentagewise, in diesel engines. The tem-

074


perature of jacket water is so low that this heat energy cannot often be utilized and hence is lost as was the case of condenser losses with a condensing steam turbine. Because of the jacket-water loss, maximum practical cycle efficiency for a diesel plant with exhaust-heat recovery will usually be lower than for steam- or gas-turbine power plants used to supply both electric power and process heat. This, coupled with the high cost of diesel fuels, usually means high over-all fuel costs for the diesel plant. Internal-combustion engines generate only about 5 per cent of the total industrial power generation. For that reason their application will not be discussed in detail here. PLANT CONSTRUCTION, INVESTMENT, A N D OPERATING COST DETAILS

If straight condensing turbines, gas turbines without exhaust-heat recovery apparatus, or diesel engines must be used (when plant requires little, if any, process steam or other forms of heat energy), the cost of generating small blocks of electric power in an industrial plant is usually higher than the cost of power purchased from the utility, except in isolated cases. For this reason, information in this section will be limited pri-

20

40 60 CENTS PER MILLION BTU

80

FIG. 16.30 Comparative costs per million Bhr for various fuels.


075

marily to a discussion of the cost of power generation in conjunction with process heat requirements. Power-plant-cycle efficiency is a direct measure of over-all fuel usage for plants generating all their power. I n an industrial plant where heat energy in the gas-turbine exhaust gases or in the exhaust and/or extraction steam from steam turbines is required in process, the fuel required (per kilowatthour) for generating electric power ranges as low as one-third the fuel used by straight condensing power plants used for power generation only. Comparative costs, in cents per million Btu, for the more common industrial power plant fuels are plotted on Fig. 16.30. POWER-GENERATION COSTS-STEAM PLANTS

Estimated costs of generating electric power in plants ranging from 1000 to 10,000 kw are shown in Fig. 16.31. These curves show the cost of power generation when condensing steam turbines are used (no steam required for process) and when noncondensing or automatic-extractiontype steam turbines exhausting and/or extracting steam for use in process are used. Even though over-all power-plant-cycle efficiencies are high (fuel cost low) when generating “by-product” electric power, the actual total cost of by-product power may be higher than purchased power. For small plants, labor costs (mils per kilowatthour) are high and sometimes raise the cost of by-product power above the cost of purchased power. Where enough by-product generation is involved to keep labor costs down and where the plant factor is high enough to minimize fixed charges (mils per kilowatthour), by-product power generation within the industrial plant is usually economical. The curves of Fig. 16.31 show the effect of rating and plant factor on the cost of power generation and indicate relative costs of by-product and “condensing” power generation. It can he seen that fixed charges, fuel, and labor are the large items of cost. Fixed charges and labor increase very rapidly as the installed generating capacity and the plant factor decrease. This emphasizes the importance of a careful study of the present and predicted future steam and electric-power requirements. By-product power costs shown in Fig. 16.31 are based on the incremental cost of installing and operating power boilers and noncondensing steam turbines t o generate electric power when exhausting the steam required for process vs. the cost of installing and operating low-pressure boilers t o supply process steam when electric power is purchased from the utility. I n other words, low-pressure boilers would he required and their initial cost (fixed charges) would be chargeable t o process requirements even though no power-generating equipment is installed in the industrial


878


plant. The cost chargeable to power generatipn would then be the added cost of pon-er boilers over low-pressure process steam boilers plus the cost of installing the turbine-generator equipment. Labor charged to power gennration would likewise he the additional boiler-room labor (if any) and the turbine-room labor required to operate the turbine generator. Fuel costs charged to power generation are based on the generating unit being credited with all heat in the turbine exhaust steam. “Condensing” power costs shown in Fig. 16.31 are based on straight condensing steam-turhine plants such as would he used to generate electric power in plank where process steam is not required. I n these costs, fixed charges are based on t,he total initial cost of the power plant, including full cost of the boiler plant. Other costs, likewise, are total power-plant operating rosts. E ~ e though n t,he cost of “condensing” kilowatts is appreciably higher than by-product kilowatts, an industrial plant which supplements power purchased from the utility with by-product gencration within the plant sometimes finds it economical to install automatic-extraction condensing turbines. With this type of t,nrhine, normal operation can be to generate by-product power only with minimum steam flow t o the exhaust (steam

PLANT FACTOR - PERCENT CONDENSING

PLANT FACTOR-PERCENT NWCONDENSING ‘‘BY PRODUCT’’

FIG. 16.32 Ertirnated total cost of generating electric power. of rating, plant factor, and labor on over-all costs.

Curves show the effect


FIG. 16.33

879

GENERATOR RAT1NG - K W I ASSUMED SAME AS PLANT RATING) Estimated component costs of power generation in plonh operating a t 50 per

cent plant fac1or.

condenser), The turbine can he controlled, automatically or manually, to generate “condensing” kilowatts only during short peaks in plant load t o eliminate or minimize short peaks on the power company. Depending upon the nature of the load and the rate structure of the power rompittry, the savirrg in power costs due t o reduced demand charges may more than offset the added cost of “condetrsirrg” power generation required for this purpose. The total (.osts of “condensing” and by-product power for steamturbine units are shown in Fig. 16.32. As you i d note, this is a summary

880


of costs taken from Fig. 16.31 for steam-turbine units rated 1000 to 10,000 kw. The bar charts on Fig. 16.33 are based on operation at 50 par cent plant factor and show the various elements of generating costs. This ix also a summary from Fig. 16.31 for steam-turbine units rated 1000 to 10,000 kw. Typical values for labor, heat rates, maintenance and investment costs for steam-turbine plants as assumed for plotting the curves of Fig. 16.31 are given in Tables 16.7 to 16.9. The following labor costs are chargeable to power generation: TABLE 16.7 Typical Steam-turbine Power-plant Heat Rates'

Plant-unit r d n g , kw..

Steam conditions.

.................

....................

Number of faedxder h e d o n . .

1,000

2,000

400 pig-750

F2 in. H(

5,000

10,000

........

At full b d i Plant heat mte.. Plant efficiency.. At three-quarter lood: Plant hsot rote.. Plant efliciemy.. At half load: Plod h o d rate.. Plant efficiency

................. ................. ................. .................

21.300 16%

18.200 18.7%

14,900 22.9%

13.900 24.6%

22,100 15.4%

18,800 18.1%

15.300 22.3%

14.300 23.8%

................. ...................

24,800 13.7%

21.200 16.1%

16,900 20.20/,

15.700 21.70/,

1,000

2.000

5,000

10.000

Plant-unit rating, kw..

.................

I

At full load: Plant heat rote.. Plant sffirlency.. At three-quarter load: Plant heat mte.. Plont efficiency.. At holf lood: Plant heat rote..

Plant efficiency..

................. ................. ................. .................

5500 62%

5200 65.6%

4700 72.6%

4500 76%

5560 61.4%

5260 65%

4750 71.9%

4550 75 %

................. .................

5850 58.5%

5600 61%

5000 68.2%

4800 71 %

I

* Btu per net kwhr.

S T E M AND POWER GENERAllON

881

Condensing plant: Four operators total per shift, or a total of sixteen on the payroll. Noncondensing "by-product" plent: Two operators per shift, or a total of eight on the payroll. An average cost of $5000 per operator w&s used. The dashed line on each curve indicates total cost if half the number of operators listed above are required so that labor costs can be easily interpolated for any number of operators desired. The typical heat rates shown in Table 16.7 were used as a basis for fuel costs. The fixed charges tabulated in Table 16.8 are generally applicable.

1

TABLE 16.8 Total Investment Chargeable to Power Generation

................. ...........

Plant-unit rding, kw.. Condensing, dollars per kw.. Noncondaring, dollars per k w . .

........

1000

$360

2000 $320

$220

$200

5000 $280 $175

10.000

$260 $160

The maintenance charges.listed in Table 16.9 were included as typical values chargeable to power generation. TABLE 16.9

-

Maintenance and Miscellaneous Costs Choraeable to Power Generotion ~~~

AontYnitroting.

Condeluing

Noncondenring*

Im

~

.......................

Dollan per yeor Per cent of inilia1 in"~lm.nt Mils per kwhr 01 100% plont focmr Dollon per yeor Per cent of initial invntmant Mils per kwhr at 100% p l m l focmr

1

_

_

1

1

-

5,000

10.000

$30,000 4.7% 1.7 mils

$45,000 3.2%

$55,000

1 .o mils

2.1% 0.6 mils

$8.000 3.6%

$1 2,000

0.P mill

0.7 mils

$1 8.000 2.1% 0.4 mils

$22,000 1.4% 0.25 mils

1.000

$20,000 5.5% 2.3 mils

I !

2,000

3%

Assumed 40 per cent of the total for B condensing plant would be chargeable to the power-generation portion of a noncondensing plant.

STEAM-POWER-PLANT INVESTMENT COSTS How much does a power-generating station cost? The cost of installing a new industrial power plant or of expanding an existing power plant varies widely. No attempt will be made to cover detailed investment costs here a t this time, but a few general comments will be given as a guide.

882


Dollars per kilowatt of installed capacity is a figure commonly used in discussing power-plant costs. I n utility plants, producing kilowatts as their only salable product, the relation of initial cost t o kilowatts installed is perhaps a reasonable yardstick for comparing costs. This same yardstick is often used in discussing industrial power plant costs even though t.he size, and henre the cost, of an industrial power plant may bear little relation to its kilowatt rating. For instance, a utility power plant and an industrial plant, each having the same kilowatt rating, might have wide variations in the requirements for various power-plant apparatus as indicated in Table 16.10. TABLE 16.10 Approximate Range of Ratings for Typical Power-plant Apparatus for a 10,000-kw Plant Plant opparohm

Turbine generator, kw Powor boilerlrl, Ib per hr Feed-water make-up and treolmont, gpm Boiler feed-water pumps. gpm

Utility plant

10.000 100-1 25,000

4-10 200-250

Industrial power plant

10.000

200-500.000 400-1000 400-1000

Boiler fans, steam piping, and other apparatus vary widely as well as the boilers, water treatment, feed-water pumps, etc., included in Table 16.10. The boiler plant is a large item in power-plant costs. With the wide variation in steam flows and other industrial power-plant requirements, the difficulty of using a value of “dollars per kilowatt installed” in estimating or even comparing the costs of industrial power plants is apparent unless the plant is described in detail. Now for the actual cost of power plants-utility plants range in cost from perhaps as low as $150 per kw t o well over $200 per kw installed. Industrial plants, keeping in mind the above warning on the use of dollars per kw, range in cost from something less than $200 per kw t o more than $300 per kw installed. Pouer-plant costs vary widely, not only because of different industrial power-plant requirements hut also because of the power-plant facilities, buildings, etc., that may already be available in the case of plant eupansion, and foundation and building requirements, etc., in the case of new plants. Because of these many variables, it is difficult t o estimate costs without carefully analyzing each plant requirement and then adding various component costs t o arrive a t total cost. This total cost can be converted to dollars per kilowatt of course. As thp rate of return on initial investment (per cent. fixed charges) is increased, initial cost becomes of increasing importance. For instance,


883

with gross fixed charges of 15 per cent, $50 per kw installed cost represents 1 mil per kwhr in power cost if the turbinegenerator is operated 85 per cent of the time (7500 hr per year) a t full load. Other rates of fixed charges mould of course change the cost per kwhr proportionally. Cost per kwhr would likewise go up as plant factor goes down. For a plant having an average annual load of 10,000 kw, 1 mil per kwhr reduction in power costs means almost $100,000 per year.

