THE THOUGHT IS IN THE QUESTION THE INFORMATION IS IN THE ANSWER

HAWKINS ELECTRICAL GUIDE
NUMBER EIGHT

QUESTIONS
ANSWERS
&
ILLUSTRATIONS

A PROGRESSIVE COURSE OF STUDY FOR ENGINEERS,
ELECTRICIANS, STUDENTS AND THOSE DESIRING TO
ACQUIRE A WORKING KNOWLEDGE OF

ELECTRICITY AND ITS APPLICATIONS

A PRACTICAL TREATISE
by
HAWKINS AND STAFF

THEO AUDEL & CO. 72 FIFTH AVE. NEW YORK.

COPYRIGHTED, 1915,
BY
THEO. AUDEL & CO.,
New York.

Printed in the United States.

TABLE OF CONTENTS
GUIDE No. 8

CHAPTER LXIII
WAVE FORM MEASUREMENT

The great importance of the wave form in alternating current work is never denied, though it has sometimes been overlooked. The application of large gas engines to the driving of alternators operated in parallel requires an accurate knowledge of the wave form, and a close conformation to a sine wave if parallel operation is to be satisfactory. It is also important that the fluctuations in magnetism of the field poles should be known, especially if solid steel pole faces be used.

If an alternator armature winding be connected in delta, the presence of a third harmonic becomes objectionable, as it gives rise to circulating currents in the winding itself, which increase the heating and lowers the efficiency of the machine.

That the importance of having a good wave form is being realized, is proved by the increasing prevalence in alternator specifications of a clause specifying the maximum divergence allowable from a true sine wave. It is however perhaps not always realized that an alternator which gives a good pressure wave on no load may give a very bad one under certain loads, and the ability of the machine to maintain a good wave form under severe conditions of load is a better criterion of its good design than is the shape of its wave at no load.

The question of wave form is of special interest to the power station engineer. Upon it depends the answer to the questions: whether he may ground his neutral wires without getting large circulating currents; whether he may safely run any combination of his alternators in parallel; whether the constants of his distributing circuit are of an order liable to cause dangerous voltage surges due to resonance with the harmonics of his pressure wave; what stresses he is getting in his insulation due to voltage surges when switching on or off, etc. It has been shown by Rossler and Welding that the luminous efficiency of the alternating current arc may be 44 per cent. higher with a flat topped than with a peaked pressure wave, while on the other hand it is well known that transformers are more efficient on a peaked wave. Also the accuracy of many alternating current instruments depends upon the wave shape.

In making insulation breakdown tests on cables, insulators, or machinery, large errors may be introduced unless the wave form at the time of the test be known. It is not sufficient even to know that the testing alternator gives a close approximation to a sine wave at no load; since if the capacity current of the apparatus under test be moderately large compared with the full load current of the testing alternator, the charging current taken may be sufficient to distort the wave form considerably, thus giving wrong results to the disadvantage of either the manufacturer or purchaser.

Fig. 2,583.—General Electric simultaneous record of three waves with common zero.

The desirability of a complete knowledge of the manner in which the pressure and current varies during the cycle, has resulted in various methods and apparatus being devised for obtaining this knowledge. The apparatus in use for such purpose may be divided into two general classes,

1. Wave indicators;
2. Oscillographs.
and the methods employed with these two species of apparatus may be described respectively as,
1. Step by step;
2. Constantly recording.
that is to say, in the first instance, a number of instantaneous values are obtained at various points of the cycle, which are plotted and a curve traced through the several points thus obtained. A constantly recording method is one in which an infinite number of values are determined and recorded by the machine, thus giving a complete record of the cycle, leaving no portion of the wave to be filled in.

Fig. 2,584.—General Electric simultaneous record of three waves with separate zeros.

Figs. 2,585 and 2,586.—Oscillograms (from paper by Morris and Catterson-Smith, Proc. I. E. E., Vol. XXXIII, page 1,023), showing how the current varies in one of the armature coils of a direct current motor. Fig. 2,585 was obtained with the brushes in the neutral position, and fig. 2,586 with the brushes shifted forward.

The various methods of determining the wave form may be further classified as:

Fig. 2,587.—Oscillogram by Bailey and Cleghorne (Proc. I.E.E., Vol. XXXVIII), showing the sparking pressure or pressure between the brush and the commutator segment at the moment of separation. The waves fall into groups of three owing to the fact that there were three armature coils in each slot.

Fig. 2,588.—Various wave forms. The sine wave represents a current or pressure which varies according to the sine law. A distorted wave is due to the properties of the circuit, for instance, the effect of hysteresis in an iron core introduced into a coil is to distort the current wave by adding harmonics so that the ascending and descending portions may not be symmetrical. A peaked wave has a large maximum as compared with its virtual value. A peaked wave is produced by a machine with concentrated winding.

Joubert's Method.—The apparatus required for determining the wave form by this step by step method, consists of a galvanometer, condenser, two, two way switches, resistance and adjustable contact maker, as shown in [fig. 2,589].

The contact maker is attached to the alternator shaft so that it will rotate synchronously with the latter. By means of the adjustable contact, the instant of "making" that is, of "closing" the testing circuit may be varied, and the angular position of the armature, at which the testing circuit is closed, determined from the scale, which is divided into degrees.

A resistance is placed in series with one of the alternator leads, such that the drop across it, gives sufficient pressure for testing.

Fig. 2,589.—Diagram illustrating Joubert's step by step method of wave form measurement.

Ques. Describe the method of making the test.

Ans. For current wave measurement switch No. 1 is placed on contact F, and for pressure wave measurement, on contact G, switch No. 2 is now turned to M and the drop across the resistance (assuming switch No. 1 to be turned to contact F) measured by charging the condenser, and then discharging it through the galvanometer by turning the switch to S. This is repeated for a number of positions of the contact maker, noting each time the galvanometer reading and position of the contact maker. By plotting the positions of contact maker as abscissæ, and the galvanometer readings as ordinates, the curve drawn through them will represent the wave form.

The apparatus is calibrated by passing a known constant current through the resistance.

Fig. 2,590.—Four part commutator method of wave form measurement. The contact device consists of two slip rings and a four part commutator. One slip ring is connected to one terminal of the source, the other to the voltmeter, and the commutator to the condenser. By adjusting R when a known direct current pressure is impressed across the terminals, the voltmeter can be rendered direct reading.

Fig. 2,591.—Modified four part commutator method of wave form measurement (Duncan's modification). By this method one contact maker can be used for any number of waves having the same frequency. Electro-dynamometers are used and the connections are made as here shown. The moving coils are connected in series to the contact maker, and the fixed coils are connected to the various sources to be investigated, then the deflection will be steady and by calibration with direct current can be made to read directly in volts.

Fig. 2,592.—Diagram illustrating the ballistic galvanometer method of wave form measurement. The test may be made as described in the accompanying text, or in case the contact breaker is belted instead of attached rigidly to the shaft, it could be arranged to run slightly out of synchronism, then by taking readings at regular intervals, points will be obtained along the curve without moving the contact breaker. If this method be used, a non-adjustable contact breaker suffices. In arranging the belt drive so as to run slightly out of synchronism, if the pulleys be of the same size, the desired result is obtained by pasting a thin strip of paper around the face of one of the pulleys thus altering the velocity ratio of the drive slightly from unity.

Ballistic Galvanometer Method.—This method, which is due to Kubber, employs a contact breaker instead of a contact maker. The distinction between these two devices should be noted: A contact maker keeps the circuit closed during each revolution for a short interval only, whereas, a contact breaker keeps the circuit open for a short interval only.

[Fig. 2,592], shows the necessary apparatus and connections for applying the ballistic galvanometer method. The contact breaker consists of a commutator having an ebonite or insulating segment and two brushes.

In operation the contact breaker keeps the circuit closed during all of each revolution, except the brief interval in which the brushes pass over the ebonite segment.

The contact breaker is adjustable and has a scale enabling its various positions of adjustment to be noted.

Ques. Describe the test.

Ans. The contact breaker is placed in successive positions and galvanometer readings taken, the switch being turned to F, [fig. 2,592], in measuring the current wave, and to G in measuring the pressure wave. The results thus obtained are plotted giving respectively current and pressure waves.

Figs.. 2,593 and 2,594.—Two curves representing pressure and current respectively of a rotary converter. Fig. 2,593, pressure wave V, fig. 2,594 current wave C. These waves were obtained from a converter which was being driven by an alternator by means of an independent motor. The rotary converter was supplying idle current to some unloaded transformers and the ripples clearly visible in the pressure wave V, correspond to the number of teeth in the armature of the rotary converter.

Ques. How is the apparatus calibrated?

Ans. By sending a constant current of known value through the resistance R.

Zero Method.—In electrical measurements, a zero method is one in which the arrangement of the testing devices is such that the value of the quantity being measured is shown when the galvanometer needle points to zero.

In the zero method either a contact maker or contact breaker may be used in connection with a galvanometer and slide wire bridge, as shown in [figs. 2,595] and [2,596].

Fig. 2,595.—Diagram illustrating zero method of wave measurement with contact maker. The voltage of the battery must be at least as great as the maximum pressure to be measured and must be kept constant.

Ques. What capacity of battery should be used?

Ans. Its voltage should be as great as the maximum pressure to be measured.

Ques. What necessary condition must be maintained in the battery?

Ans. Its pressure must be kept constant.

Ques. How are instantaneous values measured?

Ans. The bridge contact A is adjusted till the galvanometer shows no deflection, then the length AS is a measure of the pressure.

The drop between these points can be directly measured with a voltmeter if desired.

Ques. How did Mershon modify the test?

Ans. He used a telephone instead of the galvanometer to determine the correct placement of the bridge contact A.

Fig. 2,596.—Diagram illustrating zero method of wave measurement with contact breaker. The voltage of the battery must be at least as great as the maximum pressure to be measured and must be kept constant.

Ques. How can the instantaneous values be recorded?

Ans. By attaching to the contact A, a pencil controlled by an electro-magnet arranged to strike a revolving paper card at the instant of no deflection, the paper being carried on a drum.

Hospitalier Ondograph.—The device known by this name is a development of the Joubert step by step method of wave form measurement, that is to say, the principle on which its action is based, consists in automatically charging a condenser from each 100th wave, and discharging it through a recording galvanometer, each successive charge of the condenser being automatically taken from a point a little farther along the wave.

Fig. 2,597.—Diagram of Hospitalier ondograph showing mechanism and connections. It represents a development of Joubert's step by step method of wave form measurement.

As shown in the diagram, [fig. 2,597], the ondograph consists of a synchronous motor A, operated from the source of the wave form to be measured, connected by gears B to a commutator D, in such a manner that while the motor makes a certain number of revolutions, the commutator makes a like number diminished by unity; that is to say, if the speed of the motor be 900 revolutions per minute, the commutator will have a speed of 899.

The commutator has three contacts, arranged to automatically charge the condenser cc' from the line, and discharge it through the galvanometer E, the deflection of which will be proportional to the pressure at any particular instant when contact is made.

In [fig. 2,597], GG' are the motor terminals, HH' are connected to the condenser cc' through a resistance (to prevent sparking at the commutator) and I, I' are the connections to the service to be measured.

A permanent magnet type of recording galvanometer is employed. Its moving coil E receives the discharges of the condenser in rapid succession and turns slowly from one side to the other.

Fig. 2,598.—View of Hospitalier ondograph. In operation, a long pivoted pointer carrying a pen and actuated by electro-magnets, records on a revolving drum a wave form representing the alternating current, pressure or current wave.

The movable part operates a long needle (separately mounted) carrying a pen F, which traces the curve on the rotating cylinder C. This cylinder is geared to the synchronous motor to run at such a speed as to register three complete waves upon its circumference.

By substituting an electromagnetic galvanometer for the permanent magnet galvanometer, and by using the magnet coils as current coils and the moving coil as the volt coil, the instrument can be made to draw watt curves. [Fig. 2,598] shows the general appearance of the ondograph.

Cathode Ray Oscillograph.—This type of apparatus for measuring wave form was devised by Braun, and consists of a cathode ray tube having a fluorescent screen at one end, a small diaphragm with a hole in it at its middle, and two coils of a few turns each, placed outside it at right angles to one another. These coils carry currents proportional to the pressure and current respectively of the circuit under observation.

Fig. 2,599.—General Electric moving coil oscillograph complete with tracing table. The tracing table is employed for observing the waves, and by using a piece of transparent paper, the waves under observation appear as a continuous band of light which can be traced, thus making a permanent record. This is not, however, to be regarded as a recording attachment, and can not be used where instantaneous phenomena are being investigated. The synchronous motor for operating the synchronous mirror in connection with tracing and viewing attachment is wound for 100 to 115 volts, 25 to 125 cycles, and should, of course, be run from the same machine which furnishes power to the circuit under observation. A rheostat for steadying and adjusting the current should be connected in series with the motor. The beam from the vibrator mirrors striking this synchronous mirror moves back and forth over the curved glass, and gives the length of the wave; the movement of the vibrator mirror gives the amplitude, and the combination gives the wave complete. An arc lamp or projection lantern produces the image reflected by the mirrors upon the film, tracing table or screen. For the rotation of the photographic film, a small direct current shunt wound motor is ordinarily used.

The ray then moves so as to produce an energy diagram on the fluorescent screen.

Fig. 2,600.—General Electric moving coil oscillograph. The moving elements consist of single loops of flat wire carrying a small mirror and held in tension by small spiral springs. The current passing down one side and up the other, forces one side forward and the other backward, thus causing the mirror to vibrate on a vertical axis. The vibrator elements fit into chambers between the poles of electro-magnets, and are adjustable, so as to move the beam from the mirror, both vertically and horizontally. A sensitized photographic film is wrapped around a drum and held by spring clamps. The drum, with film, is placed in a case and a cap then placed over the end, making the case light, when the index is either up or down. The loading is done in a dark room. A driving dog is screwed into the drum shaft, and which, when the drum and case are in place, revolves the film past a slot. When an exposure is to be made, the index is moved from the closed position, thus opening the slot in the case and exposing the film to the beam of light from the vibrating mirrors when the electrically operated shutter is open. The slot is then closed by moving the index to "Exposed." A slide with ground glass can be inserted in place of the film case or roll holder to arrange the optical system when making adjustments. The shutter operating mechanism is arranged so as to hold the shutter open during exactly one revolution of the film drum. There are two devices connected to the shutter operating mechanism; one opens the shutter at the instant the end of the film passes the slot; the other opens immediately, at any part of the film, and both give exposure during one revolution. The first is useful when making investigations in which the events are either recurring, or their beginnings known or under control, and the second when the time of the event is not under control, such as the blowing of fuses or opening of circuit breakers.