Chapto- 17

by Donald S. Brereton

Load- and Cost-estimating Data ~

~~~~

~

.

-

This chapter is divided into two parts, one dealing with load-estimatin - ~data and the other with cost-estimating data. These two sections are coupled together to assist the power-system engineer to determine more easily the power requirements of new or additional facilities in his plant and then determine the most economical arrangement and selection of the apparatus in accordance with the sound and safe principles of powersystem design as given in the preceding chapters. It is realized that some risk is taken when information concerning the cost of electrical apparatus is published. It is hoped that when these data are used it will always be recognized that they are for estimating purposes and not for firm pricing. It is believed that the fluctuations in the prices of the apparatus listed that will occur after the publication of this handbook will not materially reduce the usefulness of this chapter. Therefore, i t is not the intention here to give an exact price of a transformer or a motor with its control but to illustrate the relative costs of the various components that make up the power system. ~~~~~~~~

LOAD-ESTIMATING DATA*

A close estimate of the location and demands of the electric load in an industrial plant is a major prerequisite for the proper design of the plant electric power system, for the success or failure of the power system lies in its ability t o meet the load requirements. Unfortunately, load estimating is seldom easy, especially in new installations, since the power system is usually being designed while the utilization equipment is also in the planning stage. Certainly, in the initial stages of planning, the exact machine ratings will seldom be known and very little can be predicted as to their probable operating schedules. The problem is one of making an * R. M. Wilson, General Electric Company, formerly of the Industrial Engmeering Section, made a substantial contribution to the material on pagee 884 to 892. 884

LOAD- AND COST-ESTIMTING DATA

885

estimate of the loads from what little information is available on the proposed plant and of supplementing these general data with load data on similar existing plants. Discussions with the plant engineers and operating personnel of the proposed plant are essential to obtain as much information as possible. General data on similar plants can be obtained from discussions with plant engineers and operating personnel of those plants. In addition to discussions with the industrial engineers, it may be helpful to consult the utility power engineers for their estimates of the load of similar industrial plants on their system. In estimating the load demand it is usually best to consider lighting and power loads separately and to combine these later to determine the demand in any one area, for in present practice light and power loads are most often taken from the same load-center substation. The load in a manufacturing plant is commonly expressed in terms of load density (volt-amperes per square foot). Different sections of the plant must nearly always he considered a t different load densities, for there will be separate areas for offices, machine shops, storage, etc. LIGHTING-LOAD ESTIMATES

Lighting loads are easier to estimate than power loads, but will represent only a small portion of the total power requirements. The load density for lighting may be estimated by considering three factors: intensity of illumination, type of lighting (mercury, fluorescent, or incandescent), and height of fixtures. Table 17.1 shows the effect of these factors on power requirements for factory lighting. For office lighting the factors should be increased by about 30 per cent. The intensity of illumination multiplied by the lighting-demand factor given in the table gives the power demand in volt-amperes per square foot of floor area. TABLE 17.1

Power Requirements for lighting* Lighting-demand fmh-rt

fixture height, f l

Incondorcent

14 14-35 35-50

I*lOW

0.12 0.13 0.15

Fluorexent

0.060 0.065 0.070

LOAD- AND COST-ESTIMATING DATA

886

For example, for a maintained illumination of 50 foot-candles in a factory area using fluorescent lighting. which is an illumination level found in general use, the lighting requires about 3 va per sq f t (50 X 0.06 = 3). Mercury lamps are the most efficient and are most often applied in high bays of manufacturing buildings where the fixtures can be mounted quite a distance above the eye level and the intensity of the source will not be unpleasant to the eyes. Fluorescent lamps are finding general use in all areas, with the exception of areas where close color definition is required. A combination of mercury and incandescent lamps is often used t o provide better color discrimination than can be provided with mercury lamps alone. In general, the lighting-load density will vary from 1 t o 8 va per sq ft, the lowest illumination being in storage areas, the highest in areas of precision manufacturing or assembly of small parts. Plant security may place emphasis on outdoor lighting; however, all outdoor lighting seldom amounts to more than about 10 per cent of the total lighting load, and unless otherwise indicated, an allowance of this additional percentage t o indoor lighting will be more than sufficient to take care of outdoor lighting. Lighting requirements will vary by industry, and Table 17.2 may he helpful in determining what some of these variations may be. The diversity factor of the lighting load mill be low, and the demand of the lighting connecting to any one load-center substation should he considered a t 100 per cent. I n terms of total plant connected lighting load, a combined demand and diversity factor of 80 per cent or higher should be used for lighting. TABLE 17.2

lighting Requirements in Various Industries

Industry Steel foundries.. Steel rolling mills; oil refining.. H e m e~ k t r i ~ a eqvipment; l wire drawing.. Avtomobilc equipmcntj baking.. Mochinc parts. Automobile osembly and p o r h . .

Lighting in Per Cent of Total Connected Load 1-3

................................... ....................... 3-5 ............. 5-8 ...................... 8-10 .................................... 10-1 5 ..................... 15-25

POWER ESTIMATES

Estimating the power load is considerably more difficult than estimating the lighting load. One way of starting on this is first to consider the major components of the power load, such as large synchronous motors and furnaces. Such loads exert the maximum influence on the power demand and are often decided early in preliminary planning since their rating and schedule of operation can be more closely predicted than can that of the smaller loads. Many of these loads will be applied a t the


887

plant primary voltage (above 600 volts), although some may be applied a t lower voltages. For large induction and 0.8-power factor synchronous motors the kva demand may be assumed t o equal the motor horsepower rating, but for 1.0-power factor synchronous motors the kva demand should he taken as 0.8 times the horsepower rating. Often the larger machines will operate alternately, and it is important to consider a diversity factor t o prevent overdesigning the system. The use of these factors is discussed in later paragraphs. A thorough knowledge of the type of manufacturing activity will aid considerably in estimating the demand of the smaller machines on a loaddensity basis since existing plants may be surveyed to obtain a suitable estimate of the load of those manufacturing operations which are similar to those planned in the new plant. These load-density estimates are especially useful for equipment operated a t 600 volts or below. In any one plant the density may vary from zero in storage areas to 35 or more va per sq ft in some machine-tool areas, spinning rooms, etc., or similar areas of concentrated loads. Table 17.3 gives representative load densities for different types of industrial plants. The table should be used for preliminary estimating only since the sine of the plant and its processes will vary considerably within a given industry category. TABLE 17.3

Estimated Load Densi:ies in Representative Industries Volt-ampere DemandLight ond Power, Va per Sq Ft

Typo of Plant Best-wgor factory and reflncry.. Papor mills.. Textile mills; engine builders. Cigarette manufacturing. General manufactwing; chemicals; electronic equipment. Small-appliance monufacfuring; machine repair l o p . . Lamp monufaauring.. Smd-device manufacturing..

..................... ...................................... ......................... ............................

.............................. .........................

19 14

I2 I1

... 10 .... 755

5 3>5

When used with good judgment, Table 17.4 can also be used to estimate the power requirements of a new industrial plant. This table shows the number of kilowatthours required per unit of output and is a general average of many industrial plants making the same type of product. When applied to any particular plant, the information, therefore, may not nearly represent artual conditions, for the degree of electrification in the plant will have a considerable bearing on the number of kilowatthours used. In order to translate these figures into terms of system capacity, it mill be necessary to know the output of the plant and the number of hours in which manufacturing processes are in operation.


888

TABLE 17.4 Approximate Power Consumption of Industry Industry

I

............................... .................... Cernmt .................................... Food,horen ................................ Pope Wood pulp .............................. Poper ond board ......................... Iron (pig). ................................. Shon ..................................... Steel ...................................... Sugar; Beet .................................... Beet .................................... Cone (011 elechic).. ....................... Cone (dlelectric with ,teem mill drives). ....... Automobiles.

Butte...................................... Corpets and rugs (wool)..

.......................... ..............................

cone, refinery.. Tobocw Cigarettes. Ciao rs

...................................

Kwhr

I

Unit

1050 136 1480 22 144

Each

384

TOn Ton

474 25 472

217

26 154 220 125 23

200 I70 8100

1000 Ib 1000 rq yd Bard

1WO Ib

TO"

1000 pairs TO"

Ton of beets Ton of reflned sugar Ton of m w wgar T a of raw sugar Ton of melt

1,000,000 (cigarettes) Ton of tobacco

1,000,000

TOTAL LOAD ESTIMATES

After the loads of the individual machines or areas of a plant have been determined, it is necessary to combine these t o obtain a total demand. This kva demand determines the system capacity which must be provided and is obtained by the use of demand and diversity factors. It is not difficult to apply these factors in a load-center system of power distribution, for each load center serves a relatively small area. For example, on a branch feeder the connected load is multiplied by a demand factor to obtain feeder demand, and the sum of feeder demands is divided by a diversity factor to obtain the combined feeder or load-center substation demand. The selection of demand and diversity factors, like load density, is based on known conditions, experience, or similar operations in existing plants. Table 17.5 shows factors for the more commonly encountered types of manufacturing loads. As the table shows, the demand factor will vary considerably with different types of loads. For example, the demand factor of a group of motors driving a conveyer belt will approach 100 per cent, while the demand factor of a group of hand tools in a small furniture factory or machine shop might he only 30 per cent. A diversity factor of unity is often used t o provide ample system

889


capacity, since the margin provided in this way is soon used by load growth; and, although the sum of the maximum demands of the individual load centers and of the equipment applied a t the distribution voltage can be divided by the diversity factor to obtain the maximum demand on the distribution system, the usual practice is to provide a system adequate for a maximum demand obtained with a diversity factor of unity or to provide even more than 100 per cent capacity in the main system to take care of load centers which would be added in the future. TABLE 17.5

Demand Factors

Load

btimoting Demand Factor, Per Cent.

MOIOW

1. General pvrp~se; m d h s 1-1; crone elwoton; v a n t i l o t i ~ ~compreuona j pYmp,, rolling mill., atc 2. Semicontinuous p m n s n ; popsr mill*i refineries; rubber mills; elc. 3. Continuous operationii textile mills.. Electric ovenii h e d e n ond fvrnoces lndudionfum.cn Arcfurn.c.. Lighthg Arc w e l d H I Rerirl.ncewo1d.r

..........................................

........ ................................

.................................... ................................................... ........................................................ ............................................................ ........................................................ ...................................................

30 60 90 80 80 100 80 30 20

* Multiply connected load by this factor to obtain total plant dernsnd, rtssuming diveraity factor is unity. Another term which is commonly used is load factor. Once in a while i t may be possible to take advantage of a low load factor and select a smaller transformer with smaller cables than could be applied based on the peak load alone. However, the transformer and cables must be of adequate size to provide a satisfactory voltage under the peak load conditions. Selection of a safe load factor requires detailed knowledge of operating characteristics of the load and of the thermal capacity of the distribution equipment. Only in rare cases should a load factor be applied. For a further explanation of these factors refer to Fig. 17.1. The feeders on load center A operate as shown in the accompanying table. The actual demand measured on the load center is 800 kva, thus the diversity factor is 1000 + 800 = 1.25. In practice the load-center rating would usually be selected at lo00 kva, which provides for future load by using a diversity factor of unity. Assuming that load centers B, C , and D are the same as A, the demand on the main substation, also assuming no diversity between load centers, is 4 X 800 = 3200 kva. The main substation would likely he rated 3750 kva, or even larger if future plans seem to dictate this, for the main substation is not so easily or economically expanded as is the additional kva available in the load-center area which can be obtained by adding a relatively small load-center substation.

LOAD- A N D COST-ESTIMTING DATA

890

MAIN SUBSTATION

DEMAND METER

13.8 K V

I CENTER

'gjJJ CONNECTED

LOAD

Feeder NO.

1

100

a0

100 300

80

3 4

125 I25 500

600

500

84

Combined

1350

2

FIG. 17.1

I

60

1000 lorsumn M) diversity)

Typical distribution-rprem arrangement.

Actual load data as recorded for groups of plants in various industries are given in Table 17.6. The comparison of group averages with the range of readings necessary to include all plants of each,group indicates the wide variation found even in similar plants and the desirability of obtaining specific data rather than relying on general estimates. Examples. Assume that a two-kiln cement plant is t o he constructed which will produce approximately 4000 barrels of standard portland


TABLE 17.6 Recorded Load Data for Demand, "a per rq ff

891

;roup of Plant! Annual kwhr per *q

ff

jurnrnary) Annual kwhr per vm demond

__

Type of plant ~

Average

Range

veroge

Range

Range

~

..................... ................... .....................

10.0 10.3 9.9

613 3-20

33.7 25.8 32.4

14-54 11-67

3.3 2.4 3.3

2-4 1-4

2-12 6-13

4-46 3-27

4.2 2.8 6.0 5.4 5.4 3.0 2.9 2.4

2-6 2-5

2-7

19.5 30.0 25.0 24.2 10.9 8.0 20.5 9.5

5-53 13-64

..................

4.8 8.7 4.0 4.5 2.0 3.0 7.4 3.6

.....

10.0

5-17 3-1 8

70.0 31 .O

8-1 95 5-50

7.0 5.2

Chcmicol Electronics. Foundry Lamp:

................. .............. ................. .............. Porcelain. ................... Printing.. .................... Small appliance. .............. General..

Wire works.. Boreworks Glass works..

Small devise General:

L a r g c o v e r 5000 k w . .

S m o l h n d e r 5000 k w

.......

9.8

.... ....

2-7

.... ....

2-1 3

-

.....

..... ..... .....

13-37

- -

4 4

1-7 1-5 1-19 1-27

-

cement per 24-hr day. From Table 17.4 the average cement plant used 22 kwhr per barrel of cement. Since a cement plant generally operated 24 hr per day, the system capacity must be at least (22 t 24) X 4000, or about 3700 kw. Assuming that the power factor is 95 per cent or better, the plant would require a system which will furnish continuously 3900 kva. Actually with an average use of 3900 kva, the system must require kva peaks somewhat greater; however, this type of operation is very steady, and the peak load may be only 10 or 20 per cent higher than the average. Thus the designer should consider a main substation rated 4500 kva or more t o take care of additional expansion. In another instance, suppose that a manufacturer proposes t o build small aircraft engines and wishes t o determine the power requirements in a new factory which will be approximately 500,000 sq ft in area, only a small portion of this being devoted t o office space. From Table 17.3 the kva demand for both light and power will be approximately 6000 kva, that is, 500,000 sq f t a t 12 va per sq ft. Because of the low demand and diversity factors in this type of operation, it is probable that the connected load-center-substation capacity will considerably exceed the main transformer capacity. First, a plant layout should be obtained on which can he spotted the major components of load. Some of these will be individual machine ratings, while others will be purely designations of load area and the type

892


of manufacturing in t,hat area. These loads can then be grouped into areas with a load of 500 to 1500 kva, taking into account the demand factors. From these groupings the load-center arrangement and ratings may be determined. SUMMARY

The load-estimating data contained herein is for general preliminary estimating only and shonld not be used when information on the specific plant for which a power system is being designed is available. However, these general data may be helpful in preliminary stages, and the application of such general principles and procedures as are given in the references listed may facilitate the design of a power system for a specific plant.

COST-ESTIMATING DATA Economics is one of the most important factors in power-Bystem engineering. There is often more than one way to satisfy a given set of requirements from a technical standpoint. The choice of the method used in such cases is largely determined by economics. The following data are presented solely for the purpose of making comparative system economic studies. Such data, under no circumstances should be used for the obtaining of appropriations, as market changes, product changes, etc., all affect the exact pricing of any particular equipment. To make proper economic comparisons requires the consideration of the installed cost of the power-system apparatus necessary to bring power from the source to the load. Problems of selection of voltage require the consideration of motors and control, particularly for the higher voltage motors and control which can operate directly at plant primary voltage. However, these approximate data are sufficiently accurate for nearly all general system comparison economic studies. INSTALLATION C O S T S

The data given in this section, except for the information on cable, do not include installation costs. These may vary widely,, depending upon working conditions, foundations required, whether the installation is new or old, whether production must be kept going while the change is made, upon labor rates, and many other factors. However, some installation figure must be assumed to get a proper over-all economic comparison between two systems. Where information known to be more accurate is not available, the following procedures have been used to obtain a reason-


893

able approximation for comparative purposes only. The following paragraphs describe accepted practices of estimating the cost of installation for different apparatus. Field-assembled Switching Equipment. Equipment that is assembled piecemeal in the field, such as standardized open switching stations, piecemeal substations, piecemeal switching equipment, should have 50 to 100 per cent added to the equipment cost to cover installation cost. This would not include foundations. Factory-assembled Switching Equipment. Factory-assembled equipment, such as load-center unit substations, metal-clad switchgear, motorcontrol centers, should have approximately 25 per cent of the cost of the equipment added for installation. This again does not cover foundations or special mounting platforms but merely the placing of the electrical equipment on an already available foundation, bolting it together, and making the electrical connections. Transformers. The installation of three-phase transformers involves the moving of the transformer into place and the making of the connections. To perform this installation function, approximately 10 per cent should be added to the cost of the three-phase transformer. Single-phase transformers require greater installation expense because there are three units to move and three units to interconnect on both the primary and secondary, as well as making the connections to the incoming and outgoing lines. The cost of the buses for the primary and secondary connections for making the delta or Y connection of the single-phase units may be included in the switchgear. However, there are still additional connections to be made. Therefore, the cost of a single-phase bank of transformers should be increased by 25 per cent to cover installation and making the connections between the transformers and lines or switchgear buses. If it is necessary to make a separate structure or if cable interconnections are necessary to form the Y or delta connection of the transformer hank, then the installation cost may he greater than 25 per cent. Motors and Control. For the purpose of system comparisons it should not, in general, be necessary to consider installation expenses in connection with motors and control to get an accurate picture of how motors and control affect the economics of a power system. For example, the difference between the installation cost of a 2300-volt motor and its control and a 4000-volt motor and its control would be nil. There may, however, be some difference between the installation cost when 440-volt motors and control are involved and it is compared with the cost of 2300-volt or 4000volt motors and control. In this case the installation cost of the motor would be essentially the same, but for smaller motors there may be an installation-cost advantage of installing the motor control in favor of 440 volts. On the other hand, for very large motors there may be an instal-

894


lation-cost advantage in installing the higher voltage motor starters. These factors may be considered. However, it is felt that they are not important enough, in general, to complicate the picture by considering them as they are of third- or fourth-order importance. Metal-enclosed Bus. The installation costs of metal-enclosed bus may vary widely, depending upon installation conditions. For estimating purposes the installation cost may he assumed to be in the range of 25 t o 50 per cent of the cost of the bus, including terminations, tapoffs, or bends. As an average estimating figure, one-third may be used for comparative purposes. PIECEMEAL-EQUIPMENT COSTS

Piecemeal-equipment costs are not given here because it would take too much space to list each item separately and would be too difficult to interpret. Furthermore, the purpose of this chapter is to cover the more general cases, i.e., those systems using modern factory-assembled equipment. Factory-assembled equipment is now almost universally employed in industrial plant power systems. The complication of presenting a simple economic story on piecemeal equipment is one of the major disadvantages of this type of equipment compared with factory-assembled switchgear, substations, or motor control. The real disadvantage is evident when the installed costs of each system are compared. The problem of piecemeal-equipment costs is further complicated by the very involved engineering and ordering and detailed expediting required, which are difficult to estimate accurately. Prices of factory-assembled equipment are published by the manufacturers, and the engineering and ordering costs are only a small percentage of the over-all cost and, therefore, are not nearly so significant as they are with piecemeal equipment. For roughly estimating the cost of a piecemeal substation or switchgear station equipment, it is necessary to include not only the large items like circuit breakers and transformers hut also all the dozens of miscellaneous items such as disconnecting switches, steel structures, insulators, buses, connectors, current transformers, potential transformers, fuses, fuse holders, etc. To obtain an accurate cost would require, first of all, a complete design for the substation. However, in estimating stages such design work is generally not feasible. As a guide in estimating the cost of a piecemeal switchgear equipment, take the cost of power circuit breakers rated 500 mva and up and increase that by 50 per cent and take the cost of power circuit breakers rated less than 500 mva and increase that by 100 per cent. This will generally be enough to include the costs of miscellaneous items mentioned previously. After taking the breaker


895

cost and increasing it 50 per cent, this total figure should be increased 50 to 100 per cent again t,o allow for installation. These substations vary so much that rules of thumb can be very misleading. Therefore, the only way to get an accurate cost is to determine each item required and its cost. One of t.he major problems in doing this under the pressure of estimating is to be sure that all items are included. With factory-assembled equipment those items are automatically included in the price, and thus the chances of error are reduced when using factory-assembled equipment as well as the engineering time required to design and estimate the job. Example 1 on page 896 is included t o elaborate some of these points for background information only. STANDARDIZED OPEN SWITCHING STATIONS

Standardized open-type switching stations or components of packaged substations are sometimes priced and sold as a unit. In such eases the manufacturer supplies all the equipment necessary for the functions to he performed by the station. To these no additions need be made to CDVCP structures, bus, disconnecting switches, etc., as is required for piecemeal substations when the parts are all purchased individually. ESTIMATING COST OF EQUIPMENT