The instrument is much used in wireless telegraphy, as it is capable of showing the characteristics of currents of very high frequency.

Fig. 2,601.—General Electric moving coil oscillograph with case removed, showing interior construction and arrangement of parts. The oscillograph is furnished complete with a three element electro-magnet galvanometer, optical system, shutter and shutter operating mechanism, film driving motor and cone pulleys, photographic and tracing attachments, 6 film holders, and the following repair parts, for vibrators: 6 extra suspension strips; 6 vibrator mirrors; 1 box gold leaf fuses; 1 bottle mirror cement; 1 bottle damping liquid.

Fig. 2,602.—Oscillogram showing the direct current pressure of a 25 cycle rotary converter (below), and (above) the pressure wave taken between one collector ring and one commutator brush. The 12 ripples per cycles in the direct current voltage are due to a 13th harmonic in the alternating current supply.

Glow Light Oscillograph.—This device consists of two aluminum rods in a partially evacuated tube, their ends being about two millimeters apart. When an alternating current of any frequency passes between them a sheath of violet light forms on one of the electrodes, passing over to the other when the current reverses during each cycle. The phenomenon may be observed or photographed by means of a revolving mirror.

Fig. 2,603.—Curves by Morris, illustrating the dangerous rush of current which may occur when switching on a transformer. The circuit was broken at F and made again at G. The current was so great as to carry the spot of light right off the photographic plate due to the fact that a residual field was left in the core after switching off, and on closing the switch again the direction of the current was such as to tend to build up the full flux in the same direction as this residual flux. The dotted lines have been drawn in to show how the actual waves were distorted from the normal.

Moving Iron Oscillograph.—This type is due to Blondel, to whom belongs the credit of working out and describing in considerable detail the principles underlying the construction of oscillographs.

The moving iron type of oscillograph consists of a very thin vane of iron suspended in a powerful magnetic field, thus forming a polarized magnet. Near this strip are placed two small coils which carry the current whose wave form is to be measured.

The moving iron vane has a very short period of vibration and can therefore follow every variation in the current.

Fig. 2,604.—Siemens-Blondel moving coil type oscillograph. The coil is in the shape of a loop of thin wire, which is suspended in the field of an electro-magnet excited by continuous current. The current to be investigated is sent through this loop, which in consequence of the interaction of current and magnetic field, begins to vibrate. The oscillations are rendered visible by directing a beam of light from a continuous current arc lamp onto a small mirror fixed to the loop. The light reflected by the mirror is in the form of a light strip, but by suitable means this is drawn out in respect of time, so that a curve truly representing the current is obtained. The loop of fine wire is stretched between two supports and is kept in tension by a spring. As the spring tension is considerable, the directive force of the vibrating system is large, and its natural periodicity very high. The mirror is fixed in the center of the loop, and has an area of 1 square mm. In order to protect the loops from mechanical injury they are built into special frames. The mirrors are of various sizes, the loop for demonstration purposes (projection device) being provided with the largest mirror and the most sensitive loop with a mirror of the smallest dimensions.

Attached to the vane is a small mirror which reflects a beam of light upon some type of receiving device.

The Siemens-Blondel oscillograph shown in [fig. 2,604], is of the moving coil type, being a development of the moving iron principle.

Moving Coil Oscillograph.—The operation of this form of oscillograph is based on the behaviour of a movable coil in a magnetic field.

Figs. 2,605 and 2,606.—Oscillograms reproduced from a paper by M. B. Field on "A Study of the Phenomena of Resonance by the Aid of Oscillograms" (Journal of E. E., Vol. XXXII). The effect of resonance on the wave forms of alternators has been the subject of much investigation and discussion; it is a matter of vital importance to the engineer in charge of a large alternating current power distribution system. Fig. 2,605 shows the pressure curve of an alternator running on a length of unloaded cable, the 11th harmonic being very prominent. Fig. 2,606 shows the striking alteration produced by reducing the length of cable in the circuit and thus causing resonance with the 13th harmonic.

It consists essentially of a modified moving coil galvanometer combined with a rotating or vibrating mirror, a moving photographic film, or a falling photographic plate. The galvanometer portion of the outfit is usually referred to as the oscillograph as illustrated in [figs. 2,608 to 2,612], representing diagrammatically the moving system.

In the narrow gap between the poles S, S of a powerful magnet are stretched two parallel conductors formed by bending a thin strip of phosphor bronze back on itself over an ivory pulley P. A spiral spring attached to this pulley serves to keep a uniform tension on the strips, and a guide piece L limits the length of the vibrating portion to the part actually in the magnetic field.

A small mirror M bridges across the two strips as shown. The effect of passing a current through such a "vibrator" is to cause one of the strips to advance while the other recedes, and the mirror is thus turned about a vertical axis.

Fig. 2,607.—General view of electro-magnet form of Duddell moving coil oscillograph, showing oil bath and electro-magnet. This instrument is specially designed to have a very high natural period of vibration (about 1/10,000 of a second) so as to be suitable for accurate research work. It is quite accurate for frequencies up to 300 per second. In the figure, A is the brass oil bath in which two vibrators are fixed; B, core of electro-magnet which is excited by two coils, one of which, C, is seen. The ends of these two coils are brought out to four terminals at D, so that the coils may be connected in series for 200 volt, or in parallel for 100 volt circuits. The bolts, E,E, hold the oil bath in position between the poles of the magnet. F,F,F (one not seen), are levelling screws; G,G, terminals of one vibrator; H, fuse; K, thermometer with bulb in center of oil bath.

Figs. 2,608 to 2,612.—Vibrator of Duddell moving coil oscillograph and section through oil bath of electro-magnet oscillograph. The vibrator consists of a brass frame W, which supports two soft iron pole pieces P,P. Between these, a long narrow groove is divided into two parts by a thin soft iron partition, which runs up the center. The current being led in by the brass wire U, passes from an insulated brass plate to the strip, which is led over an ivory guide block, down one of the narrow grooves and over another guide block, the loops round the ivory pulley O, which puts tension on the strip by the spring N, back to the guide block again, up the other narrow groove, and out by way of the insulated brass plate and lead U. Halfway up the grooves the center iron partition R is partially cut away to permit of a small mirror M, bridging across from one strip to the other, being stuck to the strips by a dot of shellac at each corner. The figure illustrates one type of vibrator in which P is removable from W for ease in repairing. In type 1, these pole pieces P,P are not removable. The vibrators are placed side by side in the gap between the poles S,S of the electro-magnet, [ see fig. 2,610]. Each vibrator is pivoted about vertical centers, the bottom center fitting in the base of the oil bath, and the one at the top being formed by a screw in the cock piece Y. It can thus be easily turned in azimuth, its position being fixed by the adjusting screw L, a spiral spring serving to keep the vibrator always in contact with this screw. Since each cock piece can be independently moved forward or backward, each vibrator can be tipped slightly in either of these directions so that complete control over the mirrors is obtained and reflected spots of light may be made to coincide with that reflected from the fixed zero mirror, which latter is fixed to a brass tongue in between the two vibrators. A plano-convex lens of 50 cm. focal length is fixed on the oil bath in front of the vibrator mirrors to converge the reflected beams of light. It will be noticed that this lens is slightly inclined so that no trouble will be given by reflections from its own surface. The normal distance from the vibrator mirrors to the scale of photographic plate is 50 cm., and at this distance, a convenient working deflection on each side of the zero line is 3 to 4 cm. This is obtained with a R.M.S. current through the strips of from .05 to .1 of an ampere according to wave form, etc. The maximum deflection on each side of the zero line should not exceed 5 cm. while the maximum R.M.S. current through the strips should in no case exceed .1 ampere.

Each strip of the loop passes through a separate gap (not shown in the figure). The whole of the "vibrator," as this part of the instrument is called, is immersed in an oil bath, the object of the oil being to damp the movement of the strips, and make the instrument dead beat. It also has the additional advantage of increasing by refraction the movement of the spot of light reflected from the vibrating mirrors.

The beam of light reflected from the mirror M is received on a screen or photographic plate, the instantaneous value of the current being proportional to the linear displacement of the spot of light so formed.

With alternating currents, the spot of light oscillates to and fro as the current varies and would thus trace a straight line.

To obtain an image of the wave form, it is necessary to traverse the photographic plate or film in a direction at right angles to the direction of the movement of the spot of light.

Fig. 2,613.—Duddell moving coil oscillograph with projection and tracing desk outfit. The outfit is designed for teaching and lecture purposes. In operation, after the beam of light from the arc lamp has been reflected from the oscillograph mirrors, it falls on a vibrating mirror which gives it a deflection proportional to time in a direction at right angles to the deflection it already has and which is proportional to the current passing through the oscillograph. It is therefore only necessary to place a screen in the path of the reflected beam of light to obtain a trace of the wave form. Since the vibrating mirror is vibrated by means of a cam on the shaft of a synchronous motor, which motor is driven from, or synchronously with, the source of supply whose wave form is being investigated, the wave form is repeated time after time in the same place on the screen, and owing to the "persistence" of vision, the whole wave appears stationary on the screen. The synchronous motor with its vibrating mirror, mentioned above, is located underneath the "tracing desk." When used in this position a wave a few centimeters in amplitude is seen through a sheet of tracing paper which is bent round a curved sheet of glass. A permanent record of the wave form can thus easily be traced on the paper. A dark box which is designed to hold a sheet of sensitized paper in place of the tracing paper, can be fitted in place of the tracing desk. Thus an actual photographic record of the wave form is obtained. If the synchronous motor be transferred from its position underneath the tracing desk to the space reserved for it close to the oscillograph, the beam of light is then received on a large mirror which is placed at an angle of about 45 degrees to the horizontal and so projects the wave form onto a large vertical screen which should be fixed about two and a half meters distant. Under these conditions a wave form of amplitude 50 cm. each side the zero line may be obtained which is therefore visible to a large audience.

Ques. How are the oscillograms obtained in the Duddell moving coil oscillograph?

Ans. In all cases the oscillograms are obtained by a spot of light tracing out the curve connecting current or voltage with time. The source of light is an arc lamp, the light from which passes first through a lens, and then, excepting when projecting on a screen, through a rectangular slit about 10 mm. long by 1 mm. wide. The position of the lamp from the lens is adjusted till an image of the arc is obtained covering the three (two moving, one fixed) small oscillograph mirrors. The light is reflected back from these mirrors and, being condensed by a lens which is immediately in front of them, it converges till an image of the slit is formed on the surface where the record is desired. All that is necessary now to obtain a bright spot of light instead of this line image is to introduce in the path of the beam of light a cylindrical lens of short focal length.

Figs. 2,614 and 2,615.—Sectional view of permanent magnet form of Duddell moving coil oscillograph. This instrument has a lower natural period of vibration (1/3000 second) than the type shown in [fig. 2,612], and therefore is not capable of accurately following wave forms of such high frequency, but it is sufficiently quick acting to follow wave forms of all ordinary frequencies with perfect accuracy. It is easier to repair, and more portable owing to the fact that the magnetic field is produced by a permanent magnet instead of an electro-magnet. This also renders the instrument suitable for use on high tension circuits without earth connection, as, owing to the fact that no direct current excitation is required, the instrument is more easily insulated than other types.

Ques. What is the function of the mirrors on the vibrating vane?

Fig. 2,616.—Diagram of connections of Duddell oscillograph to high pressure circuit. The modification necessary for high pressure circuit only applies to the vibrator which gives the pressure wave and consists in adding two more resistances, R₄ and R₅. Referring to [fig. 2,617], it will be seen that in case fuse f₂ blows, or the vibrator be accidentally broken, the full supply voltage is immediately thrown on the instrument itself. This is not permissible in high voltage work and therefore the resistance R₅ is introduced as a permanent shunt to the oscillograph vibrator. The resistance R₄ is an exact duplicate of R₂ being a 21 ohm plug resistance box for adjusting the sensitivity of the vibrator to an even figure. In practice R₅ is usually a part of R₁, and in most of the high voltage resistances, two taps are brought out near one end to serve as R₅. One of these taps is usually 50 ohms distant from the end terminal and the other only 5 ohms from the end. The use of these taps is as follows: The large resistance consisting of R₁ + R₅ is so chosen with respect to the voltage of the circuit under investigation that the current through R₁ is about .1 ampere. It should never be more than this continuously. Then R₄ is connected to the 50 ohm tap, and since the resistance of the oscillograph vibrator circuit is variable from about 5 to 26 ohms by means of R₄, the current can be controlled through the oscillograph from about .066 to .091 of an ampere, enabling an open wave form to a convenient scale to be obtained. If it now be desired to record large rises of pressure, such as may occur in cases of resonance, the height of the wave must be reduced in order to keep these rises on the plate. This is accomplished by disconnecting R₄ from the 50 ohm tap and connecting it to the 5 ohm tap, when the current through the vibrator will be from .05 to .016 of an ampere according to whether the resistance R₄ is in or out of circuit. When, instead of using the falling plate, the cinematograph camera is being used, it becomes necessary always to work on the 5 ohm tap since the width of the film is much less than that of the plate, and the current must therefore be less. In experiments where sudden rises of voltage are expected it is often advisable to keep R₁ as great as possible. That end of the resistance R₁ referred to as R₅ in the diagram should be securely connected to the supply main and no switch or fuse used. A switch may, if desired, be used in series with R₁, provided it be inserted at the point where R₁ joins the supply main remote from R₅. It will be seen that fuses f₁ and f₂ are shown. Provided that the connections are always made in accordance with the diagram, and the vibrators are always shunted by R₅ or R₃ respectively, there is not much objection to the use of these fuses, but on general principles it is wise to avoid fuses in high tension work and accordingly with each permanent magnet oscillograph, dummy fuses are supplied, which can be inserted in place of the ordinary fuses when desired. The remark previously made about keeping both vibrators and the frame of the instrument at approximately the same pressure applies with additional emphasis in high pressure work.

Ans. They simply control the direction of a beam of light in a horizontal plane in such a manner that its deflection from a zero position depends on the current passing through the instrument, and it is therefore evident that the oscillograph is not complete without means of producing a time scale.