In all estimating work it is desirable to increase the equipment cost, if net, about 10 per cent to obtain an “estimating” figure. This will then freely permit the addition of equipment already specified a t “estimating prires” with those given a t “net prices” in the various sections of apparatus handbooks. It is also sound practice t o use estimating figures which provide some degree of contingency. This percentage should be added to all parts of a piecemeal substation, since the prices of this equipment appear as net. A similar percentage figure is added in all cost data shown in this chapter.

EXAMPLES OF SYSTEM ECONOMIC COMPARISONS The following examples are principally based on using the data included on pages 906 to 924. I t is significant to point out that, when comparing piecemeal installations with factory-assembled equipment, all equipment components must be included in the piecemeal equipment. This will be home out in the first example where, if interconnect,ing buses, disconnecting switches, insulators, etc., were omitted, an entirely erroneous couclusion would be reached.

896


EXAMPLE 1

This example will compare an old-style piecemeal installation, Fig. 17.2, with a modern load-center installation, Fig. 17.3. The following procedure is used t o estimate the cost of the old-style system. Old-style Piecemeal System. The old-style piecemeal system, Fig. 17.2, has the following approximate cost. Switching Slructure. This portion of the system inchdes the three 13.8-kv power circuit breakers and the switching structure. The cost of one breaker is approximately $4200. To this should be added 10 per cent for estimating, making the estimated cost $4620 per breaker. To cover miscellaneous items such as the structure and switches, the estimated cost, of the breaker is increased by 100 per cent (breakers rated less than 500 mva), which brings a total estimated cost per breaker position of eqnipment for the switching station to $9240. If that figure is increased by 50 per cent to cover installation, the total cost per breaker installed in the station is $13,860. For the three breakers that would be approximately $41,580. Transformers. The 833-kva transformers of standard single-phase rating and design cost $4815 each. This includes the 10 per cent for estimating. The total estimated cost for three single-phase transformers would be $14,445. Adding 25 per cent for installation, the total installed cost of the 2500-kva banks would be approximately $18,060. Low-voltage Bus between Transformers and Switchgear. The estimated cost of 3000-amp low-voltage bus for outdoor installation is approximately $165 per ft. The total length of bus is assumed to be 24 ft, making the estimated cost $3960. The termination a t the indoor switchgear and the termination with bushings a t the transformers are assumed t o cost $495 each, making the total estimated cost of the bus and terminations $4950. To this is added an average installation figure of 3355 per cent, making the total installed cost of the bus $6600. Low-uollage Switchgear. The low-voltage switchgear consists of an incoming line breaker rated 100,000 amp interrupting rating, 4000 amp continuous, and from Table 17.9 the estimated cost of this breaker is $7800. The twenty 50,000-amp interrupting-capacity feeder breakers cost about $1600 apiece, making the total estimated cost of the lowvoltage switchgear $39,800, t o which should be added 25 per cent for installation, making the total installed cost of the secondary switchgear approximately $49,750. The total approximate cost of the transformers and secondary switchgear and interconnecting bus is $74,410. Cable. I n computing the total length of the secondary cables it is assumed that there is a 15-ft vertical rise in each cable from the switchgear


BP7

&* *-ti-

4 - a -

4-W-

-+


898

2-13.2KV INCOMING LINES

?$$

PLUG-IN BUS WAY /


/’ 2 5 0

MVA 7 5 0 KVA UNIT

f SUBSTATION

\

f

I

b-

L L

0

I

FIG. 17.3

Load-center system for cost comparison.

to the ceiling and that the cables are then run out t o the center of each row of busway. This requires a total of 7300 ft with an installed cost of $3.70 per f t , or a total installed cable cost of $27,010. Total Cost. Adding the above costs, the total installed cost of the oldstyle piecemeal system shown in Fig. 17.2 is $143,000. load-center System. The load-center system, Fig. 17.3, has the following approximate costs. Primary Switchgear. Referring to Table 17.8, the approximate cost of a 13.8-kv 250-mva interrupting capacity metal-clad feeder position at 1200-amp continuous rating is $7100 for either feeder or incoming line. Since there are four breakers, the total cost of the switchgear equipment is $28,400 t o which is added 25 per cent for installation, making the total approximate installed cost $35,500. Subslatiom. The four 750-kva substations have an estimated cost of approximately $23 per kva, as determined from Fig. 17.13. This cost includes six 25,000-amp interrupting-rating breakers per substation. Only five breakers per substation are required, and they are to be 15,000amp breakers instead of 25,000-amp breakers. A t $23 per kva the total cost of the 3000-kva substations would be 269,000, including six 25,000-amp feeder breakers. Deducting one feeder

F


899

breaker per substation a t $900 each will reduce the total estimated cost t.o approximately $65,400. Changing the remaining five 25,000-amp t o 15,000-amp breakers will reduce the cost of the 20 feeder breakers approximately $300 each or will reduce the total cost of the substations t o approximately $59,400. Adding 25 per cent for installation brings the approximate total installed cost to $74,250. Cable. The primary feeder cable total length, allowing 15 f t for vertical runs in each cable, would be 1020 f t of type ACV, No. 1/0 Awg, at $5 per foot installed (Table 17.12), or a total installed cost of $5100. The secondary cable total length is 1740 ft of type ACT’, 250 MCM, with an installed cost of $3.70 per f t (Table 17.13), making a total of $6438.

FIG. 17.4 Comparative cost of old-style and load-center systems and components.

900


Total Cost. Adding the component costs, the approximate total installed cost of this system is $121,288. For comparative purposes the costs of the various elements and the total costs are shown in bar chart form in Fig. 17.4. EXAMPLE 2

This example compares the cost of a radial, secondary selective, and secondary network system for a 3000-kva plant having an average load density of 10 va per sq ft. Radial System. The radial system shown in Fig. 17.5 has the following costs. The primary switchgear costs approximately $14,100 (from Table 17.8) plus installation cost of 25 per cent, making a total installed cost of $17,625. The primary switchgear is assumed t o be located about 400 ft from the edge of the building, thus the total length of primary cable is 2175 ft, and this 5-kv cable, 350 MCM interlocked armor, has an installed cost of $5.30 per ft (from Table 17.12), making a total of $11,530 for primary cable. The radial unit substations are assumed t o be 750 kva each. This is larger than is used in the network but is t o allow for diversity in loading in the various load areas. These substations in their basic form, shown on Fig. 17.13, cost about $21.50 per kva. The total cost of these suhstations, including six feeder breakers, is $96,750. However, only four feeder breakers are required. Therefore, deducting two feeder breakers per substation reduces the cost of the six units by $10,800, making a total cost of $85,950. When the 25 per cent is added for installation, the installed cost is $107,440. The secondary cable is assumed to he 250 MCM and runs from the substation up 25 f t and across t o the center of the busway. This involves 2100 f t of cable a t $3.70 per ft (from Table 17.13), or a total of $7770. The busway is assumed t o be 400-amp type FVK and has a total installed cost of about $9.50 per ft (this includes 25 per cent for installation). There is a total of 4800 f t , making the total installed cost for busway of $45,600. Adding all the above figures gives a total installed cost of the radial system of $189,965 or a total of approximately $63.30 per kva (on the 3000-kva base). Secondary Selective System. This system, Fig. 17.6, is similar to the radial system except for the normally open ties between pairs of loadcenter substations. This system has the same primary switchgear cost, ?e., $17,625; same primary cable cost, i.e., $11,530; same secondary feeder cable cost, i.e., $7770; and the same busway cost, i.e., $45,600. The unit substations will each have an additional 800-amp, 50,000-


901

3 5 0 MCM INTERLOCKED ARMOR 5 KV CAELE

AX-1-25

\

z

FEEDER BREAKERS PER SUB

# ' 4

T

LI

500 FT

FIG. 17.5

Radial recondory selective load-center system for cost comparison.

902


350 MCM INTERLOCKED 4RMOR

4160 V

5 KV CAELE

t'

4-300A

AK-1- 25

$

FEEDER BREAKERS

I t.

500 FT

600 FT

-

Ib l

FIG. 17.6 Secondary selective load-center system for cost comparison.


903

amp breaker for the tie circuit. These additional six breakers will increase the cost of the Substations approximately $9600 (from Table 17.9) or will increase the installed cost from $107,410 to $119,440 (107,440 9600 X 1.25 = $119,440). In addition there will be approximately 1800 ft of 350 MCM interlocked-armor tie cable a t an installed cost of $4.40 per ft (from Table 17.13). This is an additional $7920. The total installed cost of the secondary selective system is $209,885, or about $70 per kva. (Primary switchgear, primary cable, secondary cable, and busway are the same as for radial system.) Secondary Network System. The secondary network system, Fig. 17.7, has all transformers operating in parallel, and therefore diversity of loading in the various load areas can be taken care of by transferring load from one Substation to another. The example was chosen to use standard sizes of Substations to get a comparison that would be most favorable t o the network system. The network employs six 500-kva unit substations instead of six 750-kva unit substations as did the radial and secondary selective system. The primary switchgear, primary cable, secondary feeder, and plug-in bus costs are the same for all systems. The unit substations for this system cost about $26 per kva (total $78,000), in their basic form, from Fig. 17.13. To this must be added approximately $1500 per Substation to cover electrical operation of the main 50,000-amp circuit breaker and the relays (not covered in this chapter) necessary to enable it to provide the functions of a network protector (total of $78,000 $9000 = $87,000). I n addition, since all transformers are in parallel, the short-circuit current is over 25,000 amp, requiring 50,000-amp interrupting-capacity feeder and circuit breakers. Therefore, all feeder and tie breakers will have to be increased in cost by the difference between the 15,000-amp electrically operated breaker and the 50,000-amp manually operated breaker ($1600 - $800 = $800; $800 X 36 = $28,800). The total cost with above additions is $115,800. Furthermore, there are a total of eight breakers per substation, four feeder breakers and four tie breakers. The two additional breakers per substation (12 X $1600 = $19,200) in the larger interrupting ratings increase the total cost of the substation up to $135,000. Addin? 25 per cent for installation makes their approvimate installed cost $165,750, The tie cable for this system to tie all units together requires 3200 ft of No. 4/0-Awg cable. This allows a 25-ft vertical run for each cable plus the horizontal runs, as indicated in the diagram. This has an installed cost of $3.40 per f t (from Table 17.13), making a total cost of the tie cables of $10,880. The total installed cost of the secondary network system is $262,155 or approximately $87.40 per kva. Primary switchgear, primary cable, secondary cable, and busway are the same as for radial system.

+

+

LOAD- AND COSFESTIMATING DATA

904

4160 VOLTS 25OMVA

6

$Y

Y

s:

500 KVA (EACH)

"7;:1200

350 MCM INTERLOCKED ARMDR 5 K V CABLE

s:

AMP N W P

EAGH TI& CIRCUIT AK-1-50 TIE FEEDER BREAKERÇ PER SUBÇTATION

1-410

3-CONO

I

(0)

I

I1

500FT

I----

_-___-

I - - - - L - A 400 AMP PLUG-IN EUS 600 F T

tb) FIG. 17.7 Secondary neiwork load-center ryrtern for cost cornpariron.


905

Conclusion. Summing up these system costs, the radial system costs $63.30 per kva, the secondary selective system $70.00, and the secondary network system $87.40 per kva. I n this particular case the network system costs 38 per cent more than the radial system, whereas the secondary selective system costs only 11 per cent more than the radial system. It may seem that an unusually large amount of transformer capacity was placed in the secondary selective and radial systems compared with that in the network systems, but this was done to provide for diversity of loading. This would favor the network in the comparison, although the example shows t,hat even under the assumed conditions the network system is considerably more costly. REFERENCES 1. “Electric Power Distribution for Industrial Plants,” AIEE Committee on Industrial Power Applications, American Institute of Electrical Engineers, New York, 1955. 2. Umanskv. .,L. A,. Trends in Electrification of American Steel Industrv. “ , AIEE Technical Paper 48141. 3. “How to Estimate Electrical U’ork,” Eleclrical Conslruction and Maintenance, 1950.

906

LOAD. AND cOST.ESTIMATING DATA

COST-ESTIMATING DATA FOR USE IN COMPARATIVE SYSTEM STUDIES The following tables and curves of cost-estimating data were compiled from prices in effect a t the time of writing; however, it must be nnderstood that all prices are subject to change. For this reason the information found hereiu is not t o be used for competitive purposes, but simply for comparing costs a t different types of power system arrangements, These costs vary considerably, possibly 20 or 40 or 100 per cent (as illustrated by the preceding examples), thus a small percentage change in the net selling price of apparatus will not materially affect the power system comparisons. INDEX OF FIGURES AND TABLES Three-phase power transformers, 5000 to 50,000 kva, 13.8 t o 132 k v . . . Three-nhase ~ o w e rtransformers. 1WO to 5000 kva,. 13.800-2400Y/1386 . volts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Turbine-eenerators. 2500 to 20,000 kw. 4.16 or 13.8 k v . . . . . . . . . . . . . . . Mmter unit substations, 2500 to 25,000 kva, 13.8 kv secondary.. . . . . . . Master unit substations, 1500 to 10,000 kva, 4.16 kv secondary.. . . . . . . Load-center unit substations, 225 to 2000 kva, 480Y/277 or 208Y/l20 volts secondary.. . . . . . . . . . ......... ............. Primary switching for radial substation including steel structure and componPnts . . . . . . . . . . . . . . . . . . . . . . . Medium-voltage metalklad switchgear, ........ Low-voltage metal-enclosed switchgear, 600 volts maximum.. . . . . . . . . . Indoor metaknelosed buswsy, 4.16 or 13.8 k v . . . . . . . . . . . . . . . . . . . . . . . First cost and installed cost of 15-kv cable.. . . . . . . . . . . . . . ...... First coat and installed cost of 5-kv cable.. .................... First cost and installed cost of 600-volt cable.. . . . . . . . . . . . . . . . . . . . . . . Squirrel-cage and wound-rotor induction motors, 500 t o 5000 h p . . . . . . . Synchronous motors. 500 to 5000 h p . . . . . . . . . . . Squirrel-cage induction motors and control, 20 t p .............. Wound-rotor induction motors-and control, 20 to 500 h p . . . . . . . . . . . . . . Synchronous motors and control, 20 to 500 h p . . . . . . .........

Fig. 17.8 Fig. 17.9 Fie. 17.10 Fig. 17.11 Fig. 17.12 Fig, 17.13

Table 17.8 Table 17.9 Table 17.10 Table 17.11 Table 17.12 Table 17.13 Fig. 17.14 Fig. 17.16 Fig. 17.17 Fig. 17.18

0

2

12,500

I

66,667

W I T H AND WITHOUT LOAD RATIO

TRANSFORMER KVA (FORCED - A I R COOLED RATING) 26,667 40,000 50,000

I

D -132,000- 13,800Y17960

)

IS00

I

STANDARD KVA RATING 2500

3750

FIG. 17.9 Three-phase power transformers, 1000 lo 5000 h a , 13,800-24WY/1386 volts.

I1

K

51

FIG. 17.10 Turbine-generators, 2500 to 20,ooO kw, 4.16 or 13.8 kv.

MACHINE RATING K W

DATA: T U R B I N E GENERATORS, 3 P H - 6 0 CY-0.8 PF 600 PSI - 7 M F - 2 ’ HG

APPROXIMATE DATA FOR USE IN COMPARATIVE SYSTEM STUDIES

I

~.

.

'

FIG. 17.1 1

10 2 O . m

20.+00

1

.13.8- 500

Master unit substations, 2500 to 25,000 h a , 13.8-kv secondary.

UNIT SU~STATION R lQo00 12,000 I 15,000 lQo00 12000 14.000 16 UNIT SUBSTATION KVA

STANOARO

!700 3750 5000 75 10 4000 6000 8000

'

I- AM- 13.8- 150; $8100 2-AM- 13.8 - 250: 8300 8500 3-AM- 13.8- 500;

22.000

25,000 24000 26

a

a

P

0

a

X 0

z

$

Y

0

2 0

2

a

C

5

IC

ln I 5 a

Y P

a

x

>

2C

2:

FIG. 17.12

I

I

AM-2.414.16- 150/250

10000

*

1 5000 6000 7000 8000 9000 10000 UNIT SUBSTATION K V A Mar& unit substations, 1500 to 10,oOO kva. 4.16-ln secondary.

I

NIT SUBSTATION RATINGS 5000 7500

I

TRANS PRI VOLTAGE-l2BKV TO 6 7 K V

a n. L a

5 rg

W

a

0

5

a

K

m

L

Y

2

C

IC

20

30

4c

5c

6C

FIG. 17.13

Load-center unit substations, 225 to 2000 h a . 480Y/277 or 208Y/120volts secondary.

.<

Without Primary Breaker

Primary Switching for Radial Substations Including Steel Structure and Components

Amp

I

Mva

6,000 285 4,000 285

2OOE 200E

46 69

-

8.000 285

2OOE

34.5

300 448

3400

$3200

13.8 23 34.5 46 69

I :%I

1

$12.800 13,100 13,700

$14,600 15,600 16.100 19,100 21.200

- w

120.700 i 7,000 17,500 19,100 23,200

I500 mva

interrupting rating

20,000 30.000

High-volrage ail circuit breaker

2OOE 400E

System

1

5000 6800

4100

3400

$3100

Mva

APPW data

YOltoge

10,000 240

200E

23

10,000 150

2OOE

Amp

rating

Interrupting

High-capacity fuses

-_____

~ p p r o , sy,. data rent

~-

rating

lnterrvpting

- --__

current

Max

Low-capacity (use*

-

13.8

kv

dou,

System rollage

IApproximde Dato for Use in Comparative Syslom Studies)

TABLE 17.7

0

-

*5

P

rating, kw 110.8 pf and

and exciter

rating,

amp

14.800 15.500 18.800 15,700 I9.000 15.000 15,200 18.500 15.500 18,800

I200 I200 2000 I200 2000

1200 1200 2000 1200 2000

7.2 7.2 7.2 7.2 7.2

13.8 13.8 13.8 13.8 13.8

-

14,400 14,600

2000

11,200 11,400

2.4 4.16

4.16

1200

14.200 14.350

11.000 11,200

$10,400 10,600

__

10,000 15,000 15,000 20,000 30,000

15,000 10,000 15,000

l0,000

10,000

10.000

6,000

3,500 6,000

6,000 10.000

3.500 6,000

3,500

-~ 3,500

60 wycle,

-

6,900 7.100 91900 7,300 0,200

6,900 7,600 0,500 7,900 0,700

7,600 7,600

4,700

4,700

7,400 7,400

4,400 4,400

b3.700 3,700

-

line

cornins

or

8.600

16,900

-

10,300

8,900

7,900 8,700

5,500 5.600

5,300 5,300

$4,500 4,600

8.200 8,400

17,800

15.800 17.300

11.400

11,300

10,900 10,900

$9,700 T 9?00

start less lhon 1500 hp

15,200

14.300 14,700

16,000

14,000 15,600

9,600

9,600

9.100 9,200

8,000

$7,9001

less lhon 1500 hp

start

unloaded

reocto,

motort eutr.1

motort

full-voltage unloodod

Induction

Induction

16.000 16,500

~

less than I500 hp

start

unloaded

jpchronoi motort lUt,Dl reoc

-

9,900 10.100

10,600

9,600 10.400

7,300

7,200

7.000 7,000

16,200 6.300

,tart less than 1500 hp

Spchronow motort f"ll."oltage unloaded

m Studied

ear. 2.4 to 13.8 Kv

* A switchgear equipment will require a source of contr mwer, met :lad breaker testing accessories, and potentia which total aooroximatelv 52500. t 1500 hp'and above-add approximately $2600 for motor differential protection. $Line resotor instead of neutral reactor.

50/250..

.........{ AM-7.2-1 50. ................... AM-7.2-250.. .................. AM.7.2-250 .................... AM-7.2-500. ................... AM.7.2-500 .................... AM.13.8-150 ................... AM-13.8-250.. ................. AM-13.8-250. .................. AM-13.8400 ................... AM-13.8-530.. .................

AM-2.4/4.1&150/250..

2.4

2000

2.4 4.16

.........{

1200

2.4 4.16

I200

.........{ .........{

~

2.4 4.16

6-1 OO/l50..

AM-2.4/4.16-I

gensrator

*"er.tO#

cYlre"t

:ceder

A-t

-

Suggested ?ax. turbine-

hooker

- -

ti"* S)

us0 in Comp

'e Dato

Metal-clad Swih

ipproxi

Medium-volta

.................1

AM-2.4/4.16-100/150..

AM.2.4/4.I

AM-4.16-50..

TABLE 17.8

.

'

p

t

2

2 T

2

0,

n

p

$

0

15.000

25.000

50,000

Manual Elechisol

Manual Electrirol

Manuol Electrical

Electrical

Electrical

AK-1-15

AK-1-25

AK-1-50

AK- 1-72

AK-I-i OC

Continuow

-

100.000

4000

2000 3000

--

75,000

200-1600

50600

25-225

--

nterrvpting

Method of operotion

Type of breaker

Breaker ratings, omp

I200

900

$600 800

'eodcr

1

2600 2800

6400

11.100

8300 9500

5100

5800

5300

4300 4500

$3300 3500

A-C

4200

2100

~$2100 $2300 1600

100 amp 1600amp

transformer secondary bur tie

Incoming line

I

1

I

I

4600 5100

3800 4000

$3400 3600

,t.rt

unloaded

motor hrll-roltoge

I 7500

.....

2200

1700

1200

1000

..... 5600

$600 800

,l.Tt

'ull-rolt.age

motor

Induction

i;&

start

line reactor unlooded

motor

Synchronow Synchronou:

,rimate Datn for Use in tomparolive System Studies)

TABLE 17.9 low-voltage Metal-enclosed Switchgear, 600 Volts Maximum

917


The vertical “L”, e, can be applied as a “T”section by removing the end cover of the “L” and adding the necessary additional conductor length required. The sizes and arrangements available allow for a simple and flexible installation, which should satisfy most requirements.

TABLE 17.11

Economics of Installed 1500-volt Cable (Dollon per Foot1

Cable

Type VCL varnished-cambric leaded shielded [three conductor)

size,

A w or MCM

Type RH corono1* g*oprene shielded [three single conductors]

- - -

-g

i

E

=

-u

-

+ E

S

O

$ 0

-

0

5P

.-

88 I12 I29 I48 171 I97 227 251 304 368 454

I5.t 4.f 5.c 4.7 4.: 4.: 3.f 3.t 3.t 3.7 4.c

0,

-”

P

LI

0 0

LI

-

$

f E

-

~

> :--? -8 n -0a , -2 a=.8 2 5 L E ” -2 2- a; -“ * * -_ u -

”

- 0

LE

- -

i: s - - -

L

-a

-: -

Type ACV interlocked ormm shielded lthre. conductor1

-

D 0

E

8 0

0 0

8 6

4 2 1 0 00

000 0000 250 350 500 750 1000

-

60 79 101 115 129 147 167 190 210 252 300

3.2 3.6 4.1 4.3 4.6 5.0 5.5 6.1 6.5 7.6 8.8

$5.9 $9.1 6.3 8.C 7.6 7.: 7.8 6.1 8.2 6.3 8.6 5.1 9.1 5.4 10.3 5.4 10.8 5.1 11.9 4.7 15.4 5.1

... ...........

-

I 1

-

* Registered trade-mark of General lect

2.4 2.7 3.0 3.4 3.6 4.0 4.4 4.8 5.8 7.1 9.3

15.

5. 6.

6. 7. 8. 8. 9. 11. 13. 18.

97 126 149 I65 192 220 251 277 337 414

L2.8 14.1 3.1 4.i 3.3 4.7 3.5 5.c 3.8 5 . : 4.1 5.t 4.6 6.c 4.9 6.1 6.0 7.1 7.6 9.:

$4.2 3.5 3.1

3.0 2.7 2.5 2.4 2.3

2.2 2.2

- - - -Comoanv igle-eonductor cables in steel conduit

Rated ammms: Three-conductor or Tree or interlocked armor cable a t 40 C am ent. Cable c o d : Includes 10 per cent to .ow for shrinkage, waste, and for making u p terminations. labor a t $2.50 per hr with 38 per cent .Znstalled . . . . .cost: . Includes materials . . and . direct ”. aaairion tor insurance, ovemeaa, ana prom. Znstalled cost, dollars per ampere: These estimated data show the economic aavantage of interlocked-armor cable.

P18


TABLE 17.12

Economics of Installed 5000-volt Cable I D o I I ~ r sper Foot1

-

Cable sire.

-

A w

._ ._ c

~~

8 6 4 2 1 0 00 000 0000 250 350 500 750 1000

~

3

a

- 0

P

L?:

-

D

0 -

n o

;: -

0

2 -

~

. . . . . . ... .... .... 2 2 2 2>;

2>$

23s 2); 2)s

58 76 102 I18 134

I54

316

175 199 218 264

3);

361

3 3

t2.5 1 .? 2.8 I .7 3.7 i .a 3.9 2.1 4.2 2.4 4.4 2.7 4.8 3.1 5.9 3.: 6.3 7.6 4.c 9.3 5.t

,I.1

14.3 3.6 3.6 3.3 3.1 2.9 2.8

2 2 236 256 255

I -1 3 3

3.0

3 3 3%

2.9 2.9

. . . . . . ... .... . . . . . . ... .... ....

--- *Ii

-g

n

v

Type ACV interlockad armor (three conductor) ~

: : -e -$ - < 8 , 8 -

or

MCM

~

Type RH corono1* geoprene shielded [three single conductors)

Type VCL vornlrhod-cambric leaded (three-conductor)

I

49 69 90 118 135 I54 175 198 226 249 305

520 1

General Elect [actor or threc

iO.8 0.9 1.1 1.3

$2.2 2.5 3.4

$4.5 3.3 2.8 2.9

2.5

2.9 3.8 4.7

5.2 5.6 6.5 8.3

2.7 2.5 2.6 2.7

7.9 0.1

13.2 16.8

2.3

91 121 I39 161 185 213 244 270 334 417 529

-

$1.4 1.6 1.8 2.0 2.2 2.5 2.9 3.2 3.9 4.8 7.2

$2.6 $2.8 2.8 2.3 3.0 2.2 3 . 3 2.0 3.5 1.9 3.8 1.8 4.2 1.7 4.6 1.7 5.3 1.6 6.3 1.5 8.8 1.7 I

I

Commnv. ngle-conductor cables in stccl conduit

M',! ampere or ipterlocked-armor cable at 40 C ambient Cabls~~osl: Includes 10 per e m t to allow for shrinkage, waste, and far making up terminations. Inslalled eosl: Includes materials and direct labor at $2.50 per hr with 38 per cent addition far insurance, overhead, and profit. Inslalled cost, dollars p e r ampere: These estimated data show thc economic advantage of interlocked-armor cable.

Type RH Type VCL Conduit Venatol* gwprene varnished-combric ,ize (lhree single leaded lthrse conductor1 Coble Conductor1

Type TW

Type ACV interlocked armor lthreo conductor1

Flornenol* (lhree single

sire, .__ n

*w

n

E

or MCM

i

-:" O4

0

O

i f

-s

0, 0

: - L a z 3 2 ::

+

+ E

=

O , g L . . O : ? :

n

n

-g x : E E

- E

s o

Bu p= p= z rns

= g

- u-B .z, .g. : -?i b ;

+ E

0, u

2 2 = O 2 2 b L X 2 e t , a 5 5 - 2 :: ~ < z p_ " , " :_ : <2

c

9

9 0

O 8 = O a P 2 2 2 i 2 a p - : < ~Q S- :: ~+ +..a_

-

8 l/a l!i 4510.310.811.9 5010.511.412.8 4010.250.711.7 6 1 134 65 0.4 1.0 1.6 70 0.7 1.7 2.5 55 0.3 0.9 1.7 85 0.5 1.4 1.6 90 0.9 1.9 2.2 70 0.4 1.3 1.8101 1 0 . 8 1 2 . 0 1 l . 9 4 l?/al>1 2 I ) i l > $ l l 5 0.7 1.6 1.4120 1 . 2 2.2 2.5 95 0.6 1.4 1.5134 1 . 1 2.2 1.6 I l > 6 2 130 0.9 1.9 1.5140 1.4 2.8 2.0110 0.7 1.8 1.6154 1.2 2.3 1.5 0 2 2 I 5 0 1.0 2.4 1.6155 1.6 3.0 1.9125 0.9 2.3 1.8179 1.4 2.6 1.4 2 175 I . ? 2.6 1.5185 1.9 3 . 3 1.8145 1.0 2.4 1.7205 1.6 2.8 1.3 00 2 235 200 1.4 2.9 1.4210 2.2 4.3 2.0165 1.2 2.6 1.6237 1.8 3.0 1.3 000 2 1.8 3.8 1.7235 2.6 1.7 2.0195 1.5 3.6 1.8271 2.1 3.4 1.3 0000 2)62)$230 255 2.3 4.4 1.7270 2.9 5.7 2.1215 1.8 3.9 1.8300 2.3 3.7 1.2 250 2'13 3 310 3.0 5.8 1.9325 3.6 6.4 2.0260 2.4 5.1 2.0371 3.0 4.4 1.2 350 3 3'6380 4.0 6.8 1.8405 4.7 8.3 2.1320 3.3 6.1 1.9462 3.9 5.3 1.1 500 3 475 5.9 9.5 2.0500 6.511.0 2.2 587 5.7 7.3 1.2 750 3 3 5 4 2.2535 7.914.7 2.7 1000 4 5 545 7.411.9

_...... _._. ....

14

U

2 0

n 6

0

K

0

2 z x

w 10

n

0

9J

K

u)

P

W

K

3 18 g S

n

0

g

K

22

28

500

1000

Synchronous motors, 500 to 5oM) hp.

2000 2500 3000 MOTOR HORSEPOWER

FIG. 17.15

1500

3500

4000

4500

5oGQ

COMPARATIVE SYSTEM STUDIES DATA : MOTORS, 500 TO 5000 HP SYNCHRONOUS MOTOR l.OPF.IZ00 RPM


922

tt 0 0

M3MOd3SMOH

I 0 n

I

::

M3d S M V l l O O 31VRIXOMddW

FIG. 17.17

Wound-rotor induction motors and control, 20 to 500 hp.

u

9 0

-~

FIG. 17.18 Synchronous motors and conlrol, x ) to 500 hp

Appendix

Compiled by D. B. Armstrong

CONVERSION FACTORS Acres Acres Acre-feet Amperes per square centimeter Amperes per square inch Ampere-turns Ampere-turns per centimeter Ampere-turns per inch Atmospheres Atmospheres Atmospheres Atmospheres

BY 43,560 1.562 X 10-5 43,560 6.452 0.1550 1.257 2.540 0.3937 76.0 29.92 33.90 14.70

To Oblain Square feet Square miles Cuhic feet Amperes per square inch Amperes per square centimetei Gilberts Ampere-turns per inch Ampere-turns per centimeter Centimeters of mercury Inches of mercury Feet of water Pounds per square inch

British British British Britiah

0.2520 778.2 3.931 X lo-' 2.930 X lo-'

Kilogram-calories Foot-pounds Horsepower-hours Kilowatthours

Ccntimcters Circular mils Circular mils Circular mils

0.3937 5.067 X 10-6 7.853 X 10-' 0.7854

Inches Square centimeters Square inches Square mils

Degrees (angle) Dynes

0.01745 2.248 X

Radians Pounds

Ergs

7.378

Farads Foot-pounds Foot-pounds Foot-pounds Foot-pounds

108 1.285 x 10-3 5.050 X lo-' 0.1383 3.766 X lo-'

Gallons per minute Gallons of water a t 62 F Gallons of water per minute

2.228 8.345 500.7

Mulliply

thermal thermal thermal thermal

units units units units

x

x

925

10-8

10-8

Foot-pounds Microfarads British thermal units Horsepower-hours Kilogram-meters Kilowatthours Cubic feet per second Pounds Pounds per hour

APPENDIX

926

Gausses Gilherts

6.452 0.7958

To Obtain Lines per square inch Ampere-turns

Henrys Horsepower Horsepower Horsepower Horsepower Horsepower Horsepower (boiler) Horsepower-hours Horsepower-hours

108 42.44 33,000 1.014 10.68 0.7457 33,475 2544 1.98 x 106

Millihenrys Btu per minute Foot-pounds per minute Horsepower (metric) Kilogram-calories per minute Kilowatts Btu per hour British thermal units Foot-pounds

Inches Inches of mercury Inches of mercury Inches of water Inches of water Inches of water

2.540 1.133 0.4912 25.40 0.5781 0.0361

Centimeters Feet of water Pounds per square inch Kilograms per square meter Ounces per square inch Pounds per square inch

Joules Joules

9.480 x 10-4 0.7377

British thermal units Foot-pounds

Kilograms Kilogram-calories Kilolines Kilometers Kilowatts Kilowatts Kilowatts Kilowatthours Kilowatthours

2.205 3.968 108 3,281 0.6214 56.88 737.8 1.341 3,413 2.656 X lo6

Pounds British thermal units Maxwells Feet Miles Btu per minute Foot-pounds per second Horsepower British thermal units Foot-pounds

1og.N or In N Log,,N Lumens per square foot

0.4343 2.303 1

l0gi.N 1og.N or In N Foot-candles

Maxwells Megalines MPgohms Meters Meter-kilograms Microfarads Microhms Microhms per centimeter cube Microhms per centimeter cube Miles Miles

10-2 10' 108 39.37 7.233 10-6 10-6 0.3937 6.015 5,280 1.609

Kilolines Maxwells Ohms Inches Pound-feet Farads Ohms Microhms per inch cube Ohms per mil foot Feet Kilometers

Multiply

yonreters

BY

APPENDIX

Multiply

P17

To Obtain

BY

lo-'

Ohms Ohms Ohms per mil foot Ohms per mil loot

106 0. I662 0.06524

Megohms Microhms Microhms per centimeter cube Microhms per inch cube

Radians

57.30

Degrees

Square centimetors Square feet Square fcet Square inches Square inches Square kilometen Square meters Square miles Square miles Square millimeters Square mils

1.973 X 10' 2.296 X 1 0 F 0.09290 1.273 X lo6 6.452 0.3861 10.76 640 2.590 1.973 x 108 1.273

Circular mils AWW Square meters Circular mils Square centimeters Square milcs Squarz fcet Acres Square kilometers Circular mils Circular mils

1 1.8 1 5/9 2240 2205 2000

Ahsolute temperature C Temperature F Ahsolute temperature F Temperature C Pounds Pounds Pounds

+ 273%) + 17.8) + 460)

Temperature (C Tcmperature (C Temperature (F Temperature (F Tons (long) Tons (metric) Tons (short)

32)

EQUIVALENT VALUES OF ELECTRICAL AND HEAT UNITS Eyuioalrnt Value in Othei l i i t s 1.341 horsepower-hour 1 kilowatthour 3413 British thermal units 0.7457 kilowatt-hour 1 horsepower-hour 2541.1 British thermal units 11.341 horsepower 44,254 footIponnds per minute 1 kilowatt 737.56 foot-pounds per second 3413 British thermal units D W hour 0.7457 kilowatt 33,000 foot-pounds per minut? 1 horsepower 550 loot-pounds per second 2544.1 British thermal units Der houi 1 j0"k = 1 watt-second 778.26 foot-pounds 1 British thermal unit' = . 1054.8 j o u l ~ sor wattbeconds of t h e quantity of heat rpquired * T h e British thermal unit (Btu) is defined as to raise t h e temperature of 1 Ib of water from 32 F t o 212 F at 1 atmosphere (14.696 psi).

limt

={

=I {

APPENDIX

928

Unit Degrees absolute (Kelvin) Dcgrees absolute (Rankine) Degrees centigrade Degrees Fahrenheit

Equivalent Value i n Other Units 273 460

= Degrees centigrade = Degrees Fahrenheit = 56 (deg F - 32) = 1.8 (deg C ) 32

+ +

+

NOMENCLATURE The following tables, compiled from ASA Standsrds, represent the letter symbols and abbreviations most commonly used in power system engineering. For a complete listing of letter symbols and abbreviations, refer to ASA Standards 210.5-1949, 210.4-1943, and 210.1-1941.

The Greek Alphabet Quantities Cmnmonly Designated by the Small Gmek Letter Name Resistance- temperature coefficient alpha beta Phase constant Conductivity gamma Increment * delta Dielectric constant epsilon zeta Efficiency eta Angular phase displacement theta iota Magnetic susceptibility kappa. Wavelength lambda Permeability mu Reluctivity nu xi omicron 3.1416 pi Resistivity rho Summation. Sigma Time constant tau upsilon Magnetic flux’ phi Angular phase displacement phi chi psi Flux of displacement* omega Angular velocity commonly designated by the large Greek letter.

Large Letter A

Smell Letter

B

B

r A

E 2

H

n Y

6

r 7l

e

e

I K

x

A

x

M N

I”

I

.j

€

0

0

n

r

P

P

z

m

T T

7

” 9 9

X Y

n

x. .

+ w

* These items are

letter Symbols Used in Electrical Formulas Quantily Admittance Angular phase displacement

Symbol (8,

Y 9) (theta, phi)

Zllvstratiue Unit

Mho Radian

APPENDIX

Quantity Capacitance Charge, or quantity of electricity Conductance Current Dielectric constant Efficiency Electromotive force Energy Flux density Frequency Frequency, angular velocity Frequency, rotational Impedance Inductance Magnetic flux Magnetizing force Magnetomotive force Number of conductors or turns Number of poles Permeability Permeanee Power ReactancQ,capacitive Reactance, inductive Reactance, subtransient Reactance, transient Reluctance Resistance Resistsnee-temperature coefficient Resistivity Rotative operator, 90’ Rotative operator, 120’ Temperature

929

Illusttotiue Unit

Symbol

c

Q G I

Farad Coulomb Mho Ampere Farad per inch Volt Joule Lines per square inch Cycle per second Radian per second Revolution per second Ohm Henry Line Ampere-turn per inch Ampere-turn

Line per ampere-turn per inch Line per ampere-turn Watt Ohm Ohm

Ohms

(a)(alpha)

(P) (rho)

Ohms Ampere-turn per line Ohm Decimal parts per degree C Ohm-in

i a

Degree C Degree K (Kelvin) Dearee Time Second Wavelength Meter Work W Joule NOTE: When distinctions between maximum, instantaneous, and r m s values are necessary, Emand I, are recommended for maximum values, e and i for instantaneous values, and E and I for rms values. * In the “Industrial Power Systems Handbook,” X without the subscript L is used to denote inductive reactance. t

T

Letter Symbols Used in H e a t and Thermodynamic Formulas Quonlity Efficiency Enthalpy Entropy Volume flow

Symbol

(eta) Hi h 8, 8 , 9

(7)

Q

lllvslralive Unit Per cent Btu per pound Btu per degree F per pound Cubic feet per second

APPEN~JIX

930

Quantity Heat flow Theoretical steam rate Actual steam rate Temperature, Fahrmheit Superheat Terminal difference Velocity Specific volume Pressure, gauge Pressure, ahsolute Gas constant Specific heat Energy

Illustrative Unit Btu per hour Pounds per kilowatthour Pounds per kilowatthour Degree F Degrcc F Degree F Feet per second Cubic Feet per pound Pounds per square inch gauge Pounds per square inch absolute

Symbol

Q TSR

SR 1 , FTT

FS td

V

"

psig psia R c

U

Btu per pound

Abbreviations Used in Electrical Formulas pf = power factor hp = horsepower r = 3.1416 va = volt-ampere kva = kilovolt-ampere mva = megavolt-ampere

kv var kvar w kw mw

line-to-line voltage in kilovolts reactive volt-ampere = kilovars, reactive kilovolt-ampere = watt = kilowatt = megawatt

=

=

DEVICE NUMBERS AND FUNCTIONS FOR SWITCHGEAR The devices used in switching equipment are referred to by standarddevicenumbem. These numbers are used on diagrams, in instruction books, and in specifications to permit ready understanding of the function and operation of switching equipment. The following list of device numbers and functions &retaken from American Standard C37.2-1945. ~~

,, '

Detrice No. Function and Definition 1 Master element .. 2 Time-delay starting or closing relay 3 (Reserved for future application) 4 Master cantactor or relay 5 Stopping device 6 Starting circuit breaker, contactor, or switch 7 Anode circuit breaker 8 Control power switch 9 Reversing device 10 Unit sequence switch 11 Control power transformer 12 Overspeed device Synehronous-speed device 13 14 Underspeed device 15 Speed-regulating device 16 Battery-charging control device Series-field shunting circuit breaker or contactor 17

APPENDIX

Deuiee No. 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50* 51 * 52

53 54

55 56 57 58 59 60 61 62 63' 64 65 66

Function and Definition

Accelcrating or decelerating eontactor, circuit breaker, or relay Starting to running transition eontactor or relay Electrically operated valve Impedance relay Equalizer circuit breaker or contactor Tcmpeetture-rPgulating device Bus-tic circuit breaker, contactor, or switch Synchronizing or paralleling device Apparatus thermal device A-C undervoltage relay Resistor thermal device Isolating circuit breaker, contactor, or switch Annunciator relay Separate excitation device D-C reverse-power relay Position switch Motor-operated sequence switch Brush-operating or slip-ring short-circuiting device Polarity device Undercurrent or underpower relay Eearing thermal device Field reducing contactor Field relay Field circuit breaker, contactor, or switch Running circuit breaker, contactor, or snitch Transfer device Unit sequence starting contactor or relay D-C overvoltage rday Reverse-phase, phase-balance current, or power-rectifier misfire- relay Single- or reverse-phase voltage relay Incomplete-sequence relay A-C thermal relay A-C instantaneous overcurrent relay A-C time-delay avercurrent relay A-C circuit breaker or contactor Exciter or generator relay High-speed circuit breaker Power-factor relay Field application relay or device (Reserved for future application) (Reserved for future application) A-C overvoltage relay Voltage balance relay Current balance relay Time-delay stopping or opening relay Fluid pressure, level, or flow relay Ground protection relay Governor Notching relay

Shows change from, or addition to, ASA-C37.21945.

931

APPENDIX

932

Function and Definition Device N o A-C power directional or a-c directional overcurrent relay 67. D-C thermal relay 68 Permissive control device 69 Electrically operated rheostat 70 D-C line emergency circuit breaker or contactor 71 D-C line circuit hresker or contactor 72 Load resistor circuit breaker or contactor 73 Alarm relay 74 Positian-changing mechanism 75 D-C overcurrent relay 76 Impulse transmitter 77 Phase-angle measuring relay 78 A-C rcclosing relay 79 D-C undervoltage relay 80 Frequency device 81 D-C reclosing relay 82 Selective control, or transfer, contactor or relay 83 Operating mechanism 84* Carrier or pilot-wire receiver relay 85 Locking-out relay 86 DiEerential current relay 87 Auxiliary motor or motor generator 88 Line switch 89 Rpgulating device 90 D-C voltage directional relay 91 D-C voltage and current directional relay 92 Field-changing contactor or relay 93 Tripping or trip-free relay or contactor 94

(Reserved for special application)

98 99

Shows change from, or addition to, ASA-Cj7.2-1945.

GRAPHICAL SYMBOLS FOR POWER-SYSTEM ONE-LINE DIAGRAMS One-line diagrams are very useful in showing, by means of graphical symbols and conventional nomenclature, an over-all power system arrangement. These symbols, when used consistently and in conformance with general practice, provide a valuable

tool. The following table, compiled from ASA Standards, represents the graphical symbols most commonly used in power-system one-line diagrams. For a complete list of graphical symbols for one-line diagrams, refer to ASA Standards Y32.1.1-1951and 232.3-1946.

933

APPENDIX

Equipment

Variations Definition

Arresters lightning

Arrester plus ground (mrge arrester)

Valve-type arrester Cable terminations

Single-conductor tern naticin Three+onductor termination

Capacitor

Capacitor plus ground (surge capacitor)

Capacitor hushing

Capacitor hushing patential device

Coupling capacitor potential device

Circuit breakers air

Breaker with drawout feature

Breaker with drawout feature, and operating coil

power

Breaker with drawout fcature

Breaker with diseonnecting switches

APPENDIX

934

Variations Equipment

Basic symbol Symbol

Contact

I T

Definition

L T

Vormally open (NO)

\Tormally closed (NC)

Coil operating

k. T

:ontact with blowout coil

Drswout mounting

Fuse

High-voltage primary fuse cutout, dry type; or fuse disconnecting switch lamc as above

High-voltage primary fuse cutout, oil type

Gap protective

-DQ+

:ap plus ground (surge gap)

Synchronous generato,

Generator

3epamtcly excited d-c generator' I

D-C shunt two-wire generator' * T h e broken a part of the sy

- - - indicates where line connection

symbol is made snd is not

935

APPENDIX

Equipmeni

Basic symbol

*

Variations

Symbol

Definition

I

Generator (Cr

D-C compound two-wire generator*

I

D-C three-wire shunt generator'

I -

Ground

Mechanical connection

____ dashes c necting equipme

ihort

Meters and instruments

0 (etteror letters shall he placed within the circle t o indicate the type of instrument:

A D F GD MA

Ammeter Demand meter Frequency meter Ground detector Milliammeter PF Power-factor meter R D Recording demand meter R E D Recording

* T h e broke a part of the s

R H Varhour meter S Synehroseope T Temperature V Voltmeter VA Voleammeter VAR Varmeter \I' Wattmeter WH Watthour meter

e - - - indicates where line connection to symbol is made and is not 01.

APPENDIX

936

Equipment

triations

Basic symbol

Definition

Symbol

Induction

Motor

Synchronous

Reactor (nonmagnetic core

+ ac

pawer

Relay

Q 0 'be relay device function number ihould be placed within the circle

APPENDIX

Equipment

P37

Variations

Basic symbol Symbol

Definition

Relay protective functions The following symbols arm used to indicate protective func r be placed adja tions and cent to th, .sic relay symbol

Over Under Directional (directional over)

Balance

Differential

-

Pilot wire

cc

Carrier current

Actuating quantity: The actuating quantity is indicated by the following letters, placed either on or above the relay protective function symbol shown above: C Current' Z Distance V Voltage W Power F Frequency

+

Phase T Temperature GP Gas pressure S Synchronism

*Generally accepted practice ia to omit any designation for currentactuated devices.

APPENDIX

938 ~

riations Equipment

Relsy protectivi functions (Cont'd)

Basic symbol

Definition

Ivereurrent

bvervoltage Xreetional overcurrent )irectional residual overcurrent

Jndervoltage 'ower directional

3slanee current

Differential current

Distance Directional distance

Over frequency

Under frequency

Ovcr temperature

Phase balance

Phase rotation

APPENDIX

Equipment

Basic symbol

P39

Variations Definition

Relay proteeti, functions (Cont'd)

Pilot wire, differential curent Pilot wire, directional eomparison

Carrier pilot Positive phase sequence undervoltage Negative phase sequence overcurrent Gas-pressure relay

Resistor

Switch air break

Thermal elemen

-

o u t of step

Grounded

2Double throw

-

Switch with horn gap

APPEHDLX

940


Transformer

Basic symbol

Definition

Two-winding transformer with taps Adjustable mutual inductor. constant-current transformer Threewinding transformer

Autotransformer

Potential transformer

Current transformer

Bushing-type current transformer

step-voltage regulator or load-ratio-control autotransformer

Load-ratio-contml transformer with taps

941

APPNDIX


Basic symbol Symbol

Trsnsformer (Cont’d)

@

Definition

Single-phase induction voltage regulator

Three-phase induction voltage regulator

Transformer winding connections rhe following symbols are used to indicate transformer winding connections and may symbol:

A

Three-phw three-wire delta

P -

Three-phase three-wire delta grounded

f -

Three-phase four-wire delta grounded

A

Three-phase Y

$

Three-phase Y grounded neutral

APPENDIX

942

riatibns Equipment

Basic symbol Definition

Transformer winding connections (Cont’d)

rhree-phase zigzag

rhree-phase zigzag grounded six-phase star (or diametrical)

Six-phase Star with grounded neutral

Three-phase open delta Three-phase open delta grounded at common point

1.06

1.12

% .440

241

130

210 252 300 371

l.w

1.05

1

1.07 1.00

185

180

203

172 194 213 253 301 367

161 181 199 236

0.85 0.80

0.90

0.95

1.04 1.w

, t Temperature

281 a42

r

0.88 0.83

1.04

.... ....

378

__

0.87

1.04

213 251 297 355

159

221 261 313

136 154 174

141

125 141

151

107 121

86

111

83 I34

79

83

125

101 115

114

66

~

51 67

...

Synthetic rubber. temp 80 c. 8.W15,WO Volts ahieldedt

shielded3 49 64

ihialded

-

Synthetic rubber. tsmp 85 c. 0-8000 volt.

108 119

77 99 111

60

68

87

46 60

46

rubber. tamp 75 c. )-Bw "Oilof

Synthetic

52

Correction Factor8 foi Varioua Am1

408

::&

.... 0.93 0.84 0.92 ,... 122 0.71 0.85 0.88 0.83 National Electrical Code ratiom (eorreoted for 40 C). t R a t i n 0 not $Overed by induatry stand*rds: d u e s liated are t h w e used by the General Electrio Company. Inavlated Power Cable Eoglnears ALYIODIB~~OO (IPCEA) ratings.

1.00

480

328 373

750 1OW

%

176

i ! 418

213

218

I)

rubger or "81nished-cambrio temp 77 C. 15.0W volts

I 1 I

118

46 58 76

shzgdt Shielded?

& ::$ :$ ;

Butyi synthetic rubber or varnished cambric.

224

114

101

40 57 75

Synthetic rubber, temp 75 c. 0-600volta*

262

160

135

103 119

33 45 57 78 90

Rubber 01 thermoplasti temp 60 C. 0-6wvolts'

500

01

I N CONDUIT Three single-oonductor cables or one three-conduotor cable per Conduit, 104 F (40 C) ambient air temperature

250 350

0 00 000 0000

1

6 4 2

8

Siie. Awa MCM

SIMPLIFIED SELECTION O F CONDUCTOR SIZE

Current-carrying Capsoitis in Amperes for Cable in Conduit. in Underarovnd Ducta, and in Intarlooked Armor

0.84

1.05

1.11

1.25

I85 213 244

161

138

52 68 81 121

f "

x

3P

%

v44

APPENDIX

NATURAL TRIGONOMETRIC TABLES

-

-

And<

Nam

meea

tune. tion

de-

__ n 1

ot

~

sin COS

tan

sin CDS

sin

3

11

12 13

14

15

0.0122 0.9999 0.9999 u.0105 0.0122 0.0279 0.9996 0.0279

n ,0297 n ,9996 0.0297

o.0140 0.0157 0.9999 0.9999 n.oi40 0.0157

n ,0332 4.9995 0.9995 0.0314 0.0332 0.0314

0.0663 0.9978 0.0664

O.OBSO

n.o9nn 0.06zg

tan

0.0802 0.0819 0.0817 0.9968 0.096~ 0.0965 0.0805 0.0822 0.0840

0.0854 0.9963 0.0857

sin

0.0076

0.0993 0.1011 0.~952 0.~951 0.9949 o.o981 0.0998 0.1016 0,1149 0.1167 0.1184 0.9934 n . 9932 0.9930

0 . 1028

0.1157

0.1175

0.1323 0.9912 0.1334

0.1340 0.1357 0.9910 n.g907 0.1352 0.1370

0.1374

0.1547

tan

0.1495 0.1513 0.1530 0 . ~ 8 8 8 0.0885 0.9882 0.1512 0.1530 0.1548

sin

0.1663

0.1685 0.9857 0.17m

0.1702 0.9854 0.1727

n.1719

COQ

COS

em tan COB

COQ

0.0645 0.9979 0.0647

0.1192

0.9977 0.0682

0.9947 0.1033

0.1201 0.9928 0.1210

o.om

0.1388

n.oss0

n. 1566

0.9851

(.US

0.9860

tan

0.1691

sin tan

0.1840 0.1857 0.9829 0.9826 0 . 1 8 7 ~ 0.1890

0.1874 0.9923 0.1908

0.1891 0.~820 0.1926

sin

n.zoi1

0.2028

0.9792

tan

n.2071

0.2045 n.9789

3.2062

0.9796 u 2u53

sin em tan

0.2181 0.9759 0.2235

li" tan li" em tan

C08

0.1745

0.2089

3.9785 1.2107

0.2198 n.97s5 0.2254

0.2215 0.9751 0.2272

1.2232 1.9748 1.2290

0.2351 0.9720 0.2419

0.2368 0.9715 0.2438

0.2385 0.9711 0.2456

1.2402 1.9707 1.2475

U.2521

0.2538 0.8673 0.2623

0.2554

0.9677 0.2606

0.2642

1.2571 3.9664 3.2661

sin

0.2688

COQ

0.9632

0.2706 11.9627

tan

0.2792

0.2811

0.2723 0.9622 0 ,2830

3.2740 0.9617 0.2849

800L".

rpyright

em

COS

R% ited 1936,

n.0105

0.0628

sin

10

0.9

C"3

tan

9

0.8

sin

xin

8

0.7 ~

0.0506 0.9987 0.0507

sin

7

~

0.6

0.0488 0.9988 0.0489

tan

6

degree

0.0454 0.0471 0.9990 0.9989 0.0454 0.0472

sin

5

8

~

tan

eos

tan

4

Value of function for esoh tenth of

~

tan

2

Natural Triaonornetric Functions. 0.0-15.9"

ho

U

0.8668

---11 E n ~ i ' :rs' Har

945

APPENDIX

-

Natural Trigonometric Functions, 16.0-35.9"

And

game

gree

'une-1

de-

Value of function far each I

Of

-

0.0

I

0.1

1

0.2

I

0.3

I

1

0.5

0.29w

0.301f

0.9568 0.3038

0.3024

0.3040

0.9532 0.9527 0.3172 0.3191

0.3051 0.9521 0.3211

0.3074 0.9516 0.3230

0.3190 0,9478 0.3365

0.3206 0.9472 0.3385

0.3222 0.946f 0.3404

0.3239 0.9461 0.3424

0.3256 0.3272 0.3289 0.3305 0.3322 0.3338 0.9455 0.9449 0.9444 0.9438 0.9432 0.9426 0.3443 0.3483 0.3482 0.3502 0.3522 0.3541

0.3355 0.9421 0.3561

0.3371 0.9415 0.3581

0.3387 0.940G

0.3404 0.9403

0.360C

0.3620

0.3420 0.3437 0.3453 0.3469 0.3486 0.9397 0.9391 0.9385 0.9379 0.9373 0.3640 0.3659 0.3679 0.3699 0.3719

0.3518

0.3535

0.3759

0.9354 0.3779

0.3551 0.9341 0.37%

0.3567 0.9342 0.3819

0.3584 0.3600 0.3616 0.3633 0.3649 0.3665 0.9336 0.9330 0.9823 0.9317 0.9311 0.9304 0.3839 0.3859 0.3879 0.3899 0.3919 0.3939

0.3681 0.9298 0.3959

0.3697 0.9291 0.3979

0.3714 0.9285 0.400C

0.3730 0,9278 0,4020

0.3746 0.3762 0.3778 0.3795 0.3811 0.3827 0.9272 0.9265 0.9259 0.9252 0.9245 0.9239 0.4040 0.4061 0.4081 0.4101 0.4122 0.4142

0.3843

0.3859 0.9225

0.387:

0.9232 0.4183

0.4183

0.4204

0.3891 0.9212 0.4224

0,3907 0.3923 0.3939 0.3955 0,3971 0,3987 0.9205 0.9198 0.9191 0.9184 0.9178 0.9171 0.4245 0.4265 0.4286 0.4307 0.4327 0.4348

0.4003 0.9164 0.4369

0.4019 0.9157 0.4390

0.4036 0.915C 0.4411

0.4051 0.9143 0.4431

0.4067 0.4083 0.4099 0.4115 0.4131 0.9135 0.9128 0.9121 0.9114 0.9107 0.4452 0.4473 0.4494 0.4515 0.4538

0.4147

0.4163

0.4179

0.9100

0.9092

o.ma

0.9OaS

0.4557

0.4599

0.4195 0.9076 0.4621

0.4210 0.9070 0.4642

0.4226 0.4242 0,4258 0,4274 0.4289 0.9083 n . 9 0 ~ 6 0.9048 0 . 9 ~ 10.8033 0.4663 0.4684 0.4706 0.4727 0.4748

0.4305

0.4321 0.9018 0.4791

0.4337 0.9011 0.4813

0.4352 0,9003 0.4834

0.4368

0.4770

0.4384 0.4399 0.4415 0.4431 0.4446 0.8988 0.8980 0.8973 0.8965 0.8957 0.4877 0.4899 0.4921 0.4942 0.4964

0.4462 0.8949 0.4986

0.4478 0.8942 0.5008

0.4493 0.8934 0.5029

0.4509 0.8928 0.5051

0.4524 0.8918 0.5073

0.4540 0.4555 0,4571 0.4586 0,4602 0.8910 0.8902 0.8894 0.8886 0,8878 0.5095 0.5117 0.5139 0.5161 0.5184

0.4617 0,8870

0.5206

0.4633 0.8882 0.5228

0.4648 0.8854 0.5250

0.4664 0.8846 0.5272

0,4679 0,8838 0.5295

co8 tan

0.4695 0.4710 0.4726 0.4741 0,4756 0.8829 0,8821 0.8813 0.8805 0.8796 0.5317 0.5340 0.5362 0.5384 0.5407

0.47TP 0.8788 0.5430

0.4787 0.8780 0.5452

0.4802 0.8771 0.5475

0.4818 0.8763 0.5498

0.4833 0.8755 0.5520

ain em tan

0.4848 0.4863 0.4879 0.4894 0,4909 0.8746 0.8738 0.8729 0.8721 0.8712 0.5543 0.5566 0.5589 0.5612 0.5635

0.4924 0.8704 0.5658

0.4939 0.8695 0.5681

0.4955 0.8886 0.5704

0.4970 0.8678 0.5727

0.4955

sin

0.5000 0.5015 0.5030 0.5045 o.soeo 0.8860 0.8652 0.8643 0.8634 0.8625 0.5774 0.5797 0.5820 0.5844 0.5867

0 . 5 ~ 5 0.5wo 0.8616 0 .86m 0.5914 0.5890

3.5105 3.8599 3.5938

0.5120 0.8590

0.5135 0.8581 0.5985

0.5150 0.5165 0.5180 0.5195 0.5210 0.5225 0.8572 0.8563 0.8554 0.8545 0.8536 0.8526 0.6009 0.6032 0.6056 0.M180 0.8104 0.6128

0.5240 0.8517 0.6152

1.5255 1.8508 1.6176

0.5270 0.8499

0.5299 0.5314 0.5329 0.5344 0,5358 0.8480 0.8471 0.8462 0.8453 0.8443 0.6249 0.6273 0.6297 0.6322'0.6346

0.5373 0.8434 0.6371

0.5388 0.8425

1.5402 1.8415 1.6420

0.5417 0.8406 0.6445

0.5432

0.5446 0.5461 0.5478 0.5480 0.5505 0.8387 0.8377 0.8368 0.8358 0.8348 0.6494 0.6519 0.8544 0.6568 0.6594

0.5518

0.5534

1.5548 1.8320

0.5577 0.83M) 0.6720 0.5721 0.8202 0.6976

00s

tan em

tan sin DOS

tan

*in am

tan

ain 00s

tan

24

sin cos

tan 25

sin 008

tan

26

sin 008

tan 27

sin Eo8

tan

28

29 30

sin

COB

tan

31

ain 008

tan 32

sin 008

tan

33

sin COI

tan 34

sin OOLI

tan 35

-

0.9 ~

0.289C 0.9572

sin

23

~~

0.3000

sin

22

0.8

0.2874 0.9578

18

21

0.7

0.2857 0.9583 0.2881

17

20

__

0.6 ~

16

19

i&fT ~

0.4

bin

ws 18"

0.3502 0.9367 0.3739

0.0026

0.9361

0.6395

0.921s

0.5961

0.6200

0.6644

1.6669

0.5563 0.8310 0.6694

0.5592 0,5606 0.5621 0,5635 0.5650 0,5664 0.8290 0.8281 0.8271 0.8261 0,8251 0.8241 0.6745 0.0771 0.6796 0.6822 0.8847 0.6873

0.5678 0.8231 D.6899

1.5693

0.5707

0 573C U S750 0 5764 0 7778 0 5703 n RIQZ 11 w,i 19 h i 7 1 o 8161 o u i i i 0 7W1 0 7028 U 70>4 U 7ud0 11 7107

0.5821 0.8131

1.5816 1.8121 1.7186

0.8339 0.6619

U 5807

n

ni41

0 7133

0.8328

0.1158

1.8221 1.6824

0,8211

0.6950 0.5850 0.8111 0.7212

0.8996 0.4856

0.8669 0,5750

0.5284 0.8490 0.6224

0.8196 0.6469

0.5884 0.8100 0.7239

-

946

Angle deEre-

38

Natural Trigonometric Function 36.C55.9' Value of function or each tr

IbmB

of !no-

0.0

0.1

0.2

0.3

0.4

0.5

.lo"

gin OM

tan

0.5878 0.5892 0.5906 0.5120 0,5934 0.5948 0.8090 0.8080 0.8070 0.8059 0.8049 0.8039 0.7265 0.7292 0,7319 0.7346 0.7373 0.7400

0.6

0.7

15962 1.5918 1.8028 1.7427

1.8018 1.7454

0.9

0.8

_-

~

'.5990

0.8004

8,7481

0.7508

,.8007 0.7997

tan

0.6018 0.6032 0,6046 0.606@ 0.6074 0.7986 0.7976 0.7985 0.7950 0.1944 0.7536 0.7563 0.7590 0.7618 0.7646

0.6083 0.7934 0.7673

1.6101 1.7923 1.7701

1.6115 1.7912 1.7729

8,6129 0.6143 1 . l W 2 0.7881 1.7757 0.7785

38

ain coa tan

0.6157 0.6170 0.6184 0,6198 0.6211 0.7880 0.7869 0.7859 0.7848 0.7837

0.6225 0.7826 0.7954

1.6239 1.7815

1.6252 1.7804 1.8012

1.6266 1.7793

0.7813 0.7841 0.7869 0.7898 0.7926

1.8040

0.6280 0.7782 0.80B9

sin

0.6283 0.63W 0.6320 0,6334 0.6347 0.6361

39

COQ

1.6374 1.7705 I. 8373

1.6388 1.7894

0.6414 0.7672 0.8361

iin

37

om

1.7983

I.8302

1.6401 1.7683 1.8332

0.6428 0.6441 0.6455 0,6468 0.6481 0.6494 0.1660 0.7648 0.7638 0.7627 0.7615 0.7604 0.8391 0.8421 0.8451 0.8481 0.8511 0.8541

1.6508 1.7593

).8571

1.6521 1.7581 j.8601

1.6534 1.7570 1.8832

0.6547 0.7559 0.8862

0.6561 0,6574 0,6587 0.6600 0.6613 0.6626 0,7574 0.7536 0,7524 0.7513 0.7501 0.7490 0.8893 0.8724 0.8754 0.8785 0.8816 0.8847

1.6638 1.7478 1.8878

j.6653 ).7466 1.8910

1.6665 1.7455 1.8911

0.8678 0.7443 0.8972

tan

om

0.6681 0.6704 0.6717 0.6730 0.6741 0.6756 0.7431 0.7420 0.7408 0.7398 0.7385 0.7373 0.9W4 0.8036 0.9067 0.9089 0.9131 0.9163

1.6769 1.7361 1.9195

1.6782 1.7349 ).9228

l.6794 1.7337 ).9260

0.6807 0.7325 0.9293

43

sin cm tan

0.6820 0.6833 0.6845 0,6858 0.6871 0.7314 0.7302 0.7290 0.7278 0.7266 0.9325 0.9358 0.9391 0.9424 0.9457

0.6884 0.7254 0.9490

3.6896 3.1242 3.9523

).6909 ).7230 1.9556

1.6921 1.7218 1.9590

0.6934 0.7206 0.9623

44

ain cm tan

0.6947 0.6958 0.6912 0.6984 0,6897 0.7008 0.1193 0.7181 0.7169 0.7151 0.7145 0.7133 0.9657 0.9691 0.9725 0.9759 0.9793 0.9827

1.7022 1.7120 D.9881

1.7034 1.7108

1.7046 1.7086

).8930

0.7059 0.7083 0.8865

ain

0.7145 0.6997 1.0212

1.0247

1.7169 1.6972 1.0283

0.7181 0.6959 1.0319

tan aio

40

DOS

tan

sin 41

008

tan

42

sin

0.7771 0.7760 0.7749 0,7738 0.7727 0.8098 0.8127 0.8156 0,8185 0.8214

0.7716 0.8243

3.9896 0.7157

tam

0.7071 0.7083 0.7096 0.7108 0.7120 0.7133 0.7071 0.7059 0.7046 0.7034 0.7022 0.7008 1.wOO 1 . m 5 1.0070 1,0105 1.0141 1.0176

ain em tan

0.7193 0.7206 0.7218 0.7230 0.7242 0.6947 0.6934 0.6921 0.6908 0,6896 1.0355 1.0392 1.0428 1,0464 1.0501

0.7254 0.6884 1.0538

0.7266 0.6871 1.0575

0.7278 D.6858 1.0612

1.7290 1.6845 1.0649

0.7302 0.6833 1.0686

sin tan

0.7314 0.7325 0.7337 0.7349 0.7361 0.6820 0.6861 0.6794 0.6782 0.6769 1.0724 1.0761 1.0799 1.0837 1.0875

0.7373 0.6756 1.0813