Fig. 2,617.—Diagram of connections of Duddell oscillograph to low pressure circuit, R₁ is a high non-inductive resistance connected across the mains in series with one of the vibrators. S₂ is a switch, and f₂, the fuse (on the oscillograph in this circuit). The resistance of R₁ in ohms should be rather more than ten times the voltage of the circuit, so that a current of a little less than .1 of an ampere will pass through it. The vibrator will then give the curve of the circuit on an open scale. (For the projection oscillograph, the resistance R₁ should be only twice the supply voltage, since .5 of an ampere is required to give full scale deflection on a large screen.) To obtain the current wave form, the shunt R₃ is connected in series with the circuit under investigation and the second vibrator is connected across this shunt. Here also f₁ is a fuse, S₁ a switch, and R₂ an adjustable resistance box. The switch S₁ is however unnecessary if the plug resistance box supplied for R₂ be used, since an infinity plug is included in this box. The shunt R₃ should have a drop of about 1 volt across it in order to give a suitable working current through the vibrator. The resistance R₂ is not absolutely essential, but it is a great convenience in adjusting the current through the vibrator. It is a plug resistance box, the smallest coil being .04 of an ohm and the total 21 ohms. Being designed to carry .5 ampere continuously it can be used with any other type of Duddell oscillograph, and by its use the sensitiveness of the vibrator can be adjusted so that a round number of amperes in the shunt gives 1 mm. deflection. This adjustment is best made with direct current. It should be noted in connecting the oscillograph in circuit, that the two vibrators should be so connected to the circuit that it is impossible that a higher pressure difference than 50 volts should exist between one vibrator and the other, or between either vibrator and the frame. To ensure attention to this important point, a brass strap is provided which connects the two vibrators together and to the frame of the instrument. This does not mean that this point must necessarily be earthed since the frame of the instrument is insulated from the earth. It is advisable, however, to earth it when possible.

Figs. 2,618 and 2,619.—Two curves obtained with the falling plate camera and illustrating the discharge of a condenser through an inductive circuit. When taking curve A the resistance in the circuit was very small compared to the inductance, while before taking curve B an additional non-inductive resistance was inserted in the circuit so that the oscillations were damped out much more rapidly although the periodic time remained approximately constant.

Ques. How is the time scale produced?

Ans. Either the surface on which the beam of light falls may be caused to move in a vertical plane with a certain velocity, so that the intersection of the beam and the plane surface traces out a curve connecting current with time (a curve which becomes a permanent record if a sensitized surface be used); or, the surface may remain stationary and in the path of the horizontally vibrating beam may be introduced a mirror which rotates or vibrates about a horizontal axis, thus superposing a vertical motion proportional to time on the horizontal vibration which is proportional to current, and causing the beam of light to trace out a curve connecting current and time on the stationary surface.

Ques. What kind of recording apparatus is used with the Duddell oscillograph?

Ans. A falling plate camera, or a cinematograph film camera.

Fig. 2,620.—Synchronous motor with vibrating mirror as used with Duddell moving coil oscillograph. Since the motor must run synchronously with the wave form it is required to investigate, it should be supplied with current from the same source. The motor can be used over a wide range of frequencies (from 20 to 120). When working at frequencies below 40, it is advisable to increase the moment of inertia of the armature, and for this purpose a suitable brass disc is used. The armature carries a sector, which cuts off the light from the arc lamp during a fraction of each revolution, and a cam which rocks the vibrating mirror. It makes one revolution during two complete periods, and the cam and sector are so arranged that during 1½ periods, the mirror is turning with uniform angular velocity, while during the remaining half period, the mirror is brought back quickly to its angular position, the light being cut off by the sector during this half period.

Ques. Explain the operation of the falling plate camera.

Ans. In this arrangement a photographic plate is allowed to fall freely by the force of gravity down a dark slide. At a certain point in its fall it passes a horizontal slit through which the beams of light from the oscillograph pass, tracing out the curves on the plate as it falls.

Figs. 2,621 to 2,623.—Interior of cinematograph camera as used on Duddell moving coil oscillograph for obtaining long records. The loose side of case is shown removed and one of the reels which carry the film lying in front. The spool of film which is placed on the loose reel A, passes over the guide pulley B, then vertically downward between the brass gate D (shown open in the figure), and the brass plate C. The exposure aperture is in the plate C and can be opened or closed by a shutter controlled by the lever M. The groove in the plate C, and the springs which press the gate D flat on the plate C, prevent the film having any but a vertical motion as it passes the exposure slit. E is the sprocket driving pulley which engages with the perforations on the film and unwinds it from the reel A to reel H. Outside the case on the far side of it is secured to the axle G a three speed cone pulley. This is driven by a motor of about 1/7 horse power, which also drives, through the gears shown, the sprocket pulley E. Close to the grooved cone pulley is a lever carrying a jockey pulley L, and a brake, which latter is normally held onto the cone pulley by a spring and so causes the loose belt to slip. By pressing a lever which is attached to the falling plate camera case, the brake can be suddenly released and at the same time the jockey pulley caused to tighten the belt onto the grooved cone pulley, so that the starting and stopping of the film is controlled independently of the driving motor, and being quickly accomplished avoids waste of film. Both reels are alike and each is made in two pieces. The upper reel is loose on its axle and its motion is retarded slightly by a friction brake. The lower reel is also loose on its axle, but it is driven by means of a friction clutch, the clutch always rotating faster than the reel so that the used film delivered by the sprocket pulley E is wound up as fast as delivered. K is the front face of one reel, the boss on it pushes into the tube on the other half H, which serves not only to unite the two halves, but also to secure the end of the film which is doubled through J.

The mean speed of the plate at the moment of exposure is about 13 feet per second. This speed is very suitable for use with frequencies of from 40 to 60 periods per second. A cloth bag is used to introduce the plate to the slide.

A catch holds the plate until it is desired to let it fall. Inside the case, is a small motor, 100 or 200 volts direct current, driving four mirrors which are fixed about a common axis with their planes parallel to it.

Fig. 2,624.—Portion of oscillograph record taken with cinematograph film camera, showing the rush of current and sudden rise of voltage at the moment of switching on a high pressure feeder.

By looking through a small slot in the end of the camera into these rotating mirrors, the observer sees the wave form which the oscillograph is tracing out and is thus able to make sure that he is obtaining the particular wave form or other curve desired before exposing the plate.

Fig. 2,625.—Portion of oscillograph record taken with a cinematograph film camera showing the effect of switching off a high pressure feeder and illustrating the violent fluctuations produced by sparking at the switch contacts.

The plate falls into a second red cloth bag which is placed on the bottom of the slide. The plates used are "stereoscopic size", 6¾" × 3¼" (17.1 × 8.3 cm.).

Ques. For what use is the cinematograph camera adapted?

Ans. For long records.

For instance, in investigations, such as observation on the paralleling of alternators, the running up to speed of motors, and the surges which may occur in switching on and off cable, etc. The cinematograph camera fits on to the falling plate case and by means of which a roll of cinematograph film can be driven at a uniform speed past the exposure aperture, enabling records up to 50 metres in length to be obtained. An interior view of the cinematograph camera is shown in [fig. 2,621].

Fig. 2,626.—Curves reproduced from an article by J. T. Morris in the Electrician. "On recording transitory phenomena by the oscillograph."

Fig. 2,627.—First rush of current from an alternator when short circuited, showing unsymmetrical initial wave of current, becoming symmetrical after a few cycles. 25 cycles.

Fig. 2,628.—Pressure wave obtained from narrow exploring coil on alternator armature, indicating distribution of field flux. The terminal voltage of the alternator is very nearly a sine wave, 60 cycles; about 17 volts.

SOME OSCILLOGRAPH RECORDS

Fig. 2,629.—The waves of voltage and current of an alternating arc. A, voltage wave; B, current wave showing low power factor of the arc without apparent phase displacement. 60 cycles.

Fig. 2,630.—Rupturing 650 volt circuit. A, current wave; B, 25 cycle wave to mark time scale.

Fig. 2,631.—First rush of current from alternator when short circuited, showing unsymmetrical current wave, also wave of field current caused by short circuit current in armature. Upper curve, armature current; lower curve, field current.

Fig. 2,632.—Mazda (tungsten) lamp, showing rapid decrease to normal current as filament heats up. 25 cycles.

Fig. 2,633.—Current wave in telephone line corresponding to sustained vowel sound "i," as in machine; voice pitched at A 110.

Fig. 2,634.—Carbon lamp, showing rapid increase to normal current as filament heats up. 25 cycles.

Fig. 2,635.—Short circuit current on direct current end of rotary converter, 21,500 amperes maximum. Upper curve, direct current voltage; lower curve, direct current amperage. Duration of short circuit about .1 second.

CHAPTER LXIV
SWITCHBOARDS

General Principles of Switchboard Connections.—The interconnection of generators, transformers, lines, bus bars, and switches with their relays, in modern switchboard practice is shown by the diagrams, figs. 2,636 to [2,645]. The figures being lettered A to J for simplicity, the generators are indicated by black discs, and the switches by open circles, while each heavy line represents a set of bus bars consisting of two or more bus bars according to the system of distribution. It will be understood, also, in this connection, that the number of pole of the switches and the type of switch will depend upon the particular system of distribution employed.

Diagram A, shows the simplest system, or one in which a single generator feeds directly into the line. There are no transformers or bus bars and only one switch is sufficient.

In B, a single generator supplies two or more feeders through a single set of bus bars, requiring a switch for each feeder, and a single generator switch.

In C, two generators are employed and required and the addition of a bus section switch.

D, represents a number of generators supplying two independent circuits. The additional set of bus bars employed for this purpose necessitates an additional bus section switch, and also additional selector switches for both feeders and generators.

E, shows a standard system of connection for a city street railway system having a large number of feeders.

Figs. 2,645 and 2,646.—Diagrams illustrating general principles of switchboard connections.

This arrangement allows any group of feeders to be supplied from any group of generators.

Fig. 2,646.—Fort Wayne switchboard panel for one alternator and one transfer circuit. Diagram giving dimensions, arrangement of instruments of board, and method of wiring. The different forms of standard alternating current switchboard panels for single phase circuits made by the Fort Wayne Electric Works are designed to fulfill all the usual requirements of switchboards for this class of work. The line includes panels equipped for a single generator; for one generator and two circuits; one generator and one transfer circuit; one generator, an incandescent and an arc lighting circuit; and also feeder panels of different kinds.

It also permits the addition of a generator switch for each generator.

F, represents the simplest system with transformers.

It requires a single generator transformer bank, switch and line. The arrangement as show at F is used where a number of plants supply the same system.

G, represents a system having more than one line.

In this case a bus bar and transformer switch is used on the high tension side.

H, shows a number of generators connected to a set of low tension bus bars through generator switches, and employing a low tension transformer switch.

I, shows the connections of a system having a large number of feeders supplied by several small generators. In this case, the plant is divided into two parts, each of which may be operated independently.

J, represents the arrangement usually employed in modern plants where the generator capacity is large enough to permit of a generator transformer unit combination with two outgoing lines. By operating in parallel on the high tension side only, any generator can be run with any transformer. The whole plant can be run in parallel, or the two parts can be run separately.

Fig. 2,647.—General Electric small plant alternating current switchboard, designed for use in small central stations and isolated plants. They are for use with one set of bus bars, to which all generators and feeders are connected by means of single throw lever switches or circuit breakers, suitable provision being made for the parallel operation of the generators.

Fig. 2,648.—Crouse-Hinds voltmeter and ground detector radial switch, arranged for mounting on the switchboard. The switch proper is placed on the rear of the board with hand wheel, dial, and indicator only on the front side. The current carrying parts are of hard brass, with contact surfaces machined after assembling. The contact parts are of the plunger spring type, and the cross bar has fuse connections. Ground detector circuits are marked G+ and G- for two wire system, and G+, G-, GN+ and GN- for three wire system. When the voltmeter switch is to be used as a ground detector, two circuits are required for a two wire system, and four circuits for a three wire system, that is, a six circuit voltmeter and ground detector switch for use on a two wire system has two circuits for ground detector and four circuits for voltmeter readings. A six circuit voltmeter and ground detector switch, for use on a three wire system, has four circuits for ground detector and two circuits for voltmeter readings.

Switchboard Panels.—The term "panel" means the slab of marble or slate upon which is mounted the switches, and the indicating and controlling devices. There are usually several panels comprising switchboards of moderate or large size, these panels being classified according to the division of the system that they control, as for instance:

1. Generator panel;
2. Feeder panel;
3. Regulator panel, etc.

In construction, the marble or slate should be free from metallic veins, and for pressures above, say, 600 volts, live connections, terminals, etc., should preferably be insulated from the panels by ebonite, mica, or removed from them altogether, as is generally the case with the alternating gear where the switches are of the oil type.

Figs. 2,649 and 2,650,—Wiring diagrams of Crouse-Hinds voltmeter and ground detector switches. Fig. 2,649 voltmeter switch; fig. 2,650 voltmeter and ground detector switch. A view of the switch is shown in [fig. 2,648]; it is designed for use on two or three wire systems up to 300 volts.

The bus bars and connections should be supported by the framework at the back of the board, or in separate cells, and the instruments should be operated at low pressure through instrument transformers.

The panels are generally held in position by bolting them to an angle iron, or a strip iron framework behind them.

Generator Panel.—This section of a switchboard carries the instruments and apparatus for measuring and electrically controlling the generators. On a well designed switchboard each generator has, as a rule, its own panel.

Figs. 2,651 to 2,653.—Diagrams of connections for generator panels. Key to symbols: A, ammeter; A.S., ammeter switch; C.T., current transformer; F., fuse; F.A., direct current field ammeter; F.S., field switch; G.C.S., governor control switch; L.S., limit switch (included with governor motor); O.S., oil switch; P.I.W., polyphase indicating wattmeter; P.W.M., polyphase watthour meter; P.R., pressure receptacle; P.P., pressure plug; Rheo., rheostat; S., shunt; S.R., synchronizing receptacle; S.P., synchronizing plugs; T.B., terminal board for instrument leads; V, alternating current voltmeter.

Figs. 2,654 and 2,655.—Diagrams illustrating a simple method of determining bus capacity as suggested by the General Electric Co. Fig. 2,654 relates to any panel; the method is as follows: 1. Make a rough plan of the entire board, regardless of the number of panels to be ordered. The order of panels shown is recommended, it being most economical of copper and best adapted to future extensions. 2. To avoid confusion keep on one side of board everything pertaining to exciter buses, and on other side everything pertaining to A. C. buses. 3. With single lines represent the exciter and A. C. buses across such panels as they actually extend and by means of arrows indicate that portion of each bus which is connected to feeders and that portion which is connected to generators. Remember that "Generator" and "Feeder" arrows must always point toward each other, otherwise the rules given below do not hold. Note also that the field circuits of alternator panels are treated as D. C. feeders for the exciter bus. 4. On each panel mark its ampere rating, that is, the maximum current it supplies to or takes from the bus. For A. C. alternator panels the D. C. rating is the excitation of the machines. 5. Apply the following rules consecutively, and note their application in fig. 2,654. (For the sake of clearness ampere ratings are shown in light face type and bus capacities in large type.) A. Always begin with the tail of the arrow and treat "generator" and "feeder" sections of the bus separately. B. Bus capacity for first panel = ampere rating of panel. C. Bus capacity for each succeeding panel = ampere rating of panel plus bus capacity for preceding panel. (See sums marked above the buses in fig. 2,654.) D. For a panel not connected to a bus extending across it, use the smaller value of the bus capacities already obtained for the two adjoining panels. (See exciter bus for panel C.) E. The bus capacity for any feeder panel need not exceed the maximum for the generator panels (see A. C. bus for panel G) and vice versa (see exciter bus for panel B). Hence the corrections made in values obtained by applying rules B and C. The arrangement of panels shown in fig. 2,654 is the one which is mostly used. The above method may, however, be applied to other arrangements, one of which is shown in fig. 2,655. Here the generators must feed both ways to the feeders at either end of the board so that in determining A. C. bus capacities it is necessary to first consider the generators with the feeders at one end, and then with the feeders at the other end as shown by the dotted A. C. buses. The required bus capacities are then obtained by taking the maximum values for the two cases.

Fig. 2,656.—End view showing general arrangement of switchboards for 240, 480, and 600 volt alternating current. The cut shows a single throw oil switch mounted on the panel.

In the case of a dynamo, a good representative panel would have mounted upon it a reverse current circuit breaker, an ammeter, a double pole main switch (or perhaps a single pole switch, since the circuit breaker could also be used as a switch) a double pole socket into which a plug could be inserted to make connection with a voltmeter mounted on a swinging bracket at the end of the board; a rheostat handle, the spindle of which operates the shunt rheostat of the machine, the rheostat being placed either directly behind the spindle, if of small size, or lower down with chain drive from the hand wheel spindle, if of larger size, a field discharge switch and resistance, a lamp near the top of the panel for illuminating purposes, a fuse for the voltmeter socket, and, if desired, a watthour meter. If the dynamo be compound wound, the equalizing switch will generally be mounted on the frame of the machine, and in some cases the field rheostat will be operated from a pillar mounted in front of the switchboard gallery. If the generator be for traction purposes, the circuit breaker is more often of the maximum current type, and a lightning arrester is often added, without a choke coil, the latter as well as further lightning arresters being mounted on the feeder panels.

Figs. 2,657 and 2,658.—Two views of a feeder panel, showing general arrangement of the devices assembled thereon. A, circuit breaker; B, ammeter; C, voltmeter; D, switches.

In the case of a high pressure alternating current plant of considerable size, the bus bars oil switches, and the current and pressure transformers are generally mounted either in stoneware cells, or built on a framework in a space guarded by expanded metal walls, and no high pressure apparatus of any sort is brought on to the panels themselves.

Figs. 2,659 to 2,666.—Diagram of connections for three phase feeder panels. Key to symbols: A, ammeter; A.S., three way ammeter switch; B.A.S., bell alarm switch; C.T., current transformer; F, fuse; O.S., oil switch; P.I.W., polyphase indicating wattmeter; P.W.M., polyphase watthour meter; T.B., terminal board; T.C., trip coils for oil switch.

Feeder Panel.—The indicating and control apparatus for a feeder circuit is assembled on a panel called the feeder panel.

The most common equipment in the case of a direct current feeder panel comprises an ammeter, a double pole switch, and double pole fuses or instead of the fuses, a circuit breaker on one or both poles; in the case of a traction feeder a choke coil and a lightning arrester are often added.

Figs. 2,667 and 2,668.—Diagrams of connections for two phase and three phase installations: A and A1, ammeter; C.C., constant current transformer; C.T., current transformer; D.R., discharge resistance; F, fuse; F.S., field switch; L.A., lightning arrester; O.S., oil switch; P.P., pressure plug; P.R., pressure receptacle; P.T., pressure transformer; S and S1, plug switches; T.C., oil switch trip coil; V, voltmeter.

The equipment of a typical high pressure three phase feeder panel is an ammeter (sometimes three ammeters, one in each phase) operated by a current transformer, and oil break switch with two overload release coils, or three if the neutral of the circuit be earthed, the releases being operated by current transformers.

Fig. 2,669.—Crouse-Hinds radial ammeter switch, arranged for mounting directly on the switchboard. It is designed for use with external shunt ammeters of any make or capacity, and in connection with the required number of shunts, makes possible the taking of current readings of a corresponding number of circuits by means of one ammeter. The wiring diagram is shown in [fig. 2,670].

The switch when on a large system is often in a cell some distance behind the panel, and is then controlled by a system of levers, or by a small motor which is started and stopped by a throw over switch on the panel, in which case there is generally a lamp or lamps on the panel to show whether the switch is open or closed.

Air brake switches or links are placed between the bus bars and the oil switch to allow of the latter being isolated for inspection purposes, and as a general rule no apparatus carrying high pressure current is allowed on the front of the panel. With both direct and alternating current feeders, a watthour meter is often added to show the total consumption of the circuit.

Fig. 2,670.—Wiring diagram for Crouse-Hinds radial ammeter switch as illustrated in [fig. 2,669]. The switch proper is on the rear of the switchboard, and the hand wheel dial and indicator on the front.

A typical three phase generator panel is provided with three ammeters, one in each phase, operated from three current transformers, one to each ammeter, a volt meter, a power factor indicator, and an indicating watthour meter, all operated from one or more pressure transformers, and the necessary current transformers, the operating handle of the oil switch, which is connected to the switch itself by means of rods, two maximum releases operated by current transformers, or a reverse relay for automatically tripping the switch, lamps for indicating when the switch is tripped, a socket for taking the plug which makes connection between the secondary of a pressure transformer and the synchronizer on the synchronizing panel, and a lamp for illuminating purposes, while on the base of the panel or on a pillar at the front of the gallery is mounted the gear for the field circuit. This consists of a double pole field switch and a discharge resistance, an ammeter, a handle for the rheostat in the generator field, and (if each alternator have its own direct coupled exciter) possibly also a small rheostat for the exciter field.

NOTE.—In some cases where the capacity of the plant is not very great, the oil switch is mounted on the back of the panel, and the bus bars, current transformers, &c., on the framework, also just at the back of the panel, but under no circumstances, in good modern practice, is high pressure apparatus permitted on the front of the board. Where the capacity of the plant is very large, the oil switches are operated electrically by means of small motors, and in this case the small switch gear for starting and stopping this motor is mounted on the generator panel, also the lamp or lamps to indicate when the switch is open, and when closed.

CHAPTER LXV
ALTERNATING CURRENT WIRING

In the case of alternating current, because of its peculiar behaviour, there are several effects which must be considered in making wiring calculations, which do not enter into the problem with direct current.

Accordingly, in determining the size of wires, allowance must be made for

1. Self-induction;
2. Mutual-induction;
3. Power factor;
4. Skin effect;
5. Corona effect;
6. Frequency;
7. Resistance.

Most of these items have already been explained at such length, that only a brief summary of facts need be added, to point out their connection and importance with alternating current wiring.

Induction.—The effect of induction, whether self-induction or mutual induction, is to set up a back pressure of spurious resistance, which must be considered, as it sometimes materially affects the calculation of circuits even in interior wiring.

Self-induction is the effect produced by the action of the electric current upon itself during variations in strength.

Ques. What conditions besides variations of current strength governs the amount of self-induction in a circuit?

Ans. The shape of the circuit, and the character of the surrounding medium.

If the circuit be straight, there will be little self-induction, but if coiled, the effect will become pronounced. If the surrounding medium be air, the self-induction is small, but if it be iron, the self-induction is considerable.

Figs. 2,671 to 2,676.—The effect of self-induction. In a non-inductive circuit, as in fig. 2,672, the whole of the virtual pressure is available to cause current to flow through the lamp filament, hence it will glow with maximum brilliancy. If an inductive coil be inserted in the circuit as in fig. 2,674, the reverse pressure due to self-induction will oppose the virtual pressure, hence the effective pressure (which is the difference between the virtual and reverse pressures), will be reduced and the current flow through the lamp diminished, thus reducing the brilliancy of the illumination. The effect may be intensified to such degree by interposing an iron core in the coil as in fig. 2,676, as to extinguish the lamp.

Ques. With respect to self-induction, what method should be followed in wiring?

Ans. When iron conduits are used, the wires of each circuit should not be installed in separate conduits, because such arrangement will cause excessive self-induction.

The importance of this may be seen from the experience of one contractor, who installed feeders and mains in separate iron pipes. When the current was turned on, it was found that the self-induction was so great as to reduce the pressure to such an extent that the lamps, instead of giving full candle power, were barely red. This necessitated the removal of the feeders and main and re-installing them, so that those of the same circuit were in the same pipe.

Ques. What is mutual induction?

Ans. Mutual induction is the effect of one alternating current circuit upon another.

Fig. 2,677.—Measurement of self induction when the frequency is known. The apparatus required consists of a high resistance or electrostatic a.c. voltmeter, d.c. ammeter, and a non-inductive resistance. Connect the inductive resistance to be measured as shown, and close switch M, short circuiting the ammeter. Connect alternator in circuit and measure drop across R and across X_i. Disconnect alternator and connect battery in circuit, then open switch M and vary the continuous current until the drop across R is the same as with the alternating current, both measurements being made with the same voltmeter; read ammeter, and measure drop across X_i. Call the drop across X_i with alternating current E, and with direct current E_i, and the reading of the ammeter J. Then L = √E² + E_i² ÷ 2π f I. If the resistance X_i be known, and the ammeter be suitable for use with alternating current, the switch and R may be dispensed with.

Then L = √E² - X_i² I_i² ÷ 2π f I, where I_i is the value of the alternating current. The resistance of the voltmeter should be high enough to render its current negligible as compared with that through X_i.

Ques. How is it caused?

Ans. It is due to the magnetic field surrounding a conductor cutting adjacent conductors and inducing back pressures therein.

This effect as a rule in ordinary installations is negligible.

Transpositions.—The effect of mutual induction between two circuits is proportional to the inter-linkage of the magnetic fluxes of the two lines. This in turn depends upon the proximity of the lines and upon the general relative arrangement of the conductors.

Fig. 2,678.—Transposition diagram for two parallel lines consisting of two wires each.

Fig. 2,679.—Transposition diagram for three phase, three wire line, transposing at the vertices of an equilateral triangle. The line is originally balanced and becomes unbalanced on transposing, a procedure which should be resorted to only to prevent mutual induction.

Fig. 2,680.—Transposition diagram of three phase, three wire line, with center arranged in a straight line.

The inductive effect of one line upon another is equal to the algebraic sum of the fluxes due to the different conductors of the first line, considered separately, which link the secondary line.

The effect of mutual induction is to induce surges in the line where a difference of frequency exists between the two currents, and to induce high electrostatic charges in lines carrying little or no current, such as telephone lines.

This effect may be nullified by separating the lines and by transposing the wires of one of the lines so that the effect produced in one section is opposed by that in another. Of two parallel lines consisting of two wires each, one may be transposed to neutralize the mutual inductance.

[Fig. 2,678] shows this method. The length L' should be an even factor of L so that to every section of the line transposed there corresponds an opposing section.

Fig. 2,681.—Capacity effect in single phase transmission line. The effect is the same as would be produced by shunting across the line at each point an infinitesimal condenser having a capacity equal to that of an infinitesimal length of circuit. For the purpose of calculating the charging current, a very simple and sufficiently accurate method is to determine the current taken by a condenser having a capacity equal to that of the entire line when charged to the pressure on the line at the generating end. The effect of capacity of the line is to reduce the pressure drop, that is, improve the regulation, and to decrease or increase the power loss depending on the load and power factor of the receiver.

Fig. 2,682.—Capacity effect in a three phase transmission line. It is the same as would be produced by shunting the line at each point by three infinitesimal condensers connected in star with the neutral point grounded, the capacity of each condenser being twice that of a condenser of infinitesimal length formed by any two of the wires. The effect of capacity on the regulation and efficiency of the line can be determined with sufficient accuracy in most cases by considering the line shunted at each end by three condensers connected in star, the capacity of each condenser being equal to that formed by any two wires of the line. An approximate value for the charging current per wire is the current required to charge a condenser, equal in capacity to that of any two of the wires, to the pressure at the generating end of the line between any one wire and the neutral point.

The self inductance of lines is readily calculated from the following formula:

L = .000558 {2.303 log (2A ÷ d) + .25} per mile of circuit

where

L = inductance of a loop of a three phase circuit in henrys.

Note.—The inductance of a complete single phase circuit = L × 2 ÷ √3.

A = distance between wires;
d = diameter of wire.

Capacity.—In any given system of electrical conductors, a pressure difference between two of them corresponds to the presence of a quantity of electricity on each. With the same charges, the difference of pressure may be varied by varying the geometrical arrangement and magnitudes and also by introducing various dielectrics. The constant connecting the charge and the resulting pressure is called the capacity of the system.

All circuits have a certain capacity, because each conductor acts like the plate of a condenser, and the insulating medium, acts as the dielectric. The capacity depends upon the insulation.

For a given grade of insulation, the capacity is proportional to the surface of the conductors, and universally to the distance between them.

A three phase three wire transmission line spaced at the corners of an equilateral triangle as regards capacity acts precisely as though the neutral line were situated at the center of the triangle.

The capacity of circuits is readily calculated by applying the following formulae:

in which

C = Capacity in micro-farads;
for a metallic circuit, C = capacity between wires;

sc = Specific inductive capacity of insulating material;
= 1 for air, and 2.25 to 3.7 for rubber;

D = Inside diameter of lead sheath;

d = Diameter of conductor;

h = Distance of conductors above ground;

A = Distance between wires.

Frequency.—The number of cycles per second, or the frequency, has a direct effect upon the inductance reactance in an alternating current circuit, as is plainly seen from the formula.

X_i = 2π f L

In the case of a transmission line alone; the lower frequencies are the more desirable, in that they tend to reduce the inductance drop and charging current. The inductance drop is proportional to the frequency.