0.7385 0.6743 1.0851

0.7396 0.6730 1.0890

0.7408 0.6717 1.1028

0.7420 0.6704 1.1067

sin cm tan

0.7431 0.7443 0.7455 0.7466 0.7478 0.6691 0.6678 0.6685 0.6652 0.6639 1.1106 1.1145 1.1184-1.1224 1.1263

0.7464 0.6626 1.1303

0.7501 0.6613 1.1343

0.7513

48

1.1383

0.7524 0.6587 1.1423

0.7536 0.6574 1.1463

ain cm tan

0.7547 0.7559 0.1570 0.7581 0.7593 0.7604 0.6581 0.6547 0.6534 0.6521 0.6508 0.1.1504 1.1544 1.1585 1.1626 1.1667 1.1708

0.7615

0.7627

0.7638

49

0.6481

0.6468 0.6455

1.1750

1.1792

1.1833

0.7649 0.6441 1.1875

sin

0.7660 0.7672 0.7683 0.7694 0.7705 0.7716 0.6428 0.6414 0.6401 0.6388 0.6374 0.6361 1.1918 1.1960 1.2002 1.2045 1.2088 1.2131

0.7727 0.6347 1.2174

0.7738 0,6334

0.7749 0.6320 1.2261

0.7760

1.2218

0.7826 0.6225 1.2572

0.7837 0.6211 1.2617

0.784s 0.6198 1.2662

0.7858 (1.6184 1.2708

0.7869 0.6170 1.2753

0.7880 0.7891 0.1902 0.7912 0.7923 0.7934 0.6157 0.6143 0.6129 0.6115 0.6101 0.6088 1.2799 1.2846 1.2892 1.2938 1.2985 1.3032

0,7944 0.6074

0.7955

0.7965 0.6046 1.3175

0.7976 0.6032 1.3222

45

46

47

50

51

om

OM

em tan

ain CM

tan

52

ah

cm

tan

53

54

-

1.3079

0.66W

0.6OBo 1.3127

0.63G7 1.2305

0.7986 0.7991 0.8001 0.8018 0.8028 0.6018 0,6004 0.5980 0.5976 0.5962 1.3270 1.3319 1.3367 1.3416 1.3465

0.8039 0.8049 0.8059 0.5948 113514

0.5934 1.3564

05920 1.3613

0,8010 0.8080

ooe tan

1.3663

0.6892 1.3713

aia

0.8080 0.81W 0.8111 0.8121 0.8131 0.8141

ain

00s

tan 55

0.7771 0,7782 0.7783 0.7804 0,7815 0.6283 0.6280 0.6266 0.6252 0.6239 1.2348 1.2393 1.2437 1.2482 1.2527

0.6984

sin 00s

tan

0.5646

0.5878 0.5884 0.5850 0.5835 0.5821 1.3764 1.3814 1.3885 1.3916 1.3968

0.5801 1.4019

0.8151 0.5793 1.4071

0.8161 0.5779 1.4124

0.8171 0.5764 1.4176

0.8181 0.5750 1.4229

0.8192 0,8202 0.8211 0.8321 0.8231 0.5736 0.5721 0.5707 0.5693 0.5678 1.4281 1.4335 1.4388 1.4442 1.4496

0.8241 0.5664 1.4550

0.8251 0.5650 1.4605

0.8261 0.5635 1.4659

0.8271 0,5621 1.4715

0.8281 0.5606 1.4770

--

APPENDIX

Am! de

Ere

51

Natural Trigonometric Functions, name tion

0.0

0.1

0.2

0.3

0.4

56.W5.9'

X 6-z __

Value of function or each

Of

func-

0.5

~

0.6

-sin -0.8200 _ 0.8300 _ _ _ _ _ ~ ~0.8310 0.8320 0.8828 0.833 0.834 COB

tan ain

5:

C w

tan sin

a

ow

tan sin

58

947

D w

tan

0.7

- -

0.5582 0.5577 0.5563 0.5548 0.5534 0.551 1.4826 1.4882 1.4938 1.4884 1.5051 1.510

0.550 1.5168

0.835; 0.5401 1.522,

0.8387 0.8396 0.8406 0.8415 0.8425 0.843 0.5446 0.5432 0.5417 0.5402 0.5388 0.537, 1.5399 1.5458 1.5517 1.5577 1.5637 1.569

0.844 0.535; 1.575

0.845: 0.534. 1.5811

0.8450 0.8490 0.8488 0.8508 0.8517 0.5299 0.5284 0.5270 0.6255 0.5240 1.6003 1.6086 1.6128 1.6191 1.6255

0.852, 0.522.

0.8531 0.5211

0.854!

1.638:

1.644;

0.8572 0.8581 0.8590 0.8589 0.8607 0.5150 0.5135 0.5120 0.5105 0.5090 1.6643 1.6709 1.6775 1.6842 1.6909

0.8611 0.5071 1.687:

0.862! 0.5061 1,704)

1.6311

0.8

0.9

0.8361 0.5471 1.528:

m 7 7 0.5461 1.5340

0.846:

0.532s

0.8471 0.5314 1.5841

1.5881 0.855' 0.5181 1.651:

0.8563

0.8634 0.504) 1.711:

0.8641

0.8652

0.503( 1.7181

0.5015 1.7251

0.519!

0.5165

1.6577

60

sin ow tan

0.8660 0.8669 0.8678 0.8686 0.8695

0.5000 0.4985 0.4870 0.4855 0.4839 1.7321 1.7391 1.7461 1.7532 1.1603

0.870. 0.482l 1.767!

0.871: 0.4901 1.774;

0.8721 0.4984 1.782C

0.8726 0.4876 1.7883

0.8738 3.4863 I ,7966

61

sin cw tan

0.8746 0.8755 0.8763 0.8771 0.8780 0.4545 0.4833 0.4818 0.4502 0.4787 1.8040 1.8115 1.8190 1.8265 1.8341

0.8781 0.477: 1.8411

0.878t 0.475L 1.8491

0.880: 0.4741 1.8572

0.8813 9.4726 1.8650

1.8821 1.4710 I.8728

sin

0.8829 0.8818 0.8846 0.8854 0.8862 0.887( 0.8878 0.4695 0.4678 0.4664 0.4648 0.4633 0.4611 0.4602 1.88m 1.8887 1.8967 1.9047 1.9128 1 . 0 2 1 ~ 1.9282

0.8888 1.4586 1.9375

1. 8894 1.4571 I . 9458

) ,4555

0.8810 0.8918 0.8826 0.8934 0.8942 0.4540 0.4524 0.4508 0.4493 0.4478 1.8625 1.8711 1.8787 1.9583 l.9870

0.8849 0.4462 2.0051

D .n957 0.4446 2.0145

1.8965 1.4481 2.0233

).SO73 1.4415 1.0323

).8980 1.4399 !.0413

0.8958 0.8896 0.8003 0.9011 0.9018 0.4384 0.4368 0.4352 0.4337 0.4321 2.0503 2.0594 2.0686 2.0778 2.0872 0.9063 0.9070 0.9078 0.90e.5 0.9092

0.9026 0.4305

1.9033 ?.lo60

).9041 ).4274

).go45 ).4258

1.9056

2.0965

!.I155

!.I251 1.9121

1.4098 !.2251

62

cot/

tan sin

63 64

COB

tan sin 008

tan

sin 65

COB

ban

?in 66

10s

tan 67

sin 108

;an

/in 88

:OQ

.an

69

70

,in :DB

,an

in 8-

an 71

in 08

an

72

in

w

Bn

73

in OQ

m 'n

74

75

-

8 0

1.4289

).8802

I . 8542

I. 4242 ,.1348

0.4226 0.4210 0.4185 0.4178 0.4163 2.1445 2.1543 2.1642 2.1742 2.1842

0.81w 0.4147 2.1943

1.0107 1.4131

!.2045

1.8114 1.4115 !.2148

0.9135 0.8143 0.9150 0.9157 0.9164 0.4067 0.4051 0.4035 0.4018 0 . 4 w 3 2.2460 2.2566 2.2673 2.2781 2.2889

0.9171 0.3887 2.2998

).8178 1.3971

1.9184 1.3955

1.3220

l.0181 !.3939 ,.3332

0.9205 0.9212 0.9219 0.8225 0.9232 0.3907 0.3891 0.3875 0.3859 0.3843 2.3559 2.3673 2.3788 2.3906 2.4023

0.9239 0.3827 2.4142

1.8245 1.3811 1.4262

1.8252

1.8259

1.3795 '.4383

'.3778 4504

8.3162 -4627

0.9272 0.0278 0.9285 0.'9291 0 . 0 2 8 8 0.3746 0.3730 0.3714 0.3697 0.3681 2.4751 2.4876 2.5002 2.5129 2.5257

0.3665 2.5386

1.8311 1.3648 !.5517

1.9317 1.3633 ,5548

8.9323 '.3615

.8330 .3600

5782

,5916

18373

8.9379 '.3469 7034

,8385

,3453 ,7179

8391 ,3437 ,7326

8444 3289 ,8718

,9449 3272 8878

.9500

,9505 ,3107 .0585

1.3109

0.0304

0.9336 0.9342 0.8348 0.8354 0.9361 0.9367 0.3584 0.3567 0.3551 0.3535 0.3518 0.3502 2.6051 2.6187

2.6325 2.6464 2.6605

!.3496

2.6746

6888

1.8128

1.4083 ~.2355

,8.3923 8188 .3445

8.8255

0.9397 0.9403 0.0409 0.9415 0.9421 0.8426 0.3420 0.3404 0.3387 0.3371 0.3355 0.3338 2.7475 2.7625 2.7776 2.7829 2.8083 2.8238

8.8432 '.3322

'.3305

.a397

.a55

0.9455 0.8461 0.9266 0.9172 0.9178 0.3256 0.3239 0.3223 0.3206 0.3180 2.9042 2.9208 2.9375 2.8544 2.9714

,8159 ,3156

.0061

.Ql81 ,3140 ,0237

,3123 ,0415

,9542 .2890 1910

9548 ,2974 ,2106

,9553 2957 2305

.9558

9593 2823 ,3977

.8598 2807 ,4197

,9603 .2790 .4420

,9608 2773 ,4646

6806

9655 2605 7062

0.9511 0.9516 0.8521 0.9527 0.9532 0.3090 0.3074 0.3057 0.3040 0.3024 3.0777 3.0861 3.1146 3.1334 3.1524

0.9563 0.9568 0.8573 0.9578 0.9583 0.2924 0.2907 0.W80 0.2874 0.2557 3.2709 3.2914 3.3122 3.3332 3.3544

0.92s.3

0.3173 2.9387 0.9537

0.3W 3.1716 0.8588

0.2840 3.3759

0.9613 0.9617 0.9622 0.8827 0.9632 0.9636 0.2756 0.2740 0.2723 0.2706 0.2689 0.2672 3.4874 3.5105 3.5339 3.5576 3.5816 3.6058

1

0541 2556

6305

,9438

I

9646 2638 5554

I

.2940 2506

APPWDIX

PU

-

-

And de

Nan of tune tion

Kre

- sin 76

em

tan

sin 77

om

tan

78

sin em tan

sin

79

em

tan sin

80

00s

tan

Sin 81

COB

tan

ain 82

eon

tan sin

83

84

85

-

0.3

~

0.4

0.5

0.6

~

0.9703 0.9707 0.9711 0.9715 0.9720 0.9724 0.9728 0.2419 0.2402 0.2385 0.2368 0.2351 0.2334 0.2317 4.0108 4.0408 4.0713 4.1022 4.1335 4.1653 4.1976

0.7 0.9732 0.2300 4.2303

0.8

0.9

~

0.9736 0.2284 4.2615

0.9740 0.2267 4.2972

0.9744 0.9748 0.9751 0.9755 0.9759 0.9763 0.9767 0.9770 0.9774 0.2250 0.2232 0.2215 0.2198 0.2181 0.2164 0.2147 0.2130 0.2113 4.3315 4.3662 4.4015 4.4374 4.4737 4.5107 4.5483 4.5864 4.6252

0.9718 0.2096 4.6646

0.9781 0.9785 0.9789 0.9792 0.9796 0.9799 0 . 9 W 0.9806 0.9810 0.9813 0.2079 0.2062 0,2045 0,2028 0.2011 0.1994 0.1977 0.1959 0.1942 0.1825 4.7046 4.7453 4.7867 4.8288 4.8716 4.9152 4.9594 5.0045 5.0504 5.0970

0.9816 0.9820 0.9823 0.9826 0.9828 0.1908 0.1891 0,1874 0.1857 0.1840 5.1446 5.1929 5.2422 5.2924 5.3435

0.9833 O:l822 5.3955

0.9836 0.1805 5.4486

0.9839 0.1788 5.5026

0.9842 0.9845 0.1771 0.1754 5.5578 5.6140

0.9848 0.9851 0.9854 0.9857 0.9860 0.9863 0.9866 0.9869 0.9871 0.1736 0.1719 0.1102 0.1685 0.1668 0.1650 0.1633 0.1616 0.1599 5.6713 5.7297 5.7894 5.8502 5.9124 5.9758 6.0405 6.1066 6.1742 0.9877 0.9880 0,9882 0.9885 0.9888 0.1564 0.1547 0.1530 0.1513 0.1495 6.3138 6.3859 6.4596 6.5350 6.6122

0.9874 0.1582 6.2432

0.9890 0.9893 0.9895 0.9898 0,9900 0.1478 0.1461 0.1444 0.1426 0.1409 6.6912 6.7720 6.8548 6.9395 7.0264

0.9903 0.9905 0.9907 0.9910 0.9912 0.9914 0.9917 0.1392 0.1374 0.1357 0.1340 0.1323 0.1305 0.1288 7.1154 7.2066 7.3042 7.3962 7.4947 7.5958 7.6996

0.9919 0.9921 0.9923 0.1271 0.1253 0,1236 7.8062 7.9158 8.0285

0.9945 0.9947 0.9949 0.9951 0.9952 0.9954 0.9956 0.9957 0.9959 0.9960 0.1045 0.1028 0.1011 0.0993 0.0976 0.0958 0.0941 0.0924 0.0906 0.0889 9.5144 9.6768 9.844810.02 10.20 10.78 10.99 10.39 10.58 11.20

sin

0.9962 0.9963 0.9965 0.9966 0.9968 0.9969 0.9971 0.9972 0.9973 0.9Q74 0.0872 0.0854 0.0837 0.0819 0.0802 0.0785 0.0767 0.0750 0.0732 0.0715 1.43 11.66 11.91 12.16 12.43 12.71 13.04 13.30 13.62 13.95

ces

sin co* tan DO8

sin om

tan

89

0.2

*in em tan

tan

88

0.1

degree

0.9925 0.9928 0.9930 0.9932 0.9934 0.9936 0.9938 0.9940 0.9942 0.9943 0.1219 0.1201 0.1184 0.1167 0.1149 0.1132 0.1115 0.1097 0.1080 0,1063 8.1443 8.2636 8.3863 8.5126 8.6427 8.7769 8.9152 9.0579 9.2052 9.3572

sin

87

0.0

~

B

coa tan

tan

86

Value of function for each tenth of

sin COB

tan

0.9976 0.9977 0.9978 0.9979 0.9980 0.9981 0.9982 0.9983 0.9984 0.9985 0.0698 0.0680 0.0663 0.0645 0.0628 0.0610 0.0593 0.0576 0.0558 0.0541 4.30 14.67 15.06 15.46 15.89 16.35 16.83 17.34 17.89 18.46 0.9986 0.9987 0,9988 0.91/89 0.9990 0.9990 0.9991 0.9992 0.9993 0.9993 0.0523 0.0506 0,0488 0.0471 0.0454 0.0436 0.0419 0.0401 0.0384 0.0366 9.08 19.74 20.45 21.20 22.02 22.90 23.86 24.90 26.03 27.27 0.9994 0.9995 0.9995 0.9996 0.9996 0.9997 0.9997 0.9997 0.9998 0.9998 0.0349 0.0332 0.0314 0.1297 0.0279 0.0262 0.0244 0.0227 0,0209 0.0192 8.64 30.14 31.82 33.69 35.80 38.19 40.92 44.07 47.74 52.08 D.9998 0.9999 0.9999 0.9999 0.9999 1.000 I.WO 1.040 1.000 1.OW 0.0175 0.0157 0.0140 0,0122 0.0105 0.0487 0.0070 0.0052 0.0035 0.0017 7.29 63.66 71.62 81.85 95.49 114.6 143.2 191.0 286.5 573.0

Index Bus, arrangements of, in primary systems sectionalized bus, 717-718 star bus, 717 straight hus, 717 synchronizing bus, 717, 725-730 busway, low-voltage, 697498 for feeders, 702 plug-in type, 701 reactance of. 101 short-circuit current estimating curve for, 76-77, 8W81 medium-voltage, cost estimating d a t a for (2.4-13.8 kv indoor), 916 installation casts for, 894 capacitors connected to, 482 ground, 406, 408-417 neutral, 373 reactance of, 101-104 synchronizing, for modernization and expansion of power systems, 776,

“a,” the operator, function of, 118-119 Ahbreviations of electrical terms, 930 Aerial cable, in primary systems, 723-

724 protection of, against overvoltages from lightning, 326 (See also Cable) Air circuit breakers (see Circuit breakers) Arcing ground as intermittent line-toground short circuit, 286-289, 337-

338, 340-341 (See also Overvoltages)

Arresters, lightning (see Lightning arresters) Asymmetrical short-circuit current, d-c component of, 15-18 explanation of, 10-15 rms value of, 20 Autotransformers, groundibg of, 391-395 in light and power systems for commercial-type buildings, 749 neutral inversion of, 391P394 overvoltages caused by, 295-297, 391-

790, 794, 799 in primary power systems, 615,

717, 725-730 voltage drop in, calculation of, 242-

395 in primary substations, 711 Auxiliaries, power-plant,, 831-832, 882 Available short-circuit current, explanation of, 36-37 (See also Short-eircuit current) Balancing reactors, 784-788 BIL (basic impulse insulation level),

244 Bus-tie circuit breakers, 717-718 (See also Load-center power distribotion systems) Busway (see Bus) By-product fuel, 809, 811 By-product power (see Generation of electric power)

303-305 Cable, aerial, in primary systems, 723-

724 efficiency of, 831 supplementary-fired, 820, 867 Branch circuit design (see Commercialtype buildings; Factories) Buildings, light and power systems far (see Commercial-type buildings; Factories) Bus, arrangements of, in primary systems, 717-718 double bus, 715 ring bus, 717. 730

protection of, against overvoltages caused by lightning, 326 copper, fusing time of, 189-190 cost-estimating data for, 917-919 current-carrying capacities for, table of, 943 failure rates of, 630 heating of, during short circuit, 18W

183 impedance, reactance, resistance of,

96-110 949

950

INDEX

Cable, interlocked armor, in primary systems, 724 in secondary systems, 699-701 protection of, by current limiting fuses, 186 replacing old, 767 selection of, based on short-circuit current, curves for, 185 examples of, 186-190 tables of minimum conductor size for, 182 for capacitors, 487-488 in short-circuit calculations, eonsideration of, 45 short-circuit current limited by, 7677, 8&8l, 83-84, 87-90 synthetic-rubber-insulated, 698-699 temperature limits of, during short circuit, IX(t181 voltage drop in, calculation of, 237241 caused by motor starting, calculation of, 261-263 Capacitively grounded system (see Ungrounded system) Capacitors, constants for, 502 formulas for, 501-502 series, far power service, definition of, 492494 effect of, on voltage drop due to motor starting, 264 limitations of, 498 principle of operation of, 494-495 use of, 495-498 shunt, automatic switching of, 489492 on buses, 456, 482 cable for, selection of, 4 8 7 4 8 8 discharge of, 500-501 in equipments, kilovsr ratings of, 498500 harmonics and resonance dueto, 488489 &L induction motor terminals, limitations of, 477-479 location of, 477 ratings of, 4 7 M 8 2 thermal-overlortd current reduced by, 482 with induction motors v8. synchronous motor, 482-484 as kilovar generators, 449 kilovar ratings of, 49&500 location of, 4 5 6 4 5 8 operating characteristics of, 501 power-hill savings with, 459-460 poweraystem losses reduced by, 470-471 ~~

~

Capacitors, shunt, switching devices for, ampere ratings of, 485-487 interrupting ratings of, 487 repetitive duty of, 487 system capacity released by, 462468 economics of, 468-470 voltage ratings of, 199, 499-500 voltage rise due to, 4 7 1 4 7 4 voltage spread on, effect of, 214 (See also Power factor, improvement of) surge protective, for rotating machines, 31&311 application of, 32G333 Cascading low-voltage circuit breakers, rules far, 155-161 selection table for, 162-163 vs. selective tripping, 155-161 Case histories, of expanding power systems at higher voltage, 758-759 of overvoltages, on ungroundedneutral systems, 337-338, 401402 unusual, 297-299 of protective-device failures, 146, 147 Circuit arrangements (see Bus; Loadcenter power distribution systems; Neutral circuit arrangements far grounding; Power systems) Circuit breakers, air, large (600 volts and less), application of, 154161 application table for, 162-163 cascading of, 155-161 description of, 154 rating, basis of, 35 ratings of, 154 selection of, 161-163 selective tripping of, 15&156 standards for, 164 v8. switches and fuses, 153, 169170 basic characteristics of, 14&149 direct-acting trips on (see Direetacting trips) vs. fuses, 153 interrupting rating of, explanation of, 27, 35 low-voltage (600 volts and below), 154-166 interrupting duty on, calculation of, 32, 33, 35 momentary duty on, calculation of, 32, 33, 35 reactance of, 110

INDEX

Circuit breakers, low-voltage (GOO volts and below), short-circuit-current rating basis for, 35 main, in load-center substations, 674676 molded-case, 164-166 application of, 165 application table for, 162-163 cascading of, 165 direct-acting trips OD, 537, 54&541 ratings of, 164, 165 selection of, 166 stendards for, 166 momentary rating of, explanation of, 35 power (above 600 volts), application of, 172 application limits for, on short-eirwit-current basis, 28-29 interrupting duty on, checking of, 29,32 momentary duty an, checking of, 29,32 BS motor starters, 172 rating structure for, 24-28 ratings of, 171 selection of, 173 on GW-volt systems, application of, 35, 36 standards for, 174 types of, 171 rating, basis of, 24-28 sectionslizing, in loop feeders, 72%

721 in primary buses, 717-718 short-circuit-eurrent calculation for application of, 29-31, 35-38 short-time rating of, 27 (See also Short-circuit-current rating) switching arrangements for, in generating stations, 725-726 in main substations, 712-715 in primary buses, 715-718 in primary feeders, 718-723 types of, used in primary systems, 708-711 Codes and standards, for circuit breakers, air, large, 164 molded-case, 166 ~ o w e r 174 . for commercial-type buildings, restrictions of, 752-753 for cutouts, 178 for fuses, low-voltage, 170 power, 178 for lighting equipment, voltage limited by, 752-753

951

Codes and standards, for overcurrent protection of load-center systems, 632 for power systems, secondary, 684 for switches, low-voltage, 170 on system voltage selection, effect of, 576 for transformer protection, 632 Combined light and power systems (see Commercial-type buildings; Faetories) Commercial-type buildings, definition of, 731 light and power systems for, autotransformers in. 749 economic comparison of 480Y/277vs. 208Y/120-valt, 737-747 feeder circuits used in, 742-743, 745 lighting loads used in, 738-739 power loads used in, 739 grounding transformer for, connections of, 750 reactance of, 750 higher voltage, types and applications of, 732-735 incoming service for, types of, 731-

__

73s .

lighting-load estimates for, 885-886 National Electrical Code requirements for, 752-753 overcurrent protection for, 751-752 current-limiting fuses as, 751752 remote-control wiring system for, 74&742 Rexibilitv of. 740. 7.53-7.55 120/240-v&, j31 208Y/12&volt, 731 480Y/277-volt, components of, 7 3 6 737 fundamentals of, 735-736 12Gvolt load on, economic limit for, 747 power services for, 748-751 Conductor (see B w ; Cable; Grounding conductor; Neutral eonductor; Overhead lines) Control circuits, motor, on groundedneutral systems, 39&391 a n line-grounded systems, 390-391 safety of, 389-391 on ungrounded systems, 390-391 Conversion factors, 925-928 for amperes to kw or kva, 64 for horsepower to kva, 39 for ohms to per cent or per-unit, 38, 40 ~I

~~

~~

952

IN1>M

Coordination of relays and other protective devices (see Overcurrentprotective-device coordination) Corner-of-delta grounding, 388-389 Costa, estimatingdatafor, for busway, 916 for cable, 917-919 for generators, turbine, 909 for motor control, 914, 915, 922924 far motors, induction, 920,922, 923 synchronous, 921, 924 restrictions on use of, 892 for switchgear, low-voltage, 915 medium-voltage, 914 for transformers, 907, 908 for unit substations, load-center, 912 master, 910, 911 primary switching for, 913 for factory combined light and power systems, 582-587, 589 fixed charges in, 809, 875, 883 initial investment, return on, 809 installation, factory-assembled switching equipment, 893 field-assembled switching equipment, 893 large generating units, effect of, 80W310 metal-enclosed bus, 894 motor and control, 893, 894 steam plant, 881-882 transformer. R93 of load-center power distribution systems, 623-624, 898-900 radial, 636 secondary network, 65WG52 secondary selective. 63&640. 650652 spot secondary selective, 642 . maintenance, steam plant, 881 for modernization and expansion of power systems, examples of, 764796 of piecemeal equipment, 894-895 power, reduction by shunt capacitors of, 459-460 power generation, by-product, 811, 875-878 condensing steam, 878 fixed charges on, effect of, 875, 882883 labor on, effect of, 875, 88&881 plant factor on, effect of, 875, 880 Current-limiting fuses (see Fuses) Current transformers, reactance of, 110-112 saturation of, effect on relay perfarmsnee of, 559-561 I~

Cutouts, application of, 176 standards for, 178 types of, 174-176 (See also Oil-fused cutouts) Decrement, definition of, 1&19 fault current, effect of, on relay operation, 546-548, 561-567 Demand factors for load estimating, 889 Detector, ground, use of, 344 on portable-equipment circuits, 433 Device numbers and functions for switchgem, 930-932 Diesel-engine power plants, far industrial generation, 809, 874 for power generation, 873-874 for process heat, 873-874 Differentid relays (see Relays, differential) Direct-acting trips, a n air circuit breakers (600 volts and less). .. 533-537 operating currents of, 548 discussion of, 532-533 pickup settings of, 534 on power circuit breakers (above 600 volts), 533 reset time of, 536 thermal-magnetic type, on moldedcase breakers, 537, 54&541 time-current curves of, 535-540 coordination of, 537, 5 6 8 5 7 3 Direct-current component of short-circuit current (see Short-circuit current) Directional relays (see Relays, directional) Directly grounded system (see Grounding, solid) Disconnecting switches, reactance of, 110 (See also Switches) Distribution systems (see Commercialtype buildings; Factories; h a d center power distribution systems; Modernization and expansion of power systems; Power systems) Distribution transformers, impedance of, 96-97 (See also Transformers) Double-bus arrangement, 715 Double-throw interrupter switches on unit substations, 655-664 Dry-type transformers vs. liquid-filled, 671-672 Duplex reactors, equivalent circuit for, 65-66 use of, with large generators, 730, 784, 793

INDEX

Economic comparison of power systems, for commercial-type buildings, 737748 estimating data far, 892-895, 906924 for factory combined light and power systems, 582-587 for load-center systems; radial, secondary selective, and secondarv network. 9 W 9 0 5 for piecemeal equipment vs. loadcenter, 89G900 Efficiency, of by-product power generation, 811 engine, definition of, 844 of gas turbine, 861 importance of; 833 plant thermal, definition of, 818 power-plant, gas-turbine, 862 steam, actual, 822-823, 832-834 theoretical, 823-824 of steam boilers, 831 of steam turbine, 820, 822 Electric power generation (see Generation of electric power) Electric utility, grounding of, 366-367, ROF,

industrial ties with, 707-708, 712-715, 725, 815 Electrical fires, causes of, 406, 415 due to motective-device failure, 1 4 6 14? Electrode, grounding, definition of, 406 types of, 410 water pipe for, minimum size of, 414 Electronic equipment, voltage spread on, effect of, 211-213 Enthalpy, 828 Equipment grounding (see Grounding, equipment) Equivalent circuits, for duplex reactors, 65-66 of impedance diagrams, 5C-51 for symmetrical-component analysis, of basic sequence systems, 126 of double line-to-ground short circuit, 135 of line-to-ground short circuit, 133 of line-to-line short circuit, 130 of three-phase short circuit, 127 for three-winding transformers, 66 transformer zero sequence, 135-140 Exhaust-heat recovery boilers, use of, 820, 866 Expansion line (Mollier chart), 830 Expansion of power distribution systems (see Modernization and expansion of power systems)

953

Extraction turbines (see Turbines, steam) Factories, combined light and power systems for, economic comparisons of, 583-587, 589 480-120-volt transformers for, 593-595 higher voltage lighting for, 582-592 lamps in, lighting-system voltage for, 587 lighting-load estimates for, 885-886 office lighting supplied from, 592 other lighting supplied from, 592 for reliability, design of, 590-592 short-circuit protection of, 587-590 system considerations for, 587-595 transformers (480-120 volts) for, 593-595 277-volt lighting systems for, 582595 voltage flicker control in, 590 (See also Commercial-type buildings) Fall-of-potential method of measuring resistance to earth, 424-425 Fault current (see Short-circuit current) Fault location in ungrounded neutral system, 344-345 Feeder voltage regulators, application of, 224-225 reactance of, 112 Feeders (see Bus; Cable; Overhead lines; Power systems; Transmission circuits) Feedwater heating, 833, 862, 869 Fence, substation, grounding of, 417 Ferroresonrtnce, 284 Fires, electrical, causes of, 406, 415 due to protective-device failures, 146-147 Fixed charges in costs, 809, 875, 883 Flicker, voltage (see Voltage flicker) Fuel, by-product, 809, 811 gas-turbine, 871 Furnaces, electric resistance, power supply for, 595-597 voltage for, 575, 580 Fused switches (see Switches, fused) Fuses, characteristics of, basic, 15C-153 timecurrent, 530-532 coordination of, 531-532, 549 examples of, 568-573 current-limiting, 166170 application of, in lighting systems, 587, 590 for cable protection, 185-186 vs. large air circuit breakers, 16+170 overvoltages caused by, 294-295 performance of, 169-170

954

INDEX

Fuses, low-voltage (600 volts and below), 16~x70 application of, 167-170 vs. large air circuit breakers, 169-170 NEC t w e , 166 rating,b;asis of, 36 ratinns of. 166-167 short-circuit-current calculations for application of, 33, 36 standards for, 170 in switches, 166-170 types of, 166 power (above 600 volts), application of, in open structures, 176-177 for interrupter switches, 176 oil-cutout type of, 174-176 rating, basis of, 31 selection of, 178 short-circuit-current calculations for application of, 31-35 standards for, 178 types of, 174-176 rating, basis of, 31, 34-35 short-eireuit-current calculation for application of, 31-34 three-phase equivalent interrupting rating of, 34-35 (See also Switches, fused)

.,

I

Gas torhines (see Turbines, gas) Generating capacity, industrial and utility, 809 Generating stations, electrical layout of, 725-730 plant factor of, 808 (See also Generation of electric power; Generators; Psrellel operation of utility and industrial; Turbines) Generation of electric power, by-pfoduct, 810,812, 819, 831,840 efficiency of, 820 steam conditions on, eKeet of, 810, 819 for critical loads, 814, 817 cycles for, 818 industrial, economical arrangement for, 812 factors favoring, 811-812 in industrial plants, 809, 811, 875 vs. purchased power, 811-813 industrial use of, 808 installed capacity for, in industrial plants, 809 in utility plants, 809 plant factor for, 808 in utility plants, 809 Generator-transformer unit system, grounding of, 395-396

Generator-transformer unit system, use of, 730 voltage rating of, 198 Generators, balancing load in, 781-788 dielectric teats of, 305 double-winding, 730 minimum rating of, for system grounding, 366 neutral grounding of, 355, 357 protection of, against overvoltages caused by lightning, 326-333 reactance of, 92-94 explanation of, 8-10 short-circuit-current decrement of, 546-518. 562 effect on relay operation of, 546, ~I

~

~

561-567 ... ...

as short-circuit-current source, 3, 4 short-circuit protection of (see Overcurrent protection; Relays) short-circuit ratio of, definition of, 22 turbine, cost-estimating data for, 909 voltage drop of, due t o motor starting, calculation of, 253-260 voltage ratings of, 194, 197-198 Governing of steam turbine, by admission pressure, 852-851 automatic division of load by, 849851 by exhaust pressure, 852 by extraction pressure, 852-854 by initial pressure, 852 load regulator for, 849, 852 performance charts of, 853-854 speed corrector for, 849 speed governor for, regulation of, 849 by speed regulation, 849 Graphical symbols of electrical e q u i p ment for one-line diagrams, 932-942 Greek alphabet, 928 Ground bus, connections for, 413 definition of, 406 illustration of, 408 importance of, 410 installation of, 412-413 location of, 412-413 material for, 412 resistance to earth of, 414 size of, 411-412 for switchgear, 415, 417 Ground fault (see Short circuit, line-toground) Ground-fault current (see Short-circuit current) Ground-fault detectors, use of, 344 on portable-equipment circuits, 433 zero-sequence measuring circuits for, 142-143

INDEX

Ground-fault neutralizer, definition of, 340 overvoltages reduced by, 289 use of, 35&360 Ground indicators. use of. 344 zero-sequence measuring circuits for, 143-144 ~~~

~~~

Grounding, equipment, components for, 406-409 conductors for, 414 definition of, 339, 404 earth as a conductor for, 426 electrodes for, 410 ground bus for, 410-413 impedance in circuits for, 414-415 importance of, 405-406 of lighting fixtures, 42W421 of lightning arresters, 421 for lightning protection, 421-422 measuring resistance to earth for, 422426 of motor frames, 416 objectives of, 404-405 of outdoor circuit breakers, 416 of portable equipment, large, 426438 small, 419-420 of power-plant and dist,ribution equipment, 415-417 in secondary power systems, 694696 of semiportable equipment, 418 of small apparatus and devices, 419-420 of stationary utilization equipment, 417-419 of substation fence, 417 of switcbgear, 415-417 of transformer tanks, 416 ground-fault neutralizer, 354, 358360 line (corner-of-the-delta), 388-389 mid-phase, 389 neutral, of autotransformers, 391-395 circuit arrangement for, 370-374 equipment for, ratings of, 377-378 selection of, 378-388 of fused transformers, 712 grounding transformers for, 340, 348-352, 377-378, 385-388 methods of, 352-360 minimum rating of generators and power transformers for, 366 of portable-machinery oircuits, 428-429, 434 of public-utility supply, 395 reactors for, rating of, 378, 381385

955

Grounding, neutral, reactors for, selection of, 381-385 table for selection of. 388 resistors for, rating of, '377-381 selection of, 378-381 table for selection of, 379 selection of point for, 367-370 summary of practice for, 364-365 of systems, 600 volts and below, 36+364 2.4 kv to 15 kv, 363, 365 above 15 kv, 365, 366 reactance, calculation of ground-fault current for, 131-134, 376-377 calculation of reactor for, 381-385 definition of, 339 use of, 351, 357 resistance, calculation of ground-fault current for, 131-134, 375-376 definition of, 339 selection of resistor for, 378-381 use of, 354-357 resonant, definition of, 339-340 solid, calculation of ground-fault current for, 131-134, 377 definition of, 339 fault duty increased by, 396-399 table for estimating ground-fault current for, 362 use of, 352-355 system, advantages of, 345-348 definition of, 339 effect of, on lightning arrester selection, 314-315 equipment for, rating of, 377-378 selection of., 378-3RR ~.. industrial vs. utility practice in, 366367 influence of, on control circuit safety, 389-391 methods of, 352-366 selection of, 360-367 neutral for, obtaining, 348-352 operating experience with, examples of, 337-338, 401-402 overvoltages curbed by, 279, 280, 282, 285, 289, 296, 297 practice in, examples of, 399-401 system characteristics with various methods of, 354 Grounding conductor, connections of, 413 definition of, 406 distinction of, from neutral conductor, 409, 414 size of, 414 Grounding electrode, definition of, 406 types of, 410 water pipe for, minimum size of, 414 ~

~~~

~

~~~~~

~

~

916

INDU

Grounding of equipment (see Grounding, equipment) Grounding point, definition of, 339 selection of. 367-370 1See also Neutral m . i n t ). (hounding practice, wmniary of, :iti,l-:i65 utility w. irrduatrial, :iIiti-:iGi Groundirig reactors, rating of, :i78, 381385 selection hi, 381-385 table for, 388 Grounding resistors, rating of, 377-381 selection of, 37&381 tshle for, 379 Grounding transformers (see Transformers, grounding)

.

Harmonics, with capacitors, 488-489 caused by fluorescent lighting, 695 Higher voltage lighting systems (see Commercial-type buildings; Factories) Impedance, of bus, 101-104 of husway, 101 of cable, 96101 of circuit breakers, 110 combining of, method of, 60-63 of connected motors, 94-95 of current transformers, 116-112 of disconnect switches, 110 dividing of, 63 in equipment-grounding circuits, effect of, 414-415 of machines, induction, 91 synchronous, 91 multiplying of, 63 of overhead lines, 104-110 of reactors, 112 of regulators, feeder, 112 of rotating machines for symmetrical component analysis, 134, 135 in series, addition of, 66-61 short-circuit current limited by, 688689 surge, of cable, 301 definition of, 301 of overhead lines, 301 of transformers, distribution, 96 load-center, 97 power, 97 of transmission lines, 104-110 of turbine-generators, 92-94 zero-sequence, of transformers, 135-140 Impulse tests of insulation, 303-305 Indicator, ground, use of, 344 zero-sequence measuring circuits for, 142-143

Induction motors (see Motors, induction) Induction-type relays (see Relays) Injuries, electrical work, 405, 406 Installation costs (see Costs, installation) Insulation, basic impulse level of. 303-305 chopped-wave tests of, 305 dielectric tests of, 302-305 impulse tests of, 302-305 Interlocked-armor cable, in primary systems, 724 in secondary systems, 697, 699, 701 (See also Cable) Interlocking of interrupter switches, 631, 671 Interrupter switches (see Switches) Interrupting capacity (see Short-ciycuitcurrent rating) Interrupting duty (see Short-circuit current, calculation of) Interrupting rating (see Short-circuitcurrent ratinn) Inversion of neutral of autotransformers, 391-394

-_

Kilovars, capacitors a8 generators of, 449 control of, need for, 817 as unit of measurement, 440 (See also Capacitors, shunt) Lamps, fluorescent, harmonics from, 695 power requirements of, 885-886 voltage ratings of, 194, 198 voltage spread on, effect of, 208-210 incandescent, power requirements of, 885-886 voltage ratings of, 194, 198 voltane on. effect of. 207-208 I soread . mercury-vapor, voltage spread on, effect of, 210 voltage for, in factories, 582-587 L u g e air circuit breakers (see Circuit breakers) Letter symbols, for electrical terms, 92% 929 for heat and thermodynamic terms, 929-930 Light flicker (see Voltage flicker) Lighting, distribution systems for (see Commercial-type buildings; Factories) Lightning, grounding connections for protection against, 421422 masts for shielding against, 324-325 nature of, 299 overvoltages caused by, 299-302 on portahle equipment, 427, 435 protection against, for aerial cable, 326

P57

INDEX

Lightning, protection against, for a-c rotating machines, 326-333 for d-c rotating machines, 333 for indimtrial power systems, 299326 for metal-clad switchgear, 321-324 for rectifiers, 333-334 for substations, 324-326 for transformers, 319-321 traveling waves on transmission lines produced by, 301-302 voltage surges induced by, 300 Lightning arresters, 305-31B application proccdure for, 314-319 for direct-current systems, 312-313 discharge voltngc oi, 307-310 distribution-typc, 308-3 10 eflcct of altitude on sdcction of, 308 for effectively groundcd aystems, 315, 316 expulsion-type, :305-307 on grounded-neutral systcms, 345-346 grounding of, 421 high-voltage, 308-310 line-t y pe, 308-3 10 location of, 317-3I:l lolv-voltage, 311-3l2

opcrating principle of, 305-:iOG protective characteristics of, 307-3 I0 tolerances in, 308 for rcsistance-grounded systems, 314, 316 for rotating-machine protcction, 31031 I sparkovw voltage of, 307-3 10 station-typc, 308-3 10 t y p w of, selct.tion of, :307, :3l2,315-317 for ilngroundcd systcms, 314-316, 345-346 valve-typc, 305-807 voltage ratings of, 307-312 selrrtion of, :311-:315 Lightning protcctive eqiiiprncrit (see Capacitors,surgc protective i Lightriing arresters) Lightning S L ~ O K P St o transmission lines, direct and indnred, :iCU Line grounding (eornrr-of-dclta grounding), 388-389 Line-to-ground fault (see Short rimzit, line-to-ground j Load-break switches (see Switches, intcrrupter) Load-cmter power distrihutiun systems, advmtagcs of, 623-WI circuit arrangements for, automntirtransfer, secondary sdrrtivr, 643-645

Load-centcr power distribution systems, circuit arrangements for, cost cornparison of, 634-635 distrihuted secondary sclectivr, 643 load shifts in, caring for, 652-653 manual-transfer secondary selective, 645-646 primary d c e t i v e , 655-601 vs. radial, G51-658 vs. stxondary selective, 655 radial, G35-ti37 secondary network, 646-054 secondary sclectivc, G37-G46, 65& 654 YS. secondary network, 650-654 spot secondary nrtwork, 650 spot secondary selective, 642Mi43 types of, basic 629 relativc use of, 635 cost of, 623-624 rost-estimating data for, 898-900 ciconomic comparisons of, 896-905 above 15 kv primary vultagc, 621, 680-682

high-voltage (ahovr 600 volts), 621 looped primary iceders for, 661-664 overcurrcnt protcetion of, 632-634 principles of, 622-623 ahow 600 volts secondary voltage, 621, Gi!)-683 Load-center unit suhstation (see Unit suhststion, load-center) Load cstiniating, dPinarid factors for, table of, 889 m e of, 890-892 diversity factor for, 888 importnncc of, 881-885 for lighting, 885-886 load data for, actual, 8!11 load dcnsititts in rcpwsrntative industries for, 887 load factor for, 889 power consumption of industry for, 888

for power load, 88G-887 Load-ratio control, i i s ~of, 221, 814 Load rrgulators, turhine-gcnerator, 811, 819. 851

h o p e d primary iredrr systems, use of, 661-664, 718, 720-723 Iniv-voltagp air c i r c u i t hreakers (see Circuit hrcakers, Ion-voltage) Low-voltage switchgear, cost-cstimating data far, !J15 drawout type, application, 154-155 far unit snhstntions, li72-G74 (See also Circuit breakers)

PSB

INDEX

Low-voltage systems (see Commercialtype buildings; Factories; hfodernisation and expansion of power systems; Power systems, secondary) Magne-blast circuit breakers (see Circuit breakers, power) Magnetic air circuit breakers (see Circuit breakers, low-voltage) Master unit substation, cost-estimating data for, 910, 911 primary switching in, cost-estimating data for, 913 use of, 681-483, 815 Metal-clad switchgesr, cost-estimating data far, 914 in primary s y s t e m , 708-711 proteetion of, against overvoltages from lightning, 321-324 (See also Circuit breakers, power) Mid-phase grounding. 389 Modernization and expansion of power systems, without changing voltage, 797-799 double transformation eliminated in, 770-773 with generating systems, 773-797 with grid, high-voltage, 800-807 high voltage in, 758-763 advantages of, 759-761 lower cost of, 759 rated 600 volts, 777-784 rated 2400 volts, 78&796, 800 reduces short-circuit bottlenecks, 759-760 reduces voltage drop, 761 step-by-step basis of, 741 high-voltage backbone for, SOW807 load-center power systems in, 758 planning of, 757 rated 600 volts, 777-784 rated 2400 volts, 788-796 from 2.4 kv to 13.8 kv, 794-796 . Molded-ease circuit breakers (see Circuit breakers) hlollier chart, description of, 824, 828 examples explaining use of, 829-830, 837-839 expansion line on, 821, 830 state line on, 830 Momentary duty, 29, 32-33, 35 Momentary rating, 27, 35 Motor starters, for lighting control in factories, 587-588 selection of, above 600 volts, 179-180 for 600 volts and less, 179 short-cirenit-eurrpnt calculation for application of, 35, 36

Motor starters, types of, 178 Motors, control for, cost-estimating data for, 914, 915, 922-924 installation costs of, 893, 894 converting horsepower o f , to kvs, 39 cost-estimating data for, 920-924 dielectric tests of, 305 induction, with capacitor vs. synehronous motors, 482-484 cost-estimating data for, 920, 922, 923 installation coats for, 893, 894 limitations for capacitors on, 476479 power factor of, 474-476 ratio of, to synchronous, 95 reactance oi, 90-91 selection of capacitor rating for, 479481 as short-circuit-current source, 7-8 voltage ratings of, 104, 198 voltage spread on, effect of, 205-206 power systems for, 679-681 protection of, against overvoltages caused by lightning, 326-333 ratio of energized to installed, 95 reactance of, 90-91, 94-95 explanation of, 8-10 in short-circuit-eurrent calculations, 32-33, 4 6 5 0 starting of, calculation of voltage drop due to, 248-277 starting power factor of, 271-272 synchronous, cost-estimating data for, 921, 924 vs. induction with capacitors, 482484 installation casts for, 893, 894 kilovar output of, 452-453, 485 power-factor improvement by, 452453, 485 ratio of, to induction, 95 reactance of, 9C-91, 94-95 as short-circuit-current source, 5-6 voltage ratings of, 194, 198 voltage spread on, effect of, 206-207 voltage for, selection of, 600-611 (See also Control circuits, motor) Multiplying factors, short-circuitcurrent, table of, 32-33 use of, 20-22, 46, 43 Segative-phase-sequenee component, dcfinition of, 116-117 (See also Symmetrical-eomponent analysis) Setwork protector, 648

INDEX

Neutral circuit arrangements for grounding, of multiple power sources, 371374 practice recommended for, 370-374 selection of, 374 of single power source, 371 Neutral conductor, ,capacity of, 694-695 distinction of, from grounding eonduetor, 409, 414 Neutral grounding (see Grounding, neutral) Neutral inversion of autotransformer, 391-394 Neutral point (system neutral), definition of, 339 how to obtain, 348 selection of, 367-370 Neutralizer, ground-fault, definition of, 340 overvoltages reduced by, 289 use of, 358-360 Nomenelaturc, 928-930 Office buildings (see Commercial-type buildings; Factories) Ohms, conversion to per cent or per-unit, formula for, 38 table for, 40 Oil-fused cutouts, application of, 176 as interrupter switch and fuse, I76 on load-center unit substations, 670671 standards for, 178 One-line diagram, graphical symbols used in, 932-942 Open lines (see Overhead lines) Open-type switchgear, replacement of, 763-764 Operating experiences, of expanding power systems a t higher voltage, 758 of overvoltages, on ungrounded neutral systema, 337-338, 401-402 unusual, 297-299 of protective-device failures, 146-147 Ovens, electric, power supply for, 595-597 voltages for, 575, 580 Overcurrent protection, of combined light and power systems, 587-590 in commercial-type buildings, 751752 devices for, application of, typical, 542513 basic types of, 504 electromagnetic attraction, 506507 electromagnetic induction, 507-510 fault currents needed to predict operation of, 518-550

959

Overcurrent proteetion, devices for, operating-time classifications of, 510 (See also Direct-acting trips; Fuses; Relays) of load-center systems, 632-034 of portable-machinery circuits, 429-433 on power systems, general discussion of, 504-506, 542-543 in secondary power systcms, 685-689, 692-693 (See also Overcurrent-protectivedevice coordination) Overcurrent-protective-deviee eoordination, basic principles of, 543-544 data required for, 544-545 error currents due to current-tmnsformer saturation on, effect of, 553, 559-561 examples of, 568-573 fault-current calculations for, 545-546 fault currents needed for. 548-550 generator-current decrement on, effect of, 546-518 allowing for, 561-567 graphical proof of, 557-559 phase shift in Y-delta or delta-Y transformers on, cffeet of. 568 time-current curve8 for, air circuit breaker, 535, 538, 539 molded-case circuit breaker, 540 relay, inverse-time induction, 512 time and current settings for, current-balance relay, 567 differential relay, 507 direct-acting trip, air-eircuitbreaker, 534 molded-ease circuit-breaker, 541 general discussion of, 550 generator overcurrent relay, voltage-controlled, 567 voltage restraint, 561 ground-fault relays, residually eonneeted, 552-553 effect of error currents a n , 553 generator or trsnsformcr neutral, 554 on incoming lines and miscellaneous fccders, 552 instantaneous relays, 557 an single-motor fceders, 551 on transformer feeders. 55&551 wire-pilot relays, 567 time margins for, of direct-acting trips, 537 of fuses, 549 graphic examples of use of, 556, 564, 569, 572 of relays, 554-556

960

IN1

Overcurrent relays (see Relays, overcurrent) Overhead lines, in primary systems, 723-724 reactance of, 104-110 resistance of, 104-110 spacing of, 105, 108 voltage drop in, due to motor starting, calculation of, 261-263 Overvoltages, cause of, autotransformers as, 295-297, 391-395 contact with higher voltage systems as, 279-280 forced-current-zero interruption as, 294-295 grounded-Y broken-delta potential transformers as, 284 high-reactance line-to-ground connection as, 281-282 intermittent line-to-ground short circuit as, 286-289, 337-338, 340-341 liehtnine I as. , 299-302 neutral grounding of Y-Y transformers as, 295-297 resonant L-C circuits as, 281-285 static as, 279 switching surges as, 289-294 curbing of, by ground-fault neutralizer, 289 b y system neutral grounding, 279, 280, 282, 285, 289, 296, 297 on ungrounded systems, 278-299, 34&343 I

Panelboards in secondary power s y 6 tems, 692 Parallel operation of utility and industrial, 816-817 Per cent, conversion of ohms to, 38, 40 Per-unit method, base number for, selection of, 53, 55 comparison of, with per cent method, 53 formulas used in, 38, 54-57 ohms to, conversion of, 38 ohms used in., 54., 55 in symmetrieal-component analysis, value of, 124-125 Pilot-wire relays, operating principles of, 528-529 pilot-wire supervisory system for, 529 setting of, 567 (See also Relays) Plant factor, 808 Plastic-case circuit breakers (see Circuit breakers, molded-case) Plug-in busway, 701

Portable equipment, large, circuits for, cable skids in, 429 ground-dctector equipment on, 433 isolating transformcr in, 434 lightning protection for, 435 protectivc devices in, 429-433 relaying of, 484 simultaneous ground faults on, 433-434 ground circuit for, isolation of, 435-436 resistance of, 429 grounding of, 426-438 lightning disturbances on, 435 lightning protection for, 435 resistance to ground of, 427-429 shock hazard on, causes of, 427-429 safety rules for, 427 on ungrounded systems, 433-434 voltage rating for, 426 small, on secondary power systems, 694-696 disconnect means for, 696 (See also Grounding, equipment) Positive-phase-sequence component, definition of, 116-117 (See also Symmetrical-component analysis) Power centers (see Unit substations, load-center) Power circuit breakers (see Circuit breakers, power) Power-company service (see Purchased power) Power consumption in industry, 808,

-"-

PPP

Power factor, definition of, 441-442 fundamentals of, 440 of groups of loads, 444-446 improvement of, 446-448 advantages of, 459-474 calculation methods for, 448-449 kilowatt multiplier used in, 449-451 by capacitors with induction motors, 474-484 kilovar generators for, capacitors as, 449 location of, 4 5 6 4 5 8 synchronous machines as, 452453, 485 power-hill savings from, 4 5 9 4 6 0 power-system losses reduced by, 470-471 release of system capacity by, additional lording permitted by, 462-468

Pbl

INDEX

Power factor, irnprovcment of, release of system capacity by, definition of, 460 economics of capacitors for, 468470 examples of, 460-461, 4 6 7 4 6 8 by synchronous machines, 452-453, 485 of thermally overloaded apparatus, 460-461 voltage increased by, 471-474 (See also Capacitors, shunt) of induction motors, 474-476 leading and lagging, definition of, 443444 examples of, 443-444 low, causes of, 439 effects of, 439 measurement of, 453-456 by wattmeter readings, 454 of motor-starting current, 271-272 Power fuses (see Fuses, power) Power-generating stations, clectrical layout of, 725-730 plant factor of, 808 (See also Generation of electric power; Gencrators; Parallel operation of utilitv and industrial: Turbines) Power-generation costs (see costs, power generation) Power-plant efficiency (see Efficiency) Power supply (see Generation of electric power; Power systems, primary) Power systems, capacity released in, by capacitors, 462-468 for commercial-type buildings (see Commercial-type buildings) for critical loads, 814-817 economic comparisons of (see Economic comparisons of power systems) for factory combined light and power (see Factories) for lighting circuits, 704 load-center (see Load-center power distribution systems) losses in, reduction by capacitors of, 47M71 modernization and expansion of (see Modernization and expansion of power systems) for Dortable tools. 695 primary, bus arrangements used in, double, 715 ring, 717, 730 sectionalized, 717-718 star, 717 straight, 717 synchronizing, 717, 725-730

Power systems, primary, cable for, 723724 characteristics of, 706-707 circuit-breaker switching arrangements for, bus, 715-718 bu s-tie, 717-7 18 feeder, 718-723 generating station, 725-726 main substation, 712-715 circuit breakers used in, 708-711 feeders in, installation of, 723-724 loop, 661-664, 718, 720-723 parallel or alternate, 717, 719-720, 722, 723 radial, 718-720 tie, 718, 722 generating-station bus arrangements in, 725-730 generator-transformer unit system in, 730 generators in, double-winding, 730 main substations in, circuit arrangements of, 712-715 transformer connections for, 711712 overhead lines in, 723-724 purchased-power arrangements for, 707-708, 712-715, 725 reactors in, bus-tie, 730 generator duplex, 730 generator series, 730 synchronizing, 725-730 service reliability in, 706-707 switches in, manual, 721-722 switchgear in, metal-clad, 708-711 station-type, 70&710 transformers for, delta-Y-conneoted, 711 tertiary windings in, 711 three-winding, 711-712 Y-Y connections for, 711 protective devices for,short-eireuitcurrent (see Short-eircuit-current protective devices) relay protection for, typical layout of, 542543 secondary, cable for, interlockedarmor, 699-700 short-circuit capability of, 688 synthetic-rubber-insulated. 69% ~~