The natural period of a line, with distributed inductance and capacity, is approximately given by

P = 7,900 / √LC

where L is the total inductance in millihenrys, and C, the total capacity in micro-farads. Accordingly some lower odd harmonic of the impressed frequency may be present which corresponds with the natural period of the line. When this obtains, oscillations of dangerous magnitude may occur. Such coincidences are less likely with the lower harmonics than with the higher.

Skin Effect.—The tendency of alternating current to confine itself to the outer portions of a conductor, instead of passing uniformly through the cross section, is called skin effect. The effect is proportional to the size of the conductor and the frequency.

Ques. What effect has "skin effect" on the current?

Ans. It is equivalent to an increase of ohmic resistance and therefore opposes the current.

Figs. 2,683 to 2,687.—Skin effect and shield effect. Fig. 2,683, section of conductor illustrating skin effect or tendency of the alternating current to distribute itself unequally through the cross section of a conductor as shown by the varied shading, which represents the current flowing most strongly in the outer portions of the conductor. For this reason it has been proposed to use hollow or flat instead of solid round conductors; however, with frequency not exceeding 100, the skin effect is negligibly small in copper conductors of the sizes usually employed. In figs. 2,684 and 2,685, or 2,686 and 2,687, if two adjacent conductors be carrying current in the same direction, concentration will occur on those parts of the two conductors remote from one another, and the nearer parts will have less current, that is to say, they will be shielded. In this case, the induction due to one conductor will exert its opposing effect to the greatest extent on those parts of the other conductor nearest to it; this effect decreasing the deeper the latter is penetrated. After crossing the current axis, the induction will still decrease in magnitude, but will now aid the current in the conductor. Hence, the effect of these two conductors on one another will make the current density more uniform than is the case where the two conductors adjacent to one another are carrying current in opposite directions, as in figs. 2,685 and 2,686, therefore, the resistance and the heating for a given current will be smaller. If the two return conductors be situated on the line passing through the center of the conductors just considered, the effect will be to still further concentrate the current; the distribution symmetry will be further disturbed, and the resistance of the conductor system increased. It is therefore difficult to say which of the two cases considered holds the advantage so far as increasing the resistance is concerned. The case, however, in which the phases are mixed has much the smaller reactive drop.

If the conductor be large, or the frequency high, the central portion of the conductor carries little if any current, hence the resistance is therefore greater for alternating current than for direct current.

Ques. For what condition may "skin effect" be neglected?

Ans. For frequencies of 60 or less, with conductors having a diameter not greater than 0000 B. & S. gauge.

Ques. How is the "skin effect" calculated for a given wire?

Ans. Its area in circular mils multiplied by the frequency, gives the ratio of the wire's ohmic resistance to its combined resistance.

That is to say, the factor thus obtained multiplied by the resistance of the wire to direct current will give its combined resistance or resistance to alternating current.

The following table gives these ratio factors for large conductors.

Corona Effect.—When two wires, having a great difference of pressure are placed near each other, a certain phenomenon occurs, which is called corona effect. When the spacing or distance between the wires is small and the difference of pressure in the wires very great, a continuous passage of energy takes place through the dielectric or atmosphere, the amount of this energy may be an appreciable percentage of the power transmitted. Therefore in laying out high pressure transmission lines, this effect must be considered in the spacing of the wires.

Ques. How does the corona effect manifest itself?

Ans. It is visible at night as a bluish luminous envelope and audible as a hissing sound.

Ques. What is the critical voltage?

Ans. The voltage at which the corona effect loss takes place.

Ques. Upon what does the critical voltage depend?

Ans. Upon the radius of the wires, the spacing, and the atmospheric conditions.

Fig. 2,688.—Electromagnetic and electrostatic fields surrounding the conductors of a transmission line. The electromagnetic field is represented by lines of magnetic force that surround the conductors in circles, (the solid lines), and the electrostatic field by (dotted) circles passing from conductor to conductor across at right angles to the magnetic circles. For any given size of wire and distance apart of wires there is a certain voltage at which the critical density or critical gradient is reached, where the air breaks down and luminosity begins—the critical voltage where corona manifests itself. At still higher voltages corona spreads to further distances from the conductor and a greater volume of air becomes luminous. Incidentally, it produces noise. Now to produce light requires power and to produce noise requires power. Air is broken down and is heated in breaking down, and to heat also requires power; therefore, as soon as corona forms, power is consumed or dissipated in its formation. When this phenomenon occurs on the conductors of an alternating current circuit a change takes place in relation to current and voltage. On the wires of an alternating current transmission line, at a voltage below that where corona forms—at a voltage where wires are not luminous—considerable current, more or less depending on voltage and length of wire, flows into the circuit as capacity current or charging current.

The critical voltage increases with both the diameter of the wires, and the spacing.

The losses due to corona effect increase very rapidly with increasing pressure beyond the critical voltage.

The magnitude of the losses as well as the critical voltage is affected, by atmospheric conditions, hence they probably vary with the particular locality, and the season of the year. Therefore, for a given locality, a voltage which is normally below the critical point, may at times be above it, depending upon changes in the weather.

Such elements as smoke, fog, moisture, or other particles (dust, snow, etc.) floating in the air, increase the losses; rain, however, apparently has no appreciable effect upon the losses. It follows then that in the design of a transmission line, the atmospheric conditions of the particular locality through which the line passes should be considered.

Ques. How should live wires be spaced?

Ans. They should be so spaced as to lessen the tendency to leakage and to prevent the wires swinging together or against towers. The spacing should be only sufficient for safety, since increased spacing increases the self-induction of the line, and while it lessens the capacity, it does so only in a slight degree.

The following spacing is in accordance with average practice.

Resistance of Wires.—For quick calculation the following method of obtaining the resistance (approximately) of wires will be found convenient:

1,000 feet No. 10 B. & S. wire, which is about .1 inch in diameter (.1019), has a resistance of one ohm, at a temperature of 68° F. and weighs 31.4 pounds. A wire three sizes larger, that is No. 7, has twice the cross section and therefore one-half the resistance. A wire three sizes smaller than No. 10, that is No. 13, has one-half the cross section and therefore twice the resistance.

Thus, starting with No. 10, any number three sizes larger will double the cross sectional area and any wire three sizes smaller will halve the cross sectional area of the preceding wire. This is true to the extreme limits of the table, so that the area, weight and resistance of any wire may be at once calculated to a close approximation from this rule, intermediate sizes being obtained by interpolation.

For alternating current, the combined resistance, that is, the total resistance, including skin effect, is obtained by multiplying the resistance, as found above by the "ratio factor" ([see table page 1,894]).

Figs. 2,689 to 2,692.—Triangles for obtaining graphically, impedance, impressed pressure, etc., in alternating current circuits. For a full explanation of this method the reader is referred to Guide 5, Chapter XLVII on Alternating Current Diagrams. A thorough study of this chapter is recommended.

Impedance.—The total opposition to the flow of electricity in an alternating current circuit, or the impedance may be resolved into two components representing the ohmic resistance and the spurious resistance; these components have a phase difference of 90°, and they may be represented graphically by the two legs of a right angle triangle, of which the hypothenuse represents the impedance.

Similarly, the volts lost or "drop" in an alternating circuit may be resolved into two components representing respectively
1. The loss due to resistance.
2. The loss due to reactance.

These components have a phase difference of 90° and are represented graphically similar to the impedance components. This has been explained at considerable length in Chapter XLVII (Guide V).

Fig. 2,693.—Mechanical analogy of power factor, as exemplified by a locomotive "poling" a car off a siding. The car and locomotive are shown moving in parallel directions, and the pole AB, inclined at an angle ϕ. Now, if the length of AB be taken to represent the pressure exerted on the pole by the locomotive, then the imaginary lines AC and BC, drawn respectively parallel and at right angles to the direction of motion will represent respectively the useful and no energy (wattless) components; that is to say, if the pressure AB be applied to the car at an angle ϕ, only part of it, AC, is useful in propelling the car, the other component, BC, being wasted in tending to push the car off the track at right angles to the rails, being resisted by the flanges of the outer wheels.

Power Factor.—When the current falls out of step with the pressure, as on inductive loads, the power factor becomes less than unity, and the effect is to increase the current required for a given load. Accordingly, this must be considered in calculating the size of the wires. As has been explained, the current flowing in an alternating current circuit, as measured by an ammeter, can be resolved into two components, representing respectively the active component and the wattless component or idle current. These are graphically represented by the two legs of a right triangle, of which the hypothenuse represents the current measured by the ammeter.

This apparent current, as is evident from the triangle, exceeds the active current and lags behind the pressure by an amount represented by the angle ϕ between the hypothenuse and leg representing the energy current as shown in [fig. 2,694].

Fig. 2,694.—Diagram showing that the apparent current is more than the active current, the excess depending upon the angle of phase difference.

Fig. 2,695.—Diagram showing components of impedance volts. Compare this diagram with [figs. 2,689] and [2,671], and note that the term "reactance" is the difference between the inductance drop and the capacity drop if the circuit contain capacity, for instance, if inductance drop be 10 volts and capacity drop be 7 volts then reactance 10-7 = 3 volts.

Ques. What determines the heating of the wires on alternating current circuits with inductive loads?

Ans. The apparent current, as represented by the hypothenuse of the triangle in [fig. 2,694].

Ques. How is the apparent current obtained?

Ans. Divide the true watts by the product of the power factor multiplied by the voltage.

Example.—A certain circuit supplies 20 kw. to motors at 220 volts and .8 power factor. What is the apparent current?

Ques. What else, besides power factor, should be considered in making wire calculations for motor circuits?

Ans. The efficiency of the motor, and the heavy starting current.

The product of the efficiency of the motor multiplied by the power factor gives the apparent efficiency, which governs the size of the wires, apparatus, etc., necessary to feed the motors.

Allowance should be made for the heavy starting current required for some motors to avoid undue drop.

Ques. What are the usual power factors encountered on commercial circuits?

Ans. A mixed load of incandescent lamps and induction motors will have a power factor of from .8 to .85; induction motors above .8 to .85; incandescent and Nernst lamps .98; arc lamps, .85.

Wire Calculations.—In the calculation of alternating current circuits, the two chief factors which make the computation different from that for direct current circuits, is induction and power factor. The first depends on the frequency, and physical condition of the circuit, and the second upon the character of the load.

Ques. Under what conditions may inductance be neglected?

Figs. 2,696 to 2,698.—Example of wiring showing where inductance is negligible, and where it must be considered in wire calculations.

Ans. In cases where the wires of a circuit are not spaced over an inch apart, or in conduit work, where both wires are in the same conduit.

Under these conditions the calculation is the same as for direct current after making proper allowance for power factor.

Ques. Under what conditions must induction be considered?

Ans. On exposed circuits with wires separated several inches, particularly in the case of large wires.

Sizes of Wire.—The size of wire for any alternating circuit may be determined by slightly modifying the formula used in direct current work, and which, as derived in Guide No. 4, page 748, is

The quantity 21.6, is twice the resistance (10.8) of a foot of copper wire one mil in diameter (mil foot). This resistance (10.8) is multiplied by 2, giving the quantity 21.6, because the length of a circuit, or feet in the formula, is given as the "run" or distance one way, that is, one-half the total length of wire in the circuit, must be multiplied by 2 to get the total drop, viz.:

It is sometimes however convenient to make the calculation in terms of watts. Formula (1) may be modified for such calculation.

In modifying the formula, the "drop" should be expressed in percentage instead of actual volts lost, that is, instead of the difference in pressure between the beginning and the end of the circuit.

In any circuit the loss in percentage, or

from which

Substituting equation (2) in equation (1)

Equation (3) is modified for calculation in terms of watts as follows: The power in watts is equal to the applied voltage multiplied by the current, that is to say, the power is equal to the volts at the consumer's end of the circuit multiplied by the current, or simply

watts = volts × amperes

from which

Figs. 2,699 to 2,703.—Stranded copper cables. For conductors of large areas and in the smaller sizes where extra flexibility is required it becomes necessary to employ stranded cables made by grouping a number of wires together in either concentric or rope form. The concentric cable as here illustrated is formed by grouping six wires around a central wire thereby forming a seven wire cable. The next step is the application in a reverse direction of another layer of 12 wires and a nineteen wire cable is produced. This is again increased by a third layer of eighteen wires for a 37 wire cable and a fourth layer of 24 wires for a 61 wire cable. Successive layers, each containing 6 more wires than that preceding, may be applied until the desired capacity is obtained. The cuts show sectional views of concentric cables each formed from No. 10 B. & S. gauge wires.

Substituting this value for the current in equation (3) and remembering that the pressure taken is the volts at the consumer's end of the line

This formula (5) applies to a direct current two wire circuit, and to adapt it to any alternating current circuit it is only necessary to use the letter M instead of the number 2,160, thus

in which M is a coefficient which has various values according to the kind of circuit and value of the power factor. These values are given in the following table:

NOTE.—The above table is calculated as follows: For single phase M = 2,160 ÷ power factor² × 100; for two phase four wire, or three phase three wire, M = ½ (2,160 ÷ power factor²)× 100. Thus the value of M for a single phase line with power factor .95 = 2,160 ÷ .95² × 100 = 2,400.

It must be evident that when 2,160 is taken as the value of M, formula (6) applies to a two wire direct current circuit and also to a single phase alternating current circuit when the power factor is unity.

In the table the value of M for any particular power factor is found by dividing 2,160 by the square of that power factor for single phase and twice the square of the power factor for two phase and three phase.

Ques. For a given load and voltage how do the wires of a single and two phase system compare in size and weight, the power factor being the same in each case?

Ans. Since the two phase system is virtually two single phase systems, the four wires of the two phase systems are half the size of the two wires of the single phase system, and accordingly, the weight is the same for either system.

NOTE.—This table is for finding the value of the current in line, using the formula I = W ÷ (E × T), in which I = current in line; E = voltage between main conductors at receiving or consumers' end; W = watts. For instance, what is the current in a two phase line transmitting 1,000 watts at 550 volts, power factor .80? I = 1,000 ÷ (550 × 1.60) = 1.13.

Ques. Since there is no saving in copper in using two phases, what advantage has the two phase system over the one phase system?

Ans. It is more desirable on power circuits, because two phase motors are self-starting.

That is to say, the rotating magnetic field that can be produced by a two phase current, permits an induction motor to start without being equipped with any special phase splitting devices which are necessary on single phase motors, because the oscillating field produced by a single phase current does not produce any torque on a squirrel cage armature at rest.

Ques. For equal working conditions, what is the comparison between the single, two and three phase system as to size and weight of wires?