699

circuits for, cable in conduit, 69&699 interlocked-armor, 699-700 metal-enclosed, 6 9 7 4 9 8 open-wire, 697 codes for, 684 definition of, 684

INDEX

Pb2

Power systems, secondary, four-wire, fluorescent lighting on, 695 neutral conductor size in, 694695

impedance in, short-circuit current limited by, 688-692 large spot loads in, voltage for, 702 lighting circuits in, 704 load-center system on, effect of, 685686

for metal-fabricating-type plants, 701 motor control in, grouped, 702-703 overload protection of, Sational Electrical Code requirements for, 692493 panelboards in, 692 portable equipment in, disconnect means for, 696 grounding of, 694-696 for process-type plants, 701-704 radial circuits in, 704-705 reactors in, calculation for, 690491 disadvantages of. 689 location of, 689-690 short-circuit current limited by, 689-692 voltage drop in, 689, 691 safety to personnel in, 693494 short-circuit current protection for, 685-689 short-circuit current protection devices for, types of, 686-688 transformers far 120-volt power in, 593-595, 692, 696-697 voltage drop permissible in, 693 SOurce~in, limits on size of, 784, 793 voltage for, selection of (see Voltage, system, selection of) voltage drop in, example of, 244-248 due to motor starting, calculation of, 264-265 Power transformers, impedance of, 97 (See also Transformers) Primary power distribution systems (see Power systems, primary) Primary-selective load-center system, circuit arrangement of, 655-661 vs. radial system, 657-658 vs. secondary selective system, 655 (See also had-center power distribution systems) Process heat, characteristics of steam t o provide, 836 economics of indostrial generation affected by, 811

Process heat, gas-turbine exhaust recovery for, 820, 862, 864, 866 industrial requirements for, 820, 872 steam-turbine exhaust or extraction for, 862, 866 Protective devices, overcurrent (see Overcurrent protection, devices for; Short-circuit-current protective devices) overvoltage (see Lightning arrestera; Capacitors, surge protective) Purchased power, availability of, 813814 connections of, 707-708, 712-715, 725 for critical processes, 814-817 factors affecting use of, 811, 813 vs. industrial generation, 811-813 on plant site, effect of, 813-814 reliability of, 812, 814, 817 for remote loads, 812

RoIXo ratia, influence on transient overvoltages of, 387-388 Radial load-center System, circuit arrangement of, 635-637 (See also Load-center power distribution systems) Ratio method of measuring resistance to earth, 425-426 Reactance, of bus, 101-104 of buswrty, 101 of cable, 96-100 of circuit breakers, 110 combining of, method of, 58-60 of connected motors, 94-95 of current transformers, 110-112 of disconnect switches, 110 of machines, definition of, 8-10 induction, 91 synchronous, 91 of overhead lines, 104-110 of regulators, feeder, 112 in short-circuit-current calculetions, 44-45 suhtransient, definition of, 9 synchronous, definition of, 9 of transformers, distribution, 96 load-center, 97 power, 97 transient, definition of, 9 of transmission lines, 104-110 of turbinegenerators, 92-94 Reactance diagram, equivalent circuit, 50, 51 (See also Equivalent circuits) Reactance grounding, calci~lationof ground-fault current for, 376377

I/NDEX

Reactance grounding, calculation of reactor for, 381-385 definition of, 339

of, 354, 357 (See also Grounding, neutral) Reactors, balancing, 784-788 use

duplex, equivalent circuit for, 65-66 use of, with large generators, 730,

784, 793 grounding, rating of, 378, 381-385 selection of, 381-385 table for, 388 primary bus-tie, 730 reactance of, 112 in secondary power systems, 689-692 in series with generators, 730 synchronizing bus, short-circuit current limited by, 725-727 voltage drop caused by, 728-730 Rectifiers, protection of, against overvoltages caused by lightning, 333-

334 Reduced-frequency starting of motors, 971 _.I

Regulators, turbine load, 849-851 voltage (see Voltage regulators) Relaying (see Overcurrent-protectivedevice coordination) Relays, basic types of, 506 electromagnctic-attraction, 5 0 6

507 hinged-armature construction of,

506 operating currents of, 507 plunger (solenoid) construetion of, 507 electromagnetie-induction, 507-510 operating currents of, 507 burdens of, 514 effect on current trsnsiormers of,

559 contacts on, current carrying eapacity of, 508 coordination of, definition of, 543644

(See also Overcurrcnt-protectivedevice coordination) current-balance, operating principles

of, 529 setting of, 567 current required for operation of, 548 definition of, 505 differential, for bus protection, 526-

527 for generators, 521-524 operating principles of, 521 for transformers, 525-526 setting of, 567

963

Relays, directional, uses of, 517-518 overcurrent, operating principles of,

518-519 polarization of, 519 with voltagc restraint, 521 product-type, for ground faults, 519521 distance, operating principles of, 529-

530 ground-fault, in generator or transformer neutral, 554 residually connected, 552-553 holding coils for, 509 induction, general discussion of, 507-

515 inverse-time types of, comparison of characteristics of, 511-513 with extremely inverse-time characteristics, 512 with inverse-timc characteristics, 51 1 typical time-current BUIVCS for, 512 with very inverse-time charaeteristics, 511 overcurrent, current-tap range of, 51 I ,

514 ~~~

generator, with voltage restraint,

515-517 instantaneous hinged-armature 01 plungcr type of, 511 oueratine urineioles of. 511-515 reset timehf, 5i5 time-current operating C U ~ Y ~for, S 512-513

use of, 555555 time-delay induction, 511-515 tolerances of, operating, 511 voltage-controllrd, 517 setting of, 567 pickup current of, 514 residually connected ground-fault, 552-

553 seal-in circuit of, 50&510 target coils on, 509 time-dial (lever) on, 511 wire-pilot (a-e), 528-529 pilot-wire supervisorysystem for, 529 setting of, 567 Reuetitivo restrike as intermittent lineto-ground short circuit, 286-289,

337-338, 340-341 (See also Overvoltages) Resistance, of busway, 101 of cable, 98-101 of overhead lines, 104, 108 of transformers, 96 of transmission lines, 108 use of, for short-circuit-current calculations, 45-46

964

INDEX

Resistance t o earth, limiting values of, 414 methods of measuring, 422426 Resistance grounding, calculation of ground-fault current for, 375-376 definition of, 339 selection of resistor for, 378-381 use of, 354-357 (See also Grounding, neutral) Resistors, grounding, rating of, 377-381 selection of, 378-381 table for, 379 Resonant grounded, definition of, 339-

__

24n I

(See also Groundfault neutralizer) Restriking as intermittent line-to-ground short circuit, 286-289, 337-338, 340341 (See also Overvoltages) Ring bus for primary power systems, 717, 730 Rms value of short-circuit current, definition of, 20 Rotating machines (see Generators; Motors)

Safety to personnel, grounding of systems as related to, 345, 347, 348 on motor control circuits, 389-391 on portable equipment, 426429 on secondary power systems, 693-604 system voltage selection on, effect of, 575-576 Safety grounding (see Grounding, eauinmentl

Service reliability, of cables, 630 of combined light and power systems, 590-592 of primary selective systems, 657 in primary Systems, 706-707 of radial systems, 635-636 of secondary network systems, 653-654 of secondary selective systems, 637638, 653-651 of transformers, 630 Short circuit, bolted, for calculation of short-circuit current, 42-43 failure of circuit breaker due to, 1461 A7 __.

line-to-ground on ungrounded systom, detectors for, 344 intermittent type of, 340-341 location of, 344-345 locators for, 344 multiple occurrence of, 345 location of, for calculation of shortcircuit current, 43 (See also Short-circuit current, edculation of) Short-circuit current., asvmmetrical. d-c " component of, 15-18 explanation of, 12-15 rms value of, 20 available, explanation of, 36-37 cable conductor selection based on, 180-190 calculation of, base kva numbersfor, 55 bolted short circuits used in. 42-43 calculating board for, 58 for circuit-breaker application, 2931, 32-33, 35 combining impedances for, 60-63 combining series and parallel reactances for, 58-60 converting Y to delta or delta to Y imnedances for. 60 equivalent system impedance for, 57-63 estimating curves and tables for, 7690 examples of, 66-76 formulas for, 38-39 lor fuse application, 31, 32-33, 35-36 impedance diagram for, 41-42, 50-51 for interrupting duty, 29-31, 3233, 35 lor line-to-ground short circuit, 374-377 estimatine tables for. 362 on reactance-grounded systems, 376-377

INDEX

Short-circuit current, calculation of, for line-to-ground short circuit, on resistance-grounded systems, 375-376 on solidly svstems. 377 . arounded .. by symmetrieal-eomponent analysis, 131-134 for momentary duty, 29, 32-33, 35 motors considered in, 46-50 converting horsepower t o k v a for, 39 reactances for, 32-33, 46-50 multiplying factors for, 20-22, 46, 63 table of, 32-33 ohms for, choice of, 51-52 conversion t o common base of, 57 per cent for, choice of, 51-52 per-unit for, choice of, 51-52 reactances t o use for, 32-33, 14-46 condensrd tithle of, 32-33 in resistance-welder circuits, 69,

75 resistances t o use for, 45-46 in secondary network systems, 4445, 68-69 short-circuit location for, 43-44 in single-phase systems, 69, 75 summary of, 38-39 by symmetrical-component anslysis. for double line-to-mound short c i r u i t , 134 for line-to-ground short circuit, 131-134 for line-to-line short circuit, 127-130 for three-phase short rircuit, 1 2 6 127 in svstems above 600 volts., 44., 49., 69 for unbalanced circuit conditions, 42, 127-131 utility systems in, representation of, 57 decrement of, explanation of, 18-19 relay coordination of, effect on, 546, 561-567 X / R ratio, effect of, 19 d-e component of, 15-18 decrement of, 18 cffccts on ovcreurrent protective devices of, 507, 548, 570 explanation of, 15-18 initial magnitude of, 18 d k time constant of, 19-20 effects of, 1-3 estimating curves and tables for, for feeder circuits, husway, 76-77, 60-81 ~

965

Short-circuit current, estimating curve8 and tables for, for fccder circuits, of different construction in series, 83 parallel, 81 single-conductor cable, 83-84 single-phase, 64 three-conductor cable, 76-77, 8&81, 87-90 for unit substation, single-phase (120/240 volts), 82 three-phase (600 volts and less), 76, 78-79 fundamcntals of, 1-24 limiting effect of cable on, 87-90 line-to-ground, estimating tables for, 302 return path for, 405, 414-415 on ungrounded systems, 340-315 multiplying factors far, calculation of, 32-33 explanation of, 2&22 on portahle-machinery circuits, 427429 r m v~ a of,~20 ~ ~ SoIIrcBs of, 3-4 symmetrical, cxplanation of, 10-12 total valuc of, cnplanatian of, 22-24 from unit substations, table of, 7&82 X I R ratio in relation to, 19 Short-circuit-current protection, adequate, haw t o he sure of, 344-346 inadequate, cnsmple of, 146-147 (See also Overcurrent protection) Short-circiiit-currcnt proteetivc devices, adequate, how to be m r e of, 144145 basic types of, 148 cable se!eetion influcnccd hy, 185-186 circuit hreakers (see Circuit hreakers) fuses 5s (see Fusrs: Switches, fused) on portablc-machinery circuits, 429433 for primary power systems, 170-160 rating, basis of, 24-36 relays as (see Relays) rerpircments of, bnsie, 147-148 for secondary power systems, 154170, 170, 686-688 selection of t y p r of, 147-153 switches as, fused (see Switches, fuscd) (See also Overcurrent protection; Short-circuit-current rating) Short-eircuit-current rating, basis of, for circuit breakers, low-voltage (600 volts and below), 35 power (above 600 volts), 24-31

966

INDEX

Shart-circuit-current rating, hasis of, for fuses, low-voltage (600 volts and below), 36 power (above 600 volts), 31, 34-35 for motor starters, high-voltage, 35 low-voltage, 36 of fused switches, low-voltage, 167 of fuses, low-voltage, 167 interrupting, explanation of, 27, 35 of large air circuit breakers, 154 of molded-case circuit breakers, 166 momentary, explanation of, 27, 35 of panelboards, 692 of power circuit breakers, 26 Short-circuit ratio of generators, definition of, 22 Short-circuit study (see Short-circuit current, calculation of) Shovels, electric (see Portable equipment) Solid grounding, calculation of groundfault current for, 131-134, 377 definition of, 339 fault duty increased by, 396-399 table for estimating ground-fault current for, 362 use of, 352-355 (See also Grounding, neutral) Speed governing (see Governing of steam turbine) Standards (see Codes and standards) Star bus in primary power Rystems, 717 State line (Mollier chart), definition of, 830 Station-type switcbgear, in primary power systems, 708, 710 reduced rating of, a t 6.9 kv, 616 selection of, for 6.9-kv system, 616-617 for 13.8-kv system, 616-617 (See also Circuit breakers, powerr Steam and power generation in industrial plants, costs of, 874-883 (See also Costs) diesel plants for, 873 gas turbines for, 857, 861-866 power-plant cycles for, steam, 818, 820-824 purchased power vs., 813 steam conditions for, 835-843 (See also Steam pressures and temperatures) steam plants far, efficiency of, 832-834 steam turbines for, 820-824 performance of, 857 selection of, 844-849 steam vs. gas turbines for, 872-873 trends in. 808 (See also Generation of electric power; Turbines)

Steam pressures and tempershres, effect of, on by-product power, 810 a t exhaust and extraction of steam turbines, 837-843 on plant fuel consumption, effect of, 836-837 on power-plant efficiency, effect of, 835 selection of, 835-843 trends in, 810, 820, 835 turbine exhaust, 842 Steam rates, actual, condensing turbine, 859 extraction-turbine, 860 noncondensing-turbine, 858 theoretical, table of, condensed, 826 examples explaining use of, 831 Steam turbine (see Turbines, steam) Steam turbine-generators, cost-estimating data far, 909 reactance of, 92-94 (See also Generators) Steam-gas turbine plant (combined cycles), 868-869 Straight bus in primary power systems, 717 Substations, primary, arrangements of, 712-715 preferred transformer connections for, 711-712 protection of, against overvortages caused by lightning, 324-326 replacement of, 764-767 (See also Unit substations) Subtransient reactance, definition of, 9 Surge impedance, of cable, 301 definition of, 301 of overhead lines, 301 Switches, for capacitors, selection of, 485-487 disconnecting, reactance of, 110 fused, VS. circuit breakers, 153 interrupter above 600 volts, 176 low-voltage (600 volts and below), application of, 167-170 ratings of, avrtilable, 166-167 standards for, 170 tvoes of. 166 oil-&ed cutout type of, 174-176 (See also. Pnaea\ . . . . interruptr ir (above 600 volts), air-type, """ 176, boy classes of, 668 description of, 669 double-throw type of, 657-661 vs. single-throw, 657458 vs. two single-throw, 659-661 interlocking of, 631, 671 liquid-filled, 669 ~

INDEX

P67

Switches, interrupter (above 600 volts), in load-center power distribution systems, 6 3 M 3 1 oil-cutout type of, 176, 670-671 overcurrent protection affected by,

Symmetrical-component analysis, measurement of sequence quantities used in. 142-143 negative-sequence component in, definition of, 116-117 630-631 phase shift affecting, transformer, use of, on load-center unit substa135, 137-138 tions, 63&631 positive-sequence component in, manual, in loop feeders. 661464, 721definition of, 116-117 799 properties of symmetrical quantities oil-fused cutout type of, 174-176 used in, 119-121 Switcbgear, BIL (basic impulse insulation sequence systems used in, indepundlevel) of, 304 ence of, 121-122 device numbers and functions of, 930of short circuit, double line-to-ground, 932 134 installation coats of, 893 line-to-ground, 131-134 low-voltage, cost-estimating data for, line-to-line, 127-130 915 three-nhase. 126-127 drawout-type, where applicable, transformer zero-sequence circuits 155155 used in, 135-140 for unit substations, 672-674 zero-sequence component in, definimetal-clad, cost-estimating data for, tion of, 116-117 914 Symmetrical short-circuit current, exin primary systems, 708-711 planation of, I S 1 2 protection of, against overvoltages (See also Short-circuit current) caused by lightning, 321-324 Synchronizing bus, for modernization open-type, replacement of, 763-764 and expansion of power systems, station-type, in primary power sys776, 790, 794, 799 tems, 708, 710 in primary power systems, 615, 717, reduced rating of, at 6.9 kv, 616 725-730 selection of, for 6.S-kv system, 616Synchronous condensers, reactance of, 91 617 Synchronous converters, reactance of, 91 for 13.8kv system, 616-617 Synchronous motors (see Motors, Syn(See also Circuit breakers) chronous) Switching surges a s cause of overvoltage, Synchronous reactance, definition of, 9 289-294 (See olso Reactance) Symbols, of electrical equipment for one- System grounding (see Grounding, line diagrams, 932-942 system) letter, for electrical terms, 928-929 System neutral, definition of, 339 for beat and thermodynamic terms, how to obtain, for grounding, 348 929-930 selection of, for grounding, 367-370 Symmetrical-component analysis, (See also Grounding, neutral) “ a ” operator used in, function of, System neutral grounding (see Graund11~119 ing, neutral) basic components used in, definition System voltage (see Voltage, system) of, 115-117 physical concept of, 122-124 Tap changing under load for regulation basic system application of, 125-126 of voltage, 221, 814 characteristiesof.. imoortant. 1 1 6 1 17 Tests, dielectric, of insulation, 302-305 . example of, in composite system, Theoretical steam rates, table of, 826 la142 Three-point method of measuring refundamental equations for, derivasistance to earth, 423424 tion of, 119-121 Throwover schemes, for automatic- or table of, 123 manual-transfer secondary selective impedances for, rotating machine, load-center svstems. 643-646 134-135 for primary selective load-center syatransformer, 135-140 terns, 655-6131 introduction to, 114-116 Time constant, d-c, dehition of, 19-20

.--

~~

~~

INDEX

968

Topping turbines, noncondensing steam turbines used as, 846, 849, 854 Transformers, askarel-type, 671-672 BIL (basic impulse insulation level) of, 303-304 bus-tie, far expansion of power systems, 774-776 connections of, in main substations, 711-712 cost estimating data for, 907, 908 distribution, impedance of, 96-97 dry-type vs. liquid-filled, 671-672 failure rates of, 630 grounding, definition of, 340 for 480-volt light and power systems, 749-750 on portable-machinery circuits, 434 rating of, 377-378 selection of, 385-388 specification of, 385 Y-delta, 350 zigzag, 349-350 impedance, reactance, and resistance of, 96-97 installation costs of, 893 isolating, on portable-machinery eircuits, 434 lighting, 593-595 for loitd-center unit substations, 671672 loading of, 638-640 minimum rating of, for grounding, 366 ail-filled type, 671-672 for 120-volt power on secondary power systems, 593-595, 692, 696-697 open-ventilated dry-type, 671-672 parallel operation of, 646, 64&649 phase shift in, effect of, 135, 137-138 on relay coordination, 568 power, impedance of, 97 protection of, against overvoltages caused by lightning, 319-321 as required by Kational Electrical Code, 632 sealed dry-type, 671-072 small dry-type, for factory combined light and power systems, 593-595 as sources of short-circuit current, 8 symmetrical-eomponent impedances of, 137-140 taps on, selection of, 227-229 with tertiary windings in main substations, i l l three-winding, equivalent circuit for,

.

66 in main substations, 711-712

Transformers, voltagc drop in, calculation of, due to motor starting, 280-261 for steady-state conditions, 235-237 voltage ratings of, 194-190 method of exprcssing, 196-197 (See also Autotransformers) Transient overvoltages (see Ovcrvoltages) Transient reactance, definition of, 9 (See also Rcactsnce) Transmission circuits, impedance, resetancc, and resistance of, 105-109 spacing of lines for, 109 in system expansion, US^ of, 800-807 voltage selection for, 619-621 Transmission voltage, selection of, 61962 I (See also Voltage, system, selection of) Traveling wavcs on transmission lines produced by lightning, 301-302 Trigonomctrie functions, tables of, 944948 Trips, direct-acting (see Direct-acting trios) Tuned 'grounded (see Ground-fault neutralizer) Turbine-generators, cost-estimating data for, 909 reactance of, 92-94 (See also Generator) Turbines, gas, ambient temperature on, effect of, 862 applications for, 866-871 by-product generation with, 819 combined s t e a m g a s cycles using, 868-869 cooling-water requirements of, 861 cycles for, description of, 864 description of, 861 efficiency of, 861-862 exes8 air in, requircments for, 869 exhaust-heat recovery from, for boilers (see Boilers) for drying manufactured products, 87&871 for feedwater heating, 862, 869 for steam generation, 820, 864, 866-869 fuels for, 871 starting characteristics of, 866 vs. steam turbines, 872-873 t o top steam-turbine plants, 86% 869 steam, automatic-admission, 848, 855 automatic-extraction, 846-848, 855857 division of load and flow between sections of, 855-856

969

INDEX

Turbines, steam, sutomatic-extraetion, performance of, 857 efficiency of, 820-822 expansion lines for, 830 "8. gas turbines, 872-873 governing of (see Governing of Steam turbine) industrial, partial list of large, 810 nonextraction, 831, 844, 846 performance of, 857-859 power-plant cyclcs for, 82&823 ratings of, 855 average, in industrial plants, 810 standard kw, 855 selection of, 844-849 size of, physical, 854-855 stcam conditions of, partial list of high initial, 810 steam rates of (see Steam rates) topping, noncondensing steam turbines used as, 846, 849, 854 type of, automatic-extraetion-admission, 848

automatic-ertraction-eondensing, 846 automatic-extraction-nonconodensing, 848 condensing, 844 noneondensing, 846 topping, noncondensing steam turbines used as, 846, 849, 854 Unbalanced currents, balancing reactors for, 784-788 Ungrounded system, characteristics of, 340-345 definition of, 339 failures related to, 337-338, 401-402 ground detectors for, 344 locating isults on, 344-345 overvoltages on, damaging, 278-299, 34b343 single line-to-ground fault on, effect of, 340-345 (See also Grounding, system) Unit substations, installation costs of, 893 load-eenter, arrangement of, selection of, 666-667 centralized motor control used with, 674-675 cost-estimating data for, 912 for grounded-neutral systems, 36& 361 incoming-line section of, 6 6 7 4 7 1 interrupter switches for, 668-671 kva rating for, scleetion of, 664-666 location of, 6 7 7 4 7 8

Unit substations, load-eenter, main secondary breakers in, use of, 674-676 secondary switchgear for, 672-671 secondary switching section for, 672-676 selection and application of, 664-678 short-circuit current from, estimating tables for, 76, 78-79, 82 transformers for, 671-672 master, cost-estimating data for, 910, 911 primary switching in, cost-estimating data for, 913 use of. 681-683. 815 Unit 'system, generator-transformer, grounding of, 395-396 voltage rating of, 198 use of, 730 Utility system, cleetric, grounding of, 366-367, 395 industrial ties with, 815 Utilization equipment, grounding of, 417-419 on system voltage selection, effect of, 575 voltage ratings of (see Voltage ratings) voltage spread on, effect of (see Voltage spread) I

Voltage, improvement of, by capacitors, 471474 regulation of, on feeders, 224-225 by load-ratio control, 221, 814 a t main power sourre, 22LL222 when paralleling utility and industrial generation, 817 by t a p changing under load, 221 system, nominal, 191, 199 selection of, codes and standsrds on, effect of, 576 for commercial-type buildings (see Commercial-type buildings) factors affecting, 574, 576 above 15 kv, 619-621 for furnaces, electric, 575, 580, 595-597 far large mills, 619-621 for lighting in factories, 582-587 (See also Factories) for lighting in factory offices, 592 for modernization or expansion of power system (see Modernization and expansion of power systems) for motors, as function of horsepower, 600-610 for motors on primary system, 600-603

970

INDEX

Voltage, system, selection of, for ovens, electric, 575, 580, 595-597 for plants, large, 613-621 for primary power systems, 597619 safety an, effect of, 575-576 for secondary power systems, 576597 for serving different voltage loads at 600 volts or Less, 580-581 of single-phase vs. three-phase, 57S580 switeheertr. effect of. 612-617 for transmission circuits above 15 kv, 619-621 utilization devices, effect of, 575 of 208 Y/120 volts where applicahk, 579 of 480 volts vs. 240 volts. 57&579 of 480 volts vs. 600 volts; 577-578 of 600 volts and less, 576597 of 2400 volts, where most economical, 600, 608410 of 2400 volts to 15,oM) volts, 597-619 of 4160 volts, where most applicahle, 603-604, 60&610 of 4160 volts vs. 2400 volts, 600607 of 13,800 volts vs. 4160 volts. 599 of 13,800 volts vs. 6900 volts, 607-619 of 69 kv vs. 34.5 kv, 620-621 Voltage drop, due to motor starting, calculation of, 248-277 in cables, 261-263 in distribution systems, 260-265 examples of, 273-277 formulas for, 266272 in generators, 253-260 in overhead lines, 261-263 in power systems, 264-265 in reactors, 262 in transformers, 260-261 combining two or more, 265-266 description of, 251-253 reduced frequency on, effect of, 272 series capacitors, effect of, 264 starting-method charactcristics that affect. 249-251 in reactors, in secondary power systems, 689, 691 with synchronizing bus, 728-730 reduced by load-center power distribution system, 628 in secondary power systems, 693 steady-state, calculation of, 232-234 example of system, 244-248 ~I

Voltage drop, steady-state, calculation of, formulas for, 232-234 charts of, for husway, 242-244 for cable, 237-241 for transformers, 235-237 definition of, 201-202 effect of, 202-204 Voltage flicker, in combined light and power systems in factories, 500 limits of, 215-216 reduction of, by separate circuits, 226 by series capacitor, 227 summary of methods for, 230-231 by system design, 590 by voltage stabilizer, 226-227 Voltage identification, basic pattern of, 194 Voltage ratings, basic structure of, 1 9 s 194 of capacitors, 199, 499-500 of eenorators. 194. 197-198 of Tamps, 194, 190 of lightning arresters, 307-312 of motors, 194, 198 of systems, 194, 199-200 of transformers, 194-196 Voltage regulators, feeder, 224-225 generator, advantages of, in motor starting, 256-257 induction, 222-225 basic impulse insulation level of, 303-304 reactance of, 112 stop-type, 222-224 basic impulse insulation level of, 303-304 Voltage spread, definition of, 200 effect of, on capacitors, 214 on electronic equipment, 21 1-213 on fluorescent lamps, 208-210 on incandescent lamps, 207-208 an induction motors, 205-206 on infrared heating devices, 211 an mercury lamps, 210 an resistance heating devices, 210211 an solenoid-operated devices, 213214 on synchronous motors, 206207 recommended limits of, 214-215 reduction of, by generator voltage regulators, 226 by load-center distribution system, 217-218 by lower impedance circuits, 220 methods for, summary of, 230-231 by regulating equipment, 220-225 by shunt capacitors, 225 ~

INDEX

Voltage stabilizers, use of, 226-227 Voltngc variation (see Voltagc spread) Voltage zone, definition of, ZO(t201

97 1

X / R ratio, explanation of, 19

Y-connected autotransformers in primary substations, 711 Y-delta grounding transformer, 350 Y-Y transformer connections in primary substations, 71 1

short-eireiiit-eiirrent magnitude affected hy, 83,85, 86 X o / K oratio, influenre on transient ovcrvoltagPs of, 387-388 X , / X , ratio, overvoltage influenced hy, 280, 339, 3 8 6 3 8 8

Zero-phase-sequence component, definition of, 116-117 (See also Symmetrical-component analysis) Zigzag grounding transformer, 349-350

Contents

vii

Preface

xi

Introdnction 1. Short-circuit-current Calculating Procedures by Donald Beeman, Alan Graeme Darling, and R . H. Kaufmann

1

2. Symmetrical Components as Applied to Short-circuit-current Calculation on Three-phase Systems by R. H . Kaufmann

I11

3. Selection of A-C Short-circuit Protective Devices and Circuit Equipment by Donald Beeman and R. H . Kaufmann

144

4. Voltage-Standard Ratings, Allowable Variations, Reduction

of Variations, Calculation of Drops by W . R . Criles and Mavnord N . Halberg

101

5. System Overvoltages-Causes and Protective Measures by R. H . Kaufmann and Maynard N . Halberg

278

6. System Grounding by L. J . Carpenter and L. G . Levoy, J r .

337

7. Equipment Grounding by L. J . Carpenter, Shelby C. Coolce, J r . , R H . Kaufmann, and David Stoelzel

404

8. Power-factor Improvement by W . C . Bloomquist

430

9. System Overcurrent Protection

504

by Francis P. Brightman 10. Selection of System Voltages by Donald Reeman

574

11. Load-center Power Systems and Circuit Arrangements by Donald Beeman

622

ix

CONTENTS

x

12. Secondary Distribution Systems by Shelby C . Cooke, J r .

684

13. Primary Distribution Systems by Norman L . Hadley 14. Power Systems for Commercial Buildings by Donald S . Brereton 15. Modernization and Expansion of Existing Power Systems by Donald Beeman and L. G . Leuoy, J r .

706

113. Steam and Power Generation by William B . Wilson

808

17. Load- and Cost-estimating Data by Donald S. Brereeton

881

Appendix compiled b y D. B . Armstrong

925

Index

949

731 757

McGRAW- HILL HANDBOOKS

ARBOPTAND SXITH . National Electrical Code Handbook, 8th ed. AMERICAN SOCIETY OF MECKANICAL ENQINEERS ASME Handbook: Metals Engineering-Design ASME Handbook: Metals Properties AXERICAN SOCIETY OF TOOL ENQINEERS . Die Design Handbook AMERICAN SOCIETY OF TOOL ENGINEERS ’ Tool Engineers Handbook BEEMAN. Industrial Power Systems Handbook BERRY, BOLLAY, AND BEERS. Handbook of Meteorology BRADY. Materials Handbook, 7th ed. COXPRESSED AIR AND GAS INSTITUTE ’ Compressed Air Handbook, 2d ed. CROCKER . piping Handbook, 4th ed. CROFT. American Electricians’ Handbook, 7th ed. DAVIS . Handbook of Applied Hydraulics, 2d ed. H ENNEY . Radio Engineering Handbook, 4th ed. JOHNSON AND AWTH . Fuel8 and Combustion Handbook JWRAN . Quality-control Handbook KETCKWX . Structural Engineers’ Handbook, 3d ed. KINQ. Handbook of Hydraulics, 4th ed. KNOWLTON . Standard Handbook for Electrical Engineers, 8th ed. KWRTZ . The Lineman’s Handbook, 3d ed. LABRERTON AND MARKS . Marine Engineers’ Handbook LE G RAND ’ The New American Machinist’s Handbook L1nnm.L . Handbook of Nonferrous Metallurgy, 2 vols., 2d ed. MARKS . Mechanical Engineers’ Handbook, 5th ed. O’ROURKE . General Engineering Handbook, 2d ed. PACIFIC COAST GAS ASSOCIATION ’ Gas Engineers’ Handbook P ERRY . Chemical Business Handbook P ERRY . Chemical Engineers’ Handbook, 3d ed. STANIAR . Plant Engineering Handbook TERMAN ’ Radio Engineers’ Handbook URQUEART ’ Civil Engineering Handbook, 3d ed.

Authors

,411 :iutliors of material in the

Industrial Polx-er Systems Handbook

:Ire in the Industrial Engineering Section of the General Electric Compnn?-, Srhenedady, Kew York.

D O N ~ LH DEEM-AN Fellow-AIEE ;* Registered Professional Engineer (Kew York and California) W. C. BLOOMQUIST Fello\-.4IEE; Registered Professional Engineer (Sew York) DOXALD S. H RERETOK . Associate Member-AIEE FRAKCIS P. RRICHTHN . Associate Member-AIEE; Registered Professioiial Engineer (Sew York) I,. J. C.\RPENTEK . Member-AIEE S HELBY C. COOICE, J R. Member-AIEE; Registered ProfessionaI Engineer (Pennsylvania) W. R. CRITES Associate Member-AIEE ALAK GRAEME D.AI~LING (retired) Member-AIEE; Registered Professional Engineer (New York) XORMAN I,. HADLEY . Associate Member-AIEE MAYXORD S . HALRERG . Fellon-AIEE; Registered Professional Engineer (Kew York) R . 13. KATJFM.ANN . Fellow-MEE; Registered Professional Engineer (Kew York) L. G. LEVOU,JR. Member-AIEE; Registered Professional Engineer (Kern York) D AVID STOETZEL, JR. . Member-AIEE: Registered Professional Engineer (New York) WILLI.A~I B. WILSOK Member-AIEE; Member-ASMEt; Registered Professional Engiiieer (Massachusetts) '

'

Ameriran Institute of Electrical Engineers. Society of hlcchsnieal Engineers.

t American

T H I S BOOK I S DEDICATED TO

F. MOHLEYROBERTS AND LEONIDA.

UYAN8KY

WITHOUT WHOSE UNDER8TANDINQ GUIDANCE, ENCOUIUQEYENT, AND COUNSEL IT COULD NOT HATE B E E N WRITTEN

Preface

The engineering knowledge required to maintain and operate industrial power generation, transmission, and distribution systems has increased manyfold since electricity was first used in industrial plants t o light a few lamps and run a line shaft motor. The exacting power requirements of modern production equipment necessitate the introduction of refinements in power system design as to voltage spreads, service continuity, flexibility, etc. Since the introduction of the load-center power distrihution system, industrial power distribution practices have been almost completely revolutionized. This book outlines up-to-date industrial power system engineering practices whirh should be especially valuable to industrial plant engineers and electricians, utility power salesmen, consulting engineers, contractors, industrial power application engineers, architects, and others who have anything t o do with the planning of elertrical facilities for industrial plants or rommercial buildings. College professors and students should find use for i t vhere it is desired t o teach and study the principles of industrial power system engineering. There will be uses in other organizations as text material for discussion groups and classes involved in industrial power system engineering. Although the problems involved in electric utility generation, transmission, and distribution systems are different in many ways from those in industrial power systems, the utility engineer will find 'many of the data in this hook useful to him, particularly when designing power services for industrial plants. Many of the application data have heen presented in condensed tabular and curve form t o make practical solution of many problems easy and fast. Kumerous examples are included illustrating the nse of the principles and data outlined in this hook. The general order of chapters is that the technical information is placed in the first part of the hook and the economic or system design information in the latter part of the book. All the information on power and steam generation is placed in Chapter 16. The problems associated with system voltages are divided into three categories. Those discussed in Chapter 4 include the designation of system voltages, equipment ratings, voltage spreads, calculations of voltage drops, etc. Chapter 5 "ii

”iii

PREFACE

deals with transient overvoltages. I n Chapter 10 the economic aspects of the selection of system voltage are considered. The cost-estimating data given in Chapter 17 will naturally apply only at the date of publication of the hook. They will, however, be a valuable referenre io the years following, provided that they are adjusted according t o the industrial price index. Even though the actual costs change over a period of years, one can still use the cost data in this book for making relative rost st,udies t o compare one type of system with another. Most of the equipment for modern industrial power systems is available from several manufacturers. Because of the authors’ familiarity with, and ready acress to information on, General Electric products, they have been used as examples i n this hook. In most cases the discussion of equipment applies equally well t o similar equipment available from other manufacturers. I n writing t,his hook, it has been assumed that the reader will have a t least a working knowledge of electrical engineering principles and is familiar with the terms commonly used in this field. Commonly used symbols and abbreviations, etc., are included in the Appendix. Although this hook Ivas written by the various authors whose names appear in this book, many others contributed in various ways in its preparation. Because of space limitations it was possible t o make specific acknowledgment of such assistance in only a few cases. We wish t o express our appreciation t o all those others whose contributions in the field of industrial poirer system engineering have helped t o make this book possihle. Much of the information in this book has been previously published by the authors in various trade magazines and engineering society journals. We nish t o express our appreciation, particularly t o Industry and Power, Electrical Construction and Maintenance, Factory Management and Maintenance, Elrctrical Engineering, 4 1 E E Transactions, Electrical World, General Electric Reuiew, Power, Plant Engineering, The Plant, The Consulting Engineer, and all other magazines in which parts of this information have appeared for permission t o use the information prepared for, and originally published in, them. We also wish t o thank John Wiley & Sons, Inc., for permission t o reprint in Chapter 8 data from the hook “Capacitors for Industry” by W. C. Bloomquist, C. R. Craig, H . M. Partingtoti, and R. C. Wilson. T H E AUTHORS

Introduction When electric power stops flowing in a manufact,uring plant, the plant's production stops. The plaut owners may have invested money for the best of plant facilities, the most modern production machines, have an ample inventory of raw material, good product design, highly traiued and efficient labor, and everything else that is required to produce manufactured goods quickly and a t low cost. However, if electric power is not available when and where needed in the plant, then the owner's investment in both plant and inventory is so much idle capital. Because electric power is so important in any manufacturing process, the electric power distribution syst,em in the plant is the vital link that carries the electric power from the utility supply point to the production machine that keeps the plant's productioir moving. The electric power dist,ribution system in a plant generally costs less than 2 to 5 per cent of the tot,al plant cost, including process machinery. Since one gets so much for so little from t,he investment in the plant power system; it pays to use only the best pract.ices and equipment t o assure a maximum of profit from the over-all plant investment. The power system engineer who designs an industrial plant power distribution system must rely on his technical engineering background, and also he must, exercise considerable engineering judgment, as all phases of industrial power system enk'neering rannot be expressed i n numbers or solved by formulas. When decisioiis based on engineering judgment are required, the power system engineer should always base his consideration on the best facts availahle and not on vague hypotheses or information which does not apply t o the specific problems a t hand. One of his most valuable tools is test data made under specific controlled conditions. Field reports and operating experiences must always be weighed very carefully to be sure that all the facts and all the background are available. For example, a short-circuit test on a circuit breaker made in the laboratory under controlled conditions gives accurate knowledge of the performance of the circuit breaker because t.he test conditions can be measured. Just because a circuit breaker opened and cleared, a short circuit in a field application may or may not have been a measure of the ability of the circuit breaker t o perform under all conditions existing Xi

xii

INTRODUCTION

in that system. All conditions of a tield short circuit are generally not known, as measuring instruments are not usually continuously connected t o the circuit t o tell what the circuit current conditions were when the circuit breaker opened. Unfortunately, there is a tendency t o construe such field operation as proof that a circuit breaker is satisfactory, whereas the facts obtained from tests made under controlled conditions may show that it may be inadequate under many conditions in the particular system involved. Field data are part of the engineer’s warehouse of facts, and certain kinds of field data are extremely valuable. The engineer must be sure that he knows the conditions under which the field data are obtained and that the data are broad enough t o mean something. Referring again to the short-circuit protection picture, the operating experience in any one plant with inadequate protective equipment may have been satisfactory over a period of years. Generalizing on that one experience only or even a few more like it may lead t o the belief that inadequate circuit breakers mill never fail. More complete data may prove that inadequate equipment has failed in literally hundreds of other cases not known t o a particular individual. Therefore, any field data that are used should be extensive and represent a cross section of industry and not just a few isolated cases. To do this job effectively, the engineer must be constantly searching for facts on which t o base his decisions. Then if power system engineering is based on facts-the scientific method-many electric power system shortcomings will be overcome. OVER-ALL PLANNING

To obtain a power distribution system which is adequate to meet the service reliability requirements of a plant and yet which is lowest in cost requires that the power system engineer plan the power distribution system on an over-all inrlusive basis. While the electrical system is installed in parts, such as substations, cable, bus, switchgear, transformers, etc., the system nevertheless functions as a complete integral unit. The best way to get an over-all picture of the probable performance of an industrial power distribution system is to make a one-line diagram. This, on one small piece of paper, shows the system in its entirety. The one-line diagram gives the power system designer a reasonable idea of what service reliability will be and how the components of the system will tit together electrically t o serve the needs of the plant most effectively and economically. There are many factors which must be considered in the over-all planning of a power system. Some of the more important aspects to be considered are discussed in the following.

INTRODUCTION

xiii

SAFETY

One cannot put too much emphasis on safety. Electricity is becoming more and more an integral part of our production lines. I t is used in areas where all kinds of people work. The power system engineer is responsible for making the systems as safe as possible. Building a substation structure with a fence around it labeled "For Authorized Persons Only" is no safety guarantee. Authorized personnel make mistakes too. Three fundamental ideas help to build safe power systems: 1. Enclose all live conductors in grounded metal. 2. Use only adequate circuit-protective equipment. 3. Design the system so that working on energized conductors is not necessary. Of course, there are other important safety points. But these are three which the system design engineer can influence greatly. ECONOMICS

Economics is a very important part of power system engineering. The engineer must compare systems on the basis of cost as well as other features. I n making cost comparisons, i t is important t o include all parts of the system from the power source down to and including the utilization equipment. Cost comparisons should also include installation as well as eqnipment costs. To illustrate: System A may require less transformer kva than system B . But the conneetions of system A and other factors may require far more expensive switchgear and more cable than system B . The increased cost of switchgear and cable may more than offset savings in transformers. A comparison on the basis of transformers only would give the wrong answer. Aft,er a system has been selected, the equipment is chosen. Equipment Type I may be so constructed that its selling price is lower than the price of equipment Type 11,which is more completely assembled a t the factory. On the other hand, the increased field installation expense of Type I equipment may more than offset the difference in price. Hence, both price and installation cost must be Considered t o obtain the lowest over-all investment. Always consider the cost of the completely installed system, not just the prices of its components. PLAN FOR LOAD GROWTH

One of the greatest mistakes made in power system engineering is planning without allowance for future expansion. Failure t o plan for

xi"

INTROOUCTION

the future results in extravagance, inflexibility, and complication. The use of electricity in most plants increases. Even if new floor area is not added, more intensive manufacture within a given area increases the amount of electricity 3 to 5 per cent a year. Our economy is still expanding, and as long as that takes place we are going to have to manufacture more goods. The manufacture of more goods requires expansion of existing facilities as well as new facilities. It costs so little to plan and arrange for future load growth that it should be a must for every plant. Some important items in this respect are t o allow some reserve in substation kva size over predicted present loads, to buy main switchgear with a margin in interrupting rating to allow for increasing power supply to the plant, and last and perhaps most import,ant, t o select the proper voltage, i.e., one high enough to permit growth without bottlenecks. SIMPLICITY

Make systems simple. Many total shut,downs have resulted from complicated systems. Industrial operators do not get emergency switching practice every day. If they have complicated systems, experience proves they are apt to make mistakes in an emergency. These errors usually cause greater outages than they would with simple systems. Simple systems are easy t o understand and easy to operate during both normal and emergency conditions. The growing difficulty of obtaiuing adequately trained personnel makcs a simple power system desirable. FLEXIBILITY

Plants change manufacturing proceases from time to time. Where castings are used today, welding may be used tomorrow. Process and product alike change as demands and styles change. Any plant, and particularly a metal-fabricating plant, should be designed with the idea that complete new process layouts can be made without requiring major changes in the electric distribution system. Two great contributions toward flexibility are (1) the load-center system with small substations which may be added in small units as required and, if necessary, moved, and (2) plug-in busway, which permits the installation of permanent power distribution systems to which machine tools and other devices are merely plugged in where neceseary. SERVICE FZUABILITY Service reliability in any plant is important. Most manufacturing proeesaes are on a production-line basis, so that a shutdown of any part

INTRODUCTION

X"

of the line may hold up an entire plant. Some processes in themselves require a very high order of service reliability. High service reliability generally can be obtained in two ways. One is by providing duplicate channels to route power to any load; the other is by installing only the highest-grade electric equipment available, using the best installation methods. It might. be added that service reliability can be increased through simplicity of system design, a fact that was emphasized previously. MODERNIZATION AND EXPANSION

When plant facilities have to be expanded or modernized, the engineer is afforded an opportunity to design the ideal electrical system. First a one-line diagram should be made of the system he would have if the plant were starting new. Having made this plan, he should let it guide the future modernization and expansion. Existing equipment obviously cannot all be retired a t once. But, as additions and replacements of equipment are made, they should he on the basis of being integrated into the ideal plan and not merely as replacements of equipment in the old system. One of the most significant parts of the new program is selecting the proper voltage-not just riding along and trying to expand at existing lorn voltage. SELECTION

OF EQUIPMENT

The main fundamental in selecting equipment is to choose the best available. It costs no more in the end and pays dividends in service continuity and lower maintenance. Some widely accepted principles are: 1. Use only metal-clad equipment a t 15 kv and below. 2. Choose noninflammable transformers. 3. Use factory-assembled equipment for easier field installation and better coordination. 4. Be sure equipment ratings are adequate in every respect, i.e., voltage, current, int.errupting rating, etc. Much of the advantage of the high-quality equipment can he lost if it is carelessly installed and maintained. MAINTENANCE

Maintenance is largely in the hands of operators. However, the system designer can aid in this problem by designing systems that provide nlternate power channels, to permit one to be taken out for maintenance

i"i

INTRODUCTION

without dropping essential loads. Use drawout equipment wherever possible t o enable maintenance on circuit-breaker elements t o be done in a clean service shop. A spare element should be specified t o replace the one being maintained. MAKE A RELAY COORDINATION STUDY

The operator should check all equipment regularly. The system design engineer can help by including suggested maintenance schedules in the data given t o the operator. Relay settings can be selected by the design engineer to permit t,he operator t o set them properly. Most important of all, operation and settings of relays should be regularly checked. THINGS TO CHECK WHEN DESIGNING THE POWER SYSTEM

The discussion of items in which proper judgment should be used t o obtain a power system best suited to a given plant could be expanded. However, the foregoing points illustrate the basic thinking used by today's power system design engineers. As a guide, here are things t.o check when designing a power system for an industrial plant or building. 1. What is the nature and magnitude o,f the load? Until the process and load requirements are known, one cannot intelligently lay out a power system to meet the load requirements. Use data from existing plaut,s. See Chapter 17. 2. Where i s the power corning from? See the local utility early in the planning stages t o see what problems there are in power supply. Are there savings possible from by-product power gerieration because of by-product fuel or use of large quantities of process steam? See Chapter 16. These questions must be answered early i n the study as well as other details of what voltage is available and where is the nearest adequate utility line located in relation t o the plant site. 3 . How much will the electric power system cost? Estimates are usually required in the early stages. Be sure t o make your estimates realistic and not so low that you.later find you have t,o cut corners t o stay within the appropriations. See Chapter 17. 4. What vollage levels should be selected for the plant primary system and low-voltage system? Consider load growth, economics, flexibility, etc. Usually 13.8 or 4.16 kv is the most advantageous primary voltage and 480 volts the most all-around satisfactory for low voltage. See Chapter 10. 5 . What circuit arrangement is best suited, i.e., radial, secondary selective,

INTRODUCTION

x4i

07 secondaq network? Consider service reliability as dictated by process, provision for maintenance and possible load growth, flexibility required for process changes? etc. See Chapter 11. 6. W h a t size subsubstations are most economical? General studies give the most economical range at 500 to 1500 kva at 480 volts secondary, but in a specific plant load density, load locations, primary voltage, and many other facts affect the choice. See Chapter 11. 7. Secondary distribution. What type of feeders, cable or bus? How big should feeders be? How much load per feeder? Should plug-in busway be used or are panel hoards required? How can adequate shortcircuit protection be secured in the branch circuit protective equipment? See Chapter 12. 8. Combined Zight and power svslems. IJsually combined light and power systems are more economical, but the flicker problem must he watched, particularly if welders are present. How big should the 480120-volt transformers be? Should they be single or three phase? If fluorescent lighting is heing used, can 265-volt ballast operating off 480Y,’277-volt systems be used? Should lights be switched individually or in groups? See Chapters 10 and 11. 9. Are voltage-regulating means required? If the primary voltage varies too much, feeder or bus voltage regulators may be required. Should there be load ratio control in the main transformers? Individual regulators? Switched capacitors? Or secondary feeder regulators? See Chapter 4. 10. Short-circuit protection. Do circuit breakers, fuses, fused switches, and motor starters have adequate interrupt’ing rating? Do circuit breakers. fuses, etc., have a margin to allow for increases in plant supply? Is the system laid out so that additions can be made without overstressing circuit breakers and fuses interrupting rating-wise? See Chapters 1 through 3. 11. Grounding. Experience has shown that system neutral grounding is desirable a t all voltage levels. Is the system properly grounded a t the neutrals and are machine frames, etc.: properly grounded to earth? See Chapters 6 and 7. 12. Overcurrent protection. Are there adequate overcurrent devices properly set to provide selective tripping in the event of overcurrent conditions? Are the relays the simplest that will do the job:’ Is there adequate backup protection? Do relays, circuit breakers, and fuses operate selectively wherever practical? Are all parts of the systems adequately protected against the effects of overloads? See Chapter 9. 13. I s the lightning protection adequate? Are the lightning arresters of the proper rating, and are they properly located and grounded? Are the rotating machines adequately protected? See Chapter 5.

rriii

INTRODUCTlON

14. Are circuils properly metered to lell what loads are? Is provision made for connecting portable meters where permanent meters are not justified? Adequate metering is the only way one can check the system performance and to check when loads reach the limit,s of apparatus ratings. Remote metering, i.e., telemeteriug, gives meter readings at centralized points remote from the location of the meters. 15. Power-factor correction. Can the power bill be reduced through reduction of vars? Can you increase voltage and current-carrying capacity by using capacitors? Where should they be installed? Should they be switched? See Chapter 8. 16. Superviswy control. Are there remote water-pumping stations or other loads that should be controlled from the powerhouse or some other central points? Supervisory control will do it. THINGS TO CHECK IF POWER IS GENERATED IN THE INDUSTRIAL PLANT

17. What heat cycle should be used-gas turbine or sfearn turbine? 18. What steam presswes and temperatures are best for steam generation and process use? 19. How big should the boilers and turbines be? 20. What should the balance between extraction at various pressures be, and what should the condenser capacity be? 21. What kind of steam turbine is most suitable to the particular process under consideration? a. Straight condensing b. Extraction c. Topping d. Extraction condensing See Chapter 16.

Industrial Power Systems Handbook - Donald Beeman

Recommend Documents