Ans. Each wire of the three phase system is half the size of one of the wires of the single phase system, hence the weight of copper required for the three phase system is 75% of that required for the single phase system. Since in the two phase system half of the load is carried by each phase, each wire of the three phase system is the same size as one of the wires of the two phase system, hence, the copper required by the three phase system is 75% of that required by the two phase system.

Breaking weight of wire ÷ area = breaking weight per square inch.

Breaking weight per square inch × area = breaking weight of wire.

The weight of copper wire is 1-1/7 times the weight of iron wire of same diameter.

EXAMPLE.—What size wires must be used on a single phase circuit 2,000 feet in length to supply 30 kw. at 220 volts with loss of 4%, the power factor being .9?

The formula for circular mils is

Substituting the given values and the proper value of M from the table, in (1)

Referring to the accompanying table of the properties of copper wire, the nearest larger size wire is No. 1 B. & S. gauge having an area of 83,690 circular mils.

TABLE OF THE PROPERTIES OF COPPER WIRE

Giving weights, length and resistances of wires of Matthiessen's Standard Conductivity for both B. & S. G. (Brown & Sharpe Gauge) and B. W. G. (Birmingham Wire Gauge) from Transactions October 1903, of the American Institute of Electrical Engineers.

Drop.—In order to determine the drop or volts lost in the line, the following formula may be used

in which the % loss is a percentage of the applied power, that is, the power delivered to the consumer and not a percentage of the power at the alternator. "Volts" is the pressure at the consumer's end of the circuit.

The coefficient S has various values as given in the accompanying tables. As will be seen from the table, the value of S to be used depends upon the size of wire, spacing, power factor and frequency.

These values are accurate enough for all practical purposes, and may be used for distances of 20 miles or less and for voltages up to 25,000.

The capacity effect on very long high voltage lines, makes this method of determining the drop somewhat inaccurate beyond the limits above mentioned.

EXAMPLE.—A circuit supplying current at 440 volts, 60 frequency, with 5% loss and .8 power factor is composed of No. 2 B. & S. gauge wires spaced one foot apart. What is the drop in the line?

According to the formula

Substituting the given values, and value of S as obtained from the table for frequency 60

Current.—As has been stated, the effect of power factor less than unity, is to increase the current; hence, in inductive circuit calculations, the first step is to determine the current flowing in a circuit. This is done as follows:

and

Substituting (2) in (1)

Fig. 2,704.—Rope type of stranded copper cable which is used when a high degree of flexibility is required. The construction of this cable is the stranding together of seven groups, each containing seven wires and producing a total of 49 wires. In cases when a greater carrying capacity is desired than can be obtained through the use of the 7 × 7 or 49 wire cable, the number of groups is increased to nineteen thereby making a total of 133 wires (19 × 7).

EXAMPLE.—A 50 horse power 440 volt motor has a full load efficiency of .9 and power factor of .8. How much current is required?

Since the brake horse power of the motor is given, it is necessary to obtain the electrical horse power, thus

which in watts is

55.5 × 746 = 41,403

which is the actual load, and from which

The current therefore at 440 volts is

EXAMPLE.—A 50 horse power single phase 440 volt motor, having a full load efficiency of .92 and power factor of .8, is to be operated at a distance of 1,000 feet from the alternator. The wires are to be spaced 6 inches apart and the frequency is 60, and % loss 5. Determine: A, electrical horse power; B, watts; C, apparent load; D, current; E, size of wires; F, drop; G, voltage at the alternator.

A. Electrical horse power

or,

54.3 × 746 = 40,508 watts

B. Watts

watts = E.H.P. × 746 = 54.3 × 746 = 40,508

C. Apparent load

D. Current

E. Size of wires

From table page 1,907, nearest size larger wire is No. 00 B. & S. gauge.

F. Drop

NOTE.—Values of S are given on page 1910.

G. Voltage at alternator

alternator pressure = (volts at motor + drop) = 440 + 25.74 = 465.7 volts.

CHAPTER LXVI
POWER STATIONS

The term power station is usually applied to any building containing an installation of machinery for the conversion of energy from one form into another form. There are three general classes of station:

1. Central stations;
2. Sub-stations;
3. Isolated plants.

These may also be classified with respect to their function as

1. Generating stations;
2. Distributing stations;
3. Converting stations.

and with respect to the form of power used in generating the electric current, generating stations may be classed as

1. Steam electric;
2. Hydro-electric;
3. Gas electric, etc.

Central Stations.—It must be evident that the general type of central station to be adapted to a given case, that is to say, the general character of the machinery to be installed depends upon the kind of natural energy available for conversion into electrical energy, and the character of the electrical energy required by the consumers.

This gives rise to a further classification, as

1. Alternating current stations;
2. Direct current stations;
3. Alternating and direct current stations.

The alternators or dynamos may be driven by steam or water turbines, reciprocating engines, or gas engines, according to the character of the natural energy available.

Fig. 2,705.—Elevation of small station with direct drive, showing arrangement of the boiler and engine, piping, etc.

Ques. Why is the reciprocating engine being largely replaced by the steam turbine, especially for large units?

Ans. Because of its higher rotative speed, and absence of a multiplicity of bearings which in the case of a high speed, reciprocating engine must be maintained in close adjustment for the proper operation of the engine.

The higher speed of rotation results in a more compact unit, desirable for driving high frequency alternators.

Ques. Is the steam turbine more economical than a high duty reciprocating engine?

Ans. No.

Location of Central Stations.—As a rule, central stations should be so located that the average loss of voltage in overcoming the resistance of the lines is a minimum, and this point is located at the center of gravity of the system. In [fig. 2,706] is shown a graphical method of locating this important spot.

Fig. 2,706.—Diagram illustrating graphical method of determining the center of gravity of a system in locating the central station.

Suppose a rough canvass of prospective consumers in a district to be supplied with electric light or power shows the principal loads to be located at A, B, C, D, E, etc., and for simplicity assume that these loads will be approximately equal, so that each may be denoted by 1 for example:

The relative locations of A, B, C, D, E, etc., should be drawn to scale (say 1 inch to the 1,000 feet) after which the problem resolves itself into finding the location of the station with respect to this scale.

Fig. 2,707.—Exterior of central station at Lewis, Ia.; example of very small station located in the principal business section of a town. It also illustrates the use of a direct connected gasoline electric set. The central station is located on Main Street, which is the principal thoroughfare, and is installed in a low one story building for which a mere nominal rental charge is paid, the company having the option to buy the property later at the value of the land plus the cost of the improvements and simple interest on the same. To the front of an old frame building about 16 feet by 28 feet has been built a neat, well lighted concrete block room, about 16 feet by 16 feet, carrying the building to the lot line and affording ample space for the generating set and switchboards, and such desk room as is needed for the ordinary office business of the company. In this room, which is finished in natural pine with plastered walls, has been installed a standard General Electric 25 kw. gasoline electric generating set consisting of a four cylinder, four cycle, vertical water cooled, 43-54 H.P. gasoline engine, direct connected to a three phase, 2,300 volt, 600 R.P.M. alternator with a 125 volt exciter mounted on the same shaft and in the same frame. With the generating set is a slate switchboard panel equipped with three ammeters, one voltmeter, an instrument plug switch for voltage indication, one single pole carbon break switch, one automatic oil circuit breaker line switch and rheostats. Instrument transformers are mounted above and back of the board. For street lighting service a 4 kw. constant current transformer has been installed, and with it a gray marble switchboard panel mounted on iron frames and carrying an ammeter and a four point plug switch. On a board near the generator set are mounted in convenient reach suitable wrenches, spanners, and repair parts and tools. To cool the engine cylinders five 6 × 8 steel tanks have been installed in the old building, a pump on engine giving forced circulation.

The solution consists in first finding the center of gravity of any two of the loads, such as those at A and B. Since each of these is 1, they will together have the same effect on the system as the resultant load of 1 and 1, or 2, located at their center of gravity, this point being so chosen that the product of the loads by their respective distances from this point will in both cases be equal.

The loads being equal in this case the distances must be equal in order that the products be the same, so that the center of gravity of A + B is at G, which point is midway between A and B.

Considering, next, the resultant load of 2 at G and the load of 1 at C, the resultant load at the center or gravity of these will be 3, and this must be situated at a distance of two units from C and one unit from G so that the distance 2 times the load 1 at C equals the distance 1 times the load 2 at G. Having thus located the load 3 at H, the same method is followed in finding the load 4 at I. Then in like manner the resultant load 4 and the load 1 at E gives a load 5 at S.

The point S being the last to be determined represents, therefore, the position of the center of gravity of the entire system, and consequently the proper position of the plant in order to give the minimum loss of voltage on the lines.

Ques. Is the center of gravity of the system, as obtained in [fig. 2,706], the proper location for the central station?

Ans. It is very rarely the best location.

Ques. Why?

Ans. Other conditions, such as the price of land, difficulty of obtaining water, facilities for delivery of coal and removal of ashes, etc., may more than offset the minimum line losses and copper cost due to locating the station at the center of gravity of the system.

Fig. 2,708.—Map of Cia Docas de Santos hydro-electric system; an example of station location remote from the center of distribution. In the figure A is the intake; B, flume; C, forebay; D, penstocks; E, power house; F, narrow gauge railway; G, general store; H, point of debarkation; I, transmission line; J, dead ends; K, sub-station. Santos, in the republic of Brazil, is one of the great coffee shipping ports of the world, and for the development of its water front has required an elaborate system of quays. These have been developed by the Santos Dock Company, which holds a concession for the whole water front. The company, needing electric power for its own use, has developed a system deriving its power from a point about thirty miles from the city, where a small stream plunges down the sea coast from the mountain range that runs along it. The engineers have estimated that 100,000 horse power can be obtained from this source.

Ques. How then should the station be located?

Ans. The more practical experience the designer has had, and the more common sense he possesses, the better is he equipped to handle the problem, as the solution is generally such that it cannot be worked out by any rule of thumb method.

Fig. 2,709.—Station location. The figure shows two distribution centers as a town A and suburb B supplied with electricity from one station. For minimum cost of copper the location of the station would be at G, the center of gravity. However, it is very rarely that this is the best location. For instance at C, land is cheaper than at G, and there is room for future extension to the station, as shown by the dotted lines, whereas at G, only enough land is available for present requirements. Moreover C is near the railroad where coal may be obtained without the expense of cartage, and being located at the river, the plant may be run condensing thus effecting considerable economy. The conditions may sometimes be such that any one of the advantages to be secured by locating the station at C may more than offset the additional cost of copper.

Ques. What are the general considerations with respect to the price of land?

Ans. The cost for the station site may be so high as to necessitate building or renting room at a considerable distance from the district to be supplied.

If the price of land selected for the station be high, the running expenses will be similarly affected, inasmuch as more interest must then be paid on the capital invested.

The price or rent of real estate might also in certain instances alter the proposed interior arrangement of the station, particularly so in the case of a company with small capital operating in a city where high prices prevail. In general, however, it may be stated that whatever effect the price of real estate would have upon the arrangement, operation and location of a central station it can quite readily and accurately be determined in advance.

Ques. With respect to the cost of the land what should be especially considered?

Ans. Room for the future extension of the plant.

Although such additional space need not be purchased at the time of the original installation it is well, if possible, to make provision whereby it can be obtained at a reasonable figure when desired. The preliminary canvass of consumers will aid in deciding the amount of space advisable to allow for future extensions; as a rule, however, it is wise to count on the plant enlarging to not less than twice its original size, as often the dimensions have to be increased four and even six times those found sufficient at the beginning.

Fig. 2,710.—Section of the central station or "electricity works" at Derby, showing boiler and engine room and arrangement of bunkers, conveyor, ash pit, grates, boilers (drum, heating surface and superheater), economizer, flue, turbines, condenser pumps, etc.; also location of switchboard gallery and system of piping.

Ques. What trouble is likely to be encountered with an illy located plant after it is in operation?

Ans. It may be considered a nuisance by those residing in the vicinity, occasioning many complaints.

Fig. 2,711.—View of old and new Waterside stations. The new station at the right has an all turbine equipment of ten units, some Curtis and some Parsons machines, two have a capacity of 14,000 kw., and the remaining eight are of 12,000 kw. each. The old Riverside station, seen at the left is described on page 1940.

Thus, if the plant be placed in a residential section of the community the smoke, noise and vibration of the machines may become a nuisance to the surrounding inhabitants, and eventually end in suits for damage against the company responsible for the same. For these and the other reasons just given a company is sometimes forced to disregard entirely the location of a central station near the center of gravity of the system, and build at a considerable distance; such a proceeding would, if the distance be great, necessitate the installation of a high pressure system.

There might, however, be certain local laws in force restricting the use of high pressure currents on account of the danger resulting to life, that would prevent this solution of the problem. In such cases there could undoubtedly be found some site where the objections previously noted would be tolerated; thus, there would naturally be little objection to locating next to a stable, a brewery, or a factory of any description.

Ques. Why is the matter of water supply important for a central station?

Ans. Because, in a steam driven plant, water is used in the boilers for the production of steam by boiling, and if the engines be of the condensing type it is also used in them for creating a vacuum into which the exhaust steam passes so as to increase the efficiency of the engine above what it would be if the exhaust steam were obliged to discharge into the comparatively high pressure of the atmosphere.

The force of this will be apparent by considering that the water consumption of the engine ordinarily is from 15 to 25 lbs. of "feed water" per horse power per hour, and the amount of "circulating water" required to maintain the vacuum is about 25 to 30 times the feed water, and in the case of turbines with their 28 or 29 inch vacuum, much more. For instance, a 1,000 horse power plant running on 15 lbs. of feed water and 30 to 1 circulating water would require (1,000 × 15) × (30 + 1) = 465,000 lbs. or 55,822 gals. per hour at full capacity.

Ques. Besides price what other considerations are important with respect to water?

Ans. Its quality and the possibility of a scarcity of supply.

It is quite necessary that the water used in the boilers should be as free as possible from impurities, so as to prevent the deposition within them of any scale or sediments. The quality of the water used for condensing purposes, however, is not quite so important, although the purer it is the better.

If the plant is to be located in a city, the matter of water supply need not generally be considered, because, as a rule, it can be obtained from the waterworks; there will then, of course, be a water tax to consider and this, if large, may warrant an effort being made to obtain the water in some other way. In any event, however, the possibility of a scarcity in the supply should be reduced to a minimum.

If the plant be located in the country, some natural source of water would be utilized unless the place be supplied with waterworks, which is not generally the case. It is usual, however, to find a stream, lake or pond in the vicinity, but if none such be conveniently near, an artesian or other form of well must be sunk.

If abundance of water exist in the vicinity of the proposed installation, not only would the location of the plant be governed thereby, but the kind of power to be used for its operation would depend thereon. Thus, if the quantity of the water were sufficient throughout the entire year to supply the necessary power, water wheels might be installed and used in place of boilers and steam engines for driving the generators. The station would then, of course, be situated close to the waterfall, regardless of the center of gravity of the system.

Fig. 2,712.—View illustrating the location of a station as governed by the presence of a water falls. In such cases the natural water power may be at a considerable distance from the center of gravity of the distribution system because of the saving in generation. In the case of long distance transmission very high pressure may be used and a transformer step down sub-station be located at or near the center of gravity of the system, thus considerably reducing the cost of copper for the transmission line.

Ques. What should be noted with respect to the coal supply?

Ans. The facility for transporting the coal from the supply point to the boiler room.

In this connection, an admirable location, other conditions permitting, is adjacent to a railway line or water front so that coal delivered by car or boat may be unloaded directly into the bins supplying the boilers.

If the coal be brought by train, a side or branch track will usually be found convenient, and this will usually render any carting of the fuel entirely unnecessary.

In whatever way the coal is to be supplied, the liability of a shortage due to traffic or navigation being closed at any time of the year should be well looked into, as should also the facility for the removal of ashes, before deciding upon the final location for the plant.

Fig. 2,713.—View of a station admirably located with respect to transportation of the coal supply. As shown, the coal may be obtained either by boat or rail, and with modern machinery for conveying the coal to the interior of the station, the transportation cost is reduced to a minimum.

Fig. 2,714.—Floor plan of part of the turbine central station erected by the Boston Edison Co., showing two 5,000 kw. Curtis steam turbines in place. The complete installation contains twelve 5,000 kw. Curtis steam turbines, a sectional elevation being shown in [fig. 2,758, page 1,971].

Choice of System.—The chief considerations in the design of a central station are economy and capacity. When the current has to be transmitted long distances for either lighting or power purposes, economy is attainable only by reducing the weight of the copper conductors. This can be accomplished only by the use of the high voltage currents obtainable from alternators.

Again, where the consumers are located within a radius of two miles from the central station, thereby requiring a transmission voltage of 550 volts or less, dynamos may be employed with greater economy.

Alternating current possesses serious disadvantages for certain important applications.

For instance, in operating electric railways and for lighting it is often necessary to transmit direct current at 500 volts a distance of five or ten miles. In such cases, the excessive drop cannot be economically reduced by increasing the sizes of the line wire, while a sufficient increase of the voltage would cause serious variations under changes of load. Hence, it is common practice to employ some form of auxiliary generator or booster, which when connected in series with the feeder, automatically maintains the required pressure in the most remote districts so long as the main generators continue to furnish the normal or working voltage.

The advantage of a direct current installation in such cases over a similar plant supplying alternating current line is the fact that a storage battery may be used in connection with the former for taking up the fluctuations of the current, thereby permitting the dynamo to run with a less variable load, and consequently at higher efficiency.

Ques. Name some services requiring direct current.

Ans. Direct current is required for certain kinds of electrolytic work, such as electro-plating, the electrical separation of metals, etc., also the charging of storage batteries for electric automobiles.

Fig. 2,715.—Example of central station located remote from the distributing center and furnishing alternating current at high pressure to a sub-station where the current is passed through step down transformers and supplied at moderate pressure to the distribution system. In some cases the sub-station contains also converters supplying direct current for battery charging, electro-plating, etc.

Ques. How is direct current supplied?

Ans. Sometimes the central station is equipped with suitable apparatus for supplying both direct and alternating current. This may be accomplished in several different ways: By installing both direct and alternating current generators in the central station; by the use of double current generators or dynamotors, from which direct current may be taken from one side and alternating current from the other side; or by installing, in the sub-station of an alternating current central station, in addition to the transformers usually placed therein, a rotary converter for changing or converting alternating current into direct current.

Thus, it is evident that the character of a central station will be governed to a great extent by the class of services to be supplied.

An exception to this is where the entire output has to be transmitted a long distance to the point of utilization.

In such cases a copper economy demands the use of high tension alternating current, and its distribution to consumers may be made directly by means of step down transformers mounted near by or within the consumers' premises, or it may be transformed into low voltage alternating current by a conveniently located sub-station.

Where the current is to be used chiefly for lighting and there are only a few or no motors to be supplied, the choice between direct current and alternating current will depend greatly upon the size of the installation, direct current being preferable for small installations and alternating current for large installations.

If the current is to be used primarily for operating machinery, such as elevators, travelling cranes, machine tools and other devices of a similar character, which have to be operated intermittently and at varying speeds and loads, direct current is the more suitable; but if the motors performing such work can be operated continuously for many hours at a time under practically constant loads, as, for instance in the general work of a pumping station, alternating current may be employed with advantage.

Fig. 2,716.—Diagram illustrating diversity factor. By definition diversity factor = combined actual maximum demand of a group of customers divided by the sum of their individual maximum demands. Example, a customer has fifty (50) watt lamps and, of course, the sum of the individual maximum demands of the lamps is 2.5 kw. watts ("connected load"). The customer's maximum demand, however, is 1.5 kw. Hence, the diversity factor[A] of the customer's group of lamps is 1.5 ÷ 2.5 = .6. In the diagram the ordinates of the curves show the ratio maximum demand to connected load for various kinds of electric lighting service in Chicago.

[ [A] NOTE.—The diversity factor of a customer's group of lamps, namely, the ratio of maximum demand to connected load is usually called the demand factor of the customer.

Size of Plant.—Before any definite calculation can be made, or plans drawn, the engineer must determine the probable load. This is usually ascertained in terms of the number and distances of lamps that will be required, by making a thorough canvass of the city or town, or that portion for which electrical energy is to be supplied. The probable load that the station is to carry when it begins operation, the nature of this load, and the probable rate of increase are matters upon which the design and construction chiefly depend.

Fig. 2,717.—Load curve for one day.

Ques. What is the nature of the load carried by a central station?

Ans. It fluctuates with the time of day and also with the time of year.

Ques. How is a fluctuating load best represented?

Ans. Graphically, that is to say by means of a curve plotted on coordinate paper of which ordinates represent load values and the corresponding abscissæ time values, as in the accompanying curves.

What is the nature of a power load?

Ans. Where electricity is supplied for power purposes to a number of factories, the load is fairly steady, dropping, of course, during meal hours. In the case of traction, the average value of the load is fairly steady but there are momentarily violent fluctuations due to starting cars or trains.

Fig. 2,718.—Load curve for one year.

Ques. What is the peak load?

Ans. The maximum load which has to be carried by the station at any time of day or night as shown by the highest point of the load curve.

Ques. Define the load factor.

Ans. The machinery of the station evidently must be large enough to carry the peak load, and therefore considerably in excess of that required for the average demand. The ratio of the average to the maximum load is called the load factor.

There are two kinds of load factor: the annual, and the daily.

The annual load factor is obtained as a percentage by multiplying the number of units sold (per year) by 100, and dividing by the product of the maximum load and the number of hours in the year. The daily load factor is obtained by taking the figures for 24 hours instead of a year.

Fig. 2,719.—Load curve of plant supplying power for the operation of motors in a manufacturing district. The horizontal dotted lines show suitable power ratings. A properly designed steam plant has a large overload capacity, a hydraulic plant has a small overload capacity, and a gasoline engine plant has no overload capacity. Accordingly, the peak of the load (maximum load) may be 25 or 30 per cent. in excess of the rated capacity of a steam plant, not more than 5 or 10 per cent. in excess of the rated capacity of a hydraulic plant, not at all in excess of the rated capacity of a gas engine plant.

Ques. What must be provided in addition to the machinery required to supply the peak load?

Ans. Additional units must be installed for use in case of repairs or break down of some of the other units.

EXAMPLE.—What would be the boiler horse power required to generate 5,000 kw. under the following conditions: Efficiency of generators 85%; efficiency of engines 90%; feed water of engines and auxiliaries 15 lbs. per I. H. P.; boiler pressure 175 lbs.; temperature of feed water 150° Fahr? With a rate of combustion of 15 lbs. of coal per sq. foot of grate per hour and an evaporation (from and at 212°) of 8 lbs. of water per lb. of coal, what area of grate would be required and how much heating surface?

5,000 kw. = 5,000 ÷ .746 = 6,702 electrical horse power

To obtain this electrical horse power with alternators whose efficiency is 85% requires

6,702 ÷ .85 = 7,885 brake horse power at the engine

This, with mechanical efficiency of 90% is equivalent to

7,885 ÷ .9 = 8,761 indicated horse power

Since 15 lbs. of feed water are required for the engines and auxiliaries per indicated horse power per hour, the total feed water or evaporation required to generate 5,000 kw. is

15 × 8,761 = 131,415 lbs. per hour.

that is to say, the boilers must be of sufficient capacity to generate 131,415 lbs. of steam per hour from water at a temperature of 150° Fahr. This must be multiplied by the factor of evaporation for steam at 175 lbs. pressure from feed water at a temperature of 150°, in order to get the equivalent evaporation "from and at 212°."

The formula for the factor of evaporation is

in which
H = total heat of steam at the observed pressure;
h = total heat of feed water of the observed temperature;
965.7 = latent heat, of steam at atmospheric pressure.

Substituting in (1) values for the observed pressure and temperature as obtained from the steam table

for which the equivalent evaporation "from and at 212°" is

131,415 × 1.117 = 146,791 lbs. per hour

One boiler horse power being equal to an evaporation of 34½ lbs. of water from a feed water temperature of 212° Fahr., into steam at the same temperature, the boiler capacity is accordingly

148,105 ÷ 34.5 = 4,293 boiler horse power.

The rate of evaporation is given at 8 lbs. of water (from and at 212° Fahr.), for which the fuel required is

148,105 ÷ 8 = 18,513 lbs. of coal per hour.

For a rate of combustion of 15 lbs. of coal per hour per square foot of grate,

grate area = 18,513 ÷ 15 = 1,234 sq. ft.

For stationary boilers the usual ratio of heating surface to grate area is 35:1, accordingly the heating surface corresponding to this ratio is

1,234 × 35 = 43,190 sq.ft.

The above calculation is based on a rate of evaporation of 8 lbs. of water per lb. of coal and a rate of combustion of 15 lbs. of coal per sq. ft. of grate. For other rates the required grate area may be obtained from the following table:

General Arrangement of Station.—In designing an electrical station, it is preferable that whatever rooms or divisions of the interior space are desired should determine the total outside dimensions of the plant in the original plans of the building than that these latter dimensions be fixed and the rooms, etc., be fitted in afterward.

NOTE.—An approximate rule for the conditions of ordinary practice is a saving of 1 per cent. made by each increase of 11° in the temperature of the feed water. This corresponds to .0909 per cent. per degree. The calculation of saving is made as follows: Boiler pressure, 100 lbs. gauge; total heat in steam above 32° = 1,185 B.T.U. feed water, original temperature 60°, final temperature 209°F. Increase in heat units, 150. Heat units above 32° in feed water of original temperature = 28. Heat units in steam above that in cold feed water, 1,185-28 = 1,157. Saving by the feed water heater = 150 ÷ 1,157 = 12.96 per cent. The same result is obtained by the use of the table. Increase in temperature 150° × tabular figure .0864 = 12.96 per cent. Let total heat of 1 lb. of steam at the boiler pressure = H; total heat of 1 lb. of feed water before entering the heater = h', and after passing through the heater = h"; then the saving made by the heater is (h"-h') ÷ (H-h').

Under usual conditions the plans of an electrical station are readily drawn, as they are generally of a simple nature. The engines and generators will occupy the majority of the space, and these are usually placed in one large room; in some stations, however, they are located respectively in two adjacent rooms. The boilers are generally located in a room apart from the engines and dynamos, and in some cases a separate building is provided for them; the pumps, etc., must be installed not far from the boilers, and space must also be allowed near the boilers for coal and ashes.

Fig. 2,720.—Floor plan of an electrical station having a belted drive with counter shaft.

Fig. 2,720 shows the floor plan of an electrical station, in which a countershaft and belted connections are used between the engines and generators. Referring first to the plan of the building itself, A represents the engine and dynamo room, B denotes the boiler room, C the office, D the store room, and E the chimney connected with the boilers by means of the uptake w. Referring next to the apparatus installed, S, S, S, S represents a battery of four boilers; these are connected by steam piping VV to the two steam engines, M and M, which are belted to the countershaft O. Belted to the countershaft are the generators, T, T, T, T, the circuits from which are controlled on the switchboard, H.

Ques. What are the objections to the arrangement shown in [fig. 2,720].?

Ans. The large space required by the belt drive especially in locations where land is expensive. Another objection is the frictional loss due to the belt drive with its countershaft, etc.

Fig. 2,721.—Elevation of station having a belted drive with countershaft, as shown in plan in [fig. 2,720].

Ques. What are the desirable features of the belt drive?

Ans. High speed generators may be used, thus reducing the first cost, and the multiplicity of speeds and flexibility of the system resulting from the use of a friction clutch.

Thus in [fig. 2,720], each pulley may be mounted on the counter shaft O with a friction clutch. A jaw clutch may also be provided at Z, thus permitting the shaft O to be divided into two sections. It is therefore possible by this arrangement to cause either of the engines to drive any one of the generators, or all of them, or both of the engines to drive all of the generators simultaneously.

Ques. Under what condition is the counter shaft belt drive particularly valuable?

Ans. In case of a break down of any one of the engines or generators, and also when it becomes necessary to clean them without interrupting the service.

Fig. 2,722.—Plan of station arranged for extension. The space required for a central station depends upon the number and kind of lights to be supplied, and upon the character and arrangement of the machinery. In calculating the size of building required, two things must be carefully considered: first, the building must be adapted to the plant to be installed in the beginning; and second, it must be arranged so that enlargement can be made without disarranging or interfering with the plant already in existence. This is usually best secured by providing for expansion in one or two definite directions, the building being made large enough to accommodate additional units that will be necessary at some future time because of the growth of the community and consequent increased demand for electric current.

Ques. How may the design in [fig. 2,720] be modified for the installation of a storage battery?

Ans. If a storage battery be necessary, a partition may be constructed across the room A, as indicated by the dotted lines, and the battery installed in the room thus formed.

Fig. 2,723.—Interior of old Riverside station showing at the right, seven 6,000 horse power alternators driven by reciprocating engines, and at the left, a number of turbine units aggregating 90,000 horse power.

Ques. Mention a few details in the general arrangement of the building [fig. 2,720].

Ans. Two doors to the room A may conveniently be provided at K and L, the former connecting with the boiler room B, and the latter serving as the main entrance to the station. There is little that need be added to what has already been stated regarding the boiler room B. The door at F provides for the entrance of coal and the removal of ashes, while at P, the pump and heaters may conveniently be located. In the office C, visitors may be received, the station reports made out, bulletins issued from time to time, and whatever engineering problems arise may here be solved on paper by the engineer in charge of the plant. The store room D will be found convenient for various supplies, tools and appliances needed in the operation of the station. These may here be kept under lock and key and the daily waste and loss resulting from carelessness avoided.

Ques. What important point should be noted in locating the engines and boilers?

Ans. They should be so placed that the piping between them will be as short and direct as possible.

Ques. Why?

Ans. The steam pipe should be short to reduce the loss of heat between engine and boiler to a minimum, and both short and direct to avoid undue friction and consequent drop in pressure of the steam in passing through the pipe to the engine.

Entirely too little attention is given to this matter on the part of designers and it cannot be too strongly emphasized that, for economy, the steam pipe between an engine and boiler should be as short and direct as possible, having regard of course, for proper piping methods.

Ques. What should be provided for the steam pipe?

Ans. A heavy covering of approved material should be placed around the pipe to reduce the loss of heat by radiation. For this purpose hair felt, mineral wool and asbestos are used.

Fig. 2,724.—View of engine and condenser, showing how to arrange the piping to secure good vacuum. Locate the condenser as near the engine as possible; use easy bends instead of elbows; place the pump below bottom of condenser so the water will drain to pump. At A is a relief valve, for protection in case the condenser become flooded through failure of the pump, and at B is a gate valve to shut off condenser in case atmospheric exhaust is desired to permit repairs to be made to condenser during operation. A water seal should be maintained on the relief valve and special attention should be given to the stuffing box of the gate valve to prevent air leakage. The discharge valve of the pump should be water sealed.

Ques. How should the piping be arranged between the engine and condenser, and why?

Ans. It should be as short and direct as possible; especially should elbows be avoided so that the back pressure on the engine piston will be reduced as near as can be to that of the condenser.

That is to say, in order to get nearly the full effect of the vacuum in the condenser the frictional resistance of the piping should be reduced to a minimum.

Where 90° turns are necessary, easy bends should be used instead of sharp elbows. The force of this argument must be apparent by noting the practice of steam turbine builders of placing the turbine right up against the condenser, and remembering that a high vacuum is necessary to the economical working of a turbine. See fig. 1,445, page 1,182.

Ques. What are the considerations respecting the number and type of engine to be used?

Ans. In the illustration [fig. 2,720], two engines M and M' are employed, one belted to each end of the countershaft O. These engines should be of similar or identical pattern; for a small output they may be either simple or compound, as the conditions of fuel expenditure may dictate, but if the output be large, triple expansion engines or turbines are advisable.

Fig. 2,725.—"Dry pipe" for horizontal boiler: it is connected to the main outlet and its upper surface is perforated with small holes, the far end being closed. With this arrangement steam is taken from the boiler over a large area, so that it will contain very little moisture. All horizontal boilers without a dome should be fitted with a dry pipe; most engineers do not realize the importance of obtaining dry steam for engine operation.

Corliss or similar slow speed engines may advantageously be used in either case. In all cases the engine should be run condensing unless the cost for circulating water is prohibitive; even in such cases cooling towers may be installed and effect a saving.

In operation, during the greater part of the day, one engine running two or perhaps three of the generators, will carry the load, but when the load is particularly heavy, as in the morning and evening, both engines and all the generators may be required to meet the demands.

Fig. 2,726.—Method of connecting a header to a battery of boilers. Where two or more boilers are connected to a single header, the use of a reliable non-return boiler stop valve is necessary, and in some countries their installation is compulsory. A non-return boiler stop valve will instantly close should the pressure in the boiler to which it is attached suddenly decrease below that in the header, and thereby prevent the entrance of steam from the other boilers of the battery. This sudden decrease in pressure may be caused by a ruptured fitting or the blowing out of a tube, in which event an ordinary stop valve taking the place of a non-return boiler stop valve would be inadequate, as the loss of steam from the other boilers of the battery would be tremendous before an ordinary valve could be reached and closed, assuming that it would be possible to do so, which in the majority of cases it is not. Should it be desired to cut out a boiler for cleaning or repairs, the non-return boiler stop valve will not permit steam to enter the boiler from the header, even should the handwheel be operated for this purpose, as it cannot be opened by hand, but can, however, be closed. A non-return boiler stop valve should be attached to each boiler and connected to an angle valve on the header. A pipe bend should be used for connecting the valves, as this will allow for expansion and contraction. The pipe should slope a trifle downward toward the header and a suitable drain provided. This drain should be opened and all water permitted to escape before the angle valve is opened, thereby preventing any damage due to water hammer.

By exercising a little ingenuity in shifting the load on different machines at different times, both engines and dynamos, may readily be cleaned and repaired without interrupting the service.

Ques. For economy what kind of steam should be used?

Ans. Super-heated steam.

The saving due to the use of superheated steam is about 1% for every ten degrees Fahr. of super-heat. It should be used in all cases.

Ques. How should the machines be located?

Ans. Sufficient space should be allowed between them that cleaning and repairing may be done easily, quickly and effectually.

Figs. 2,727 and 2,728.—Method of preventing vibration and of supporting pipes. The figures show top and side views of a main header carried in suitable frames fitted with adjustable roller. While the pipe is illustrated as resting on the adjustable rollers, nevertheless the rollers may also be placed at the sides or on top of the pipe to prevent vibration, or in cases where the thrust from a horizontal or vertical branch has to be provided for. This arrangement will take care of the vibration without in any way preventing the free expansion and contraction of the pipe.

Ques. How should the switchboard be located?

Ans. In [fig. 2,720], the switchboard H is mounted against the wall dividing the room A from the room B, and is in line with the machines.

The advantages arising from a switchboard thus installed are, that the switchboard attendant working thereon can obtain at any time an unobstructed view of the performance of each individual machine, and he has in consequence a much better control of them; then, too, while he is engaged at the engines or generators he can also see the measuring instruments on the switchboard, and ascertain approximately the readings upon them.

In cases of emergency it is sometimes necessary for the engineer in charge of a plant to be in several places at the same time in order to prevent an accident, and that this seemingly impossibility may be approximated as nearly as possible, it is essential that the controlling devices be located as closely together as is consistent, and that no moving belt or pulley intervene between them.

These conditions are well satisfied in [fig. 2,720], and owing to the short distances between the generators and the switchboard the drop of voltage in each of the conducting wires between them will be low.

This latter advantage is worthy of notice in a station generating large currents at a low pressure. To offset the advantages just mentioned, the location of the switchboard in line with the machines introduces an element of danger to the switchboard, its apparatus, and the attendant, on account of the possible bursting of a flywheel or other parts of the machines from centrifugal force.

Figs. 2,729 and 2,730.—Points on placing stop valves. The first and most important feature is to ascertain whether the valve will act as a water trap for condensed steam. Fig. 2,729 illustrates a common error in the placing of valves, as this arrangement permits of an accumulation of condensed steam above the valve when closed, and should the engineer be careless and open the valve suddenly, serious results might follow owing to water-hammer. Fig. 2,730 illustrates the correct method of placing the valve. It sometimes occurs, however, that it is not convenient to place the valve as shown in fig. 2,730 and that fig. 2,729 is the only manner in which the valve can be placed. In such cases, the valve should have a drain, and this drain should always be opened before the large valve is opened.

If the switchboard be placed in the dotted position at H', or, in fact, at the opposite end of the room A, the damage to life and property that might result from the effects of centrifugal force would be eliminated, but in place thereof would be the disadvantages of an obstructed view of the machines from the switchboard, an obstructed view of the switchboard from the machines, inaccessibility between these two, and a greater drop of voltage in the majority of the conducting wires between the generators and the switchboard.

Ques. Describe a second arrangement of station with belt drive and compare it with the design shown in [fig. 2,720].

Fig. 2,731.—Plan of electrical station with belt drive without counter shaft. The installation here represented consists of two boilers, S, etc., and three sets of engines and generators, T, M, etc. Sufficient allowance has been made in the plans, however, for future increase of business, as additional space has been provided for an extra engine and generator set, as indicated by the dotted lines.

Ans. A floor plan somewhat different from that presented in [fig. 2,720] is shown in [fig. 2,731]. Here a belt drive is employed, but no countershaft is used. Each generator, therefore, is dependent upon its respective engine, and in consequence the flexibility obtained by the use of a countershaft is lost. On the other hand, there is less loss of mechanical power between the engines and generators in the driving of the latter, and less floor space is necessary in the room A. If, however, the floor area of this room be made the same as in the previous arrangement and the same number of machines are to be installed, they may be spaced further apart, affording in consequence considerably more room for cleaning and repairing them.

In operation, the normal conditions should be such that any two of the engine and generator sets may readily carry the average load, the third set to be used only as a reserve either to aid the other two when the load is unusually heavy or to replace one of the other sets when it becomes necessary to clean or repair the latter.

The switchboard may perhaps be best located at H, as a similar position on the opposite side of the room A would bring it beneath one or more of the steam pipes and thus endanger it should a possible leakage occur from these pipes. If located at H, however, it will be in line with the machines, and therefore will be subject to the disadvantages previously mentioned for such cases; consequently it might be as well to place it at the further end of the room, either against the partition (shown dotted) of the storage battery room if this be built, or else (if no storage battery is to be installed), against the end wall itself. The nearer end of the room A would not be very desirable for the switchboard installation on account of being so far removed from the machines, and therefore more or less inaccessible from them. Outside of what has now been mentioned, the division of the floor plan and the arrangement therein is practically the same as in [fig. 2,720], accordingly what has already been stated regarding the former installation applies, therefore, with equal force to the present installation.

Ques. Describe a plant with direct drive.

Ans. This type of drive is shown in [fig. 2,732]. Each engine is directly connected to a generator, that is, the main shafts of both are joined together in line so that the generator is driven without the aid of a belt.

Ques. What is the advantage of direct drive?

Ans. The great saving in floor space, which is plainly shown in [fig. 2,732], the portion A' representing the saving which results over the installations previously illustrated in figs. [2,720] and [2,731].

Ques. How could the floor space be further reduced?

Ans. By employing vertical instead of horizontal engines.

Ques. What should be done before drawing the plans for the station?

Ans. The types of the various machines and apparatus to be installed should, as nearly as possible, be selected in advance so that their approximate dimensions may serve as a guide in drawing up the plans of the building.

Fig. 2,732.—Plan of electrical station containing direct connected units. As shown, space is provided for an extra boiler and engine and generator set, as indicated by the dotted lines. Space also exists for a storage battery room if necessary, and the partition dividing this room from the engine and dynamo room is shown by a dotted line as in previous cases.

Owing to the great difference in these dimensions for the various types, and in fact for the same types as manufactured by different concerns, no definite rules regarding the necessary space required can here be given. In a general way, however, the author has endeavoured to indicate by the drawings the relative amounts of space that ordinarily would be considered sufficient.

Ques. What is the disadvantage of direct drive?

Ans. A more expensive generator is required because it must run at the same speed as the engine, which is relatively low as compared with that of a belted generator.

Station Construction.—The construction or rearrangement of the building intended for the plant is a problem that under ordinary conditions would be solved by an architect, or at least by an architect with the assistance of an electrical or mechanical engineer, still there are many installations where the electrical engineer has been compelled to design the building.

In such instances he should be equipped with a general knowledge of the construction of buildings.

Foundations.—The foundation may be either natural or artificial; that is, it may be composed of rock or soil sufficiently solid to serve the purpose unaided, or it may be such as to require strengthening by means of wood or iron beams, etc. In either case any tendency toward a considerable settling or shifting of the foundation due to the action of water, frost, etc., after the station has been completed must be well guarded against. To this end special attention should be given to the matter of drainage.

Ques. How should the foundation be constructed for the machines?

Ans. The foundations constructed for the machines should be entirely separate from that built for the walls of the building, so that the vibrations of the former will not affect the latter.

If there be several engines and dynamos to be installed, it is best to construct two foundations, one for the engines and one for the dynamos. If, however, there be considerable distance between the units, it may be advisable to build a separate foundation for each engine and for each dynamo. The material of which these foundations are composed should if the machines be of 20 horse power or over, possess considerable strength and be impervious to moisture. Brick, stone and concrete are desirable for the purpose, and only the best quality of cement mortar should be employed. Care must be taken that lime mortar is not used in place of cement mortar, as the former is not well adapted to withstand the vibrations of the machines without crumbling.

Fig. 2,733.—Angle for foundation footing. In ordinary practice the footing courses upon which the walls of the building proper rest, consist of blocks or slabs of stone as large as are available and convenient to handle. Footings of brick or concrete are also used in very soft soils; footings consisting of timber grillage are often employed. A grillage of iron or steel beams has also been used successfully. The inclination of the angle φ, of footing should be about as follows: for metal footings 75°; for stone, 60°; for concrete, 45°; for brick, 30°. Damp proof courses of slate, or layer of asphalt are laid in or on the foundations or lower walls to prevent moisture arising or penetrating by capillary attraction.

Ques. Describe a method of constructing foundations.

Ans. An excavation is made to the desired depth and a form inserted corresponding to the desired dimensions for the foundation. A template is placed on top locating all the centers, with iron pipes suspended from these centers, two or three sizes larger than the anchor bolts. At the lower end of the pipes are core boxes. Concrete is poured into the mould thus formed, and when hard, the forms are removed thus leaving the solid foundation. The anchor bolts are inserted through the pipes and passed through iron plates at the lower end as shown in [fig. 2,734], being secured by nuts. By using pipe of two or three bolt diameters a margin is provided for adjustment so the bolts will pass through the holes in the frame of the machine thus allowing for any slight errors in laying out the centers on the template.

Fig. 2,734.—Concrete foundation showing method of installing the anchor bolts.

Ques. What is the object of the openings in the bottom of the foundation?

Ans. In case of a defective bolt, it may be replaced by a new one without injury to the foundation.

Walls.—Regarding the material for the walls of the station iron, stone, brick and wood may be considered. Of these, iron in the form of sheets or plates would be entirely fireproof, but being itself a conductor would introduce difficulties in maintaining a high insulation resistance of the current carrying circuits; it would also make the building difficult to heat in winter and to keep cool in summer. Stone in the form of limestone, granite or sandstone, as a building material is desirable for solidity and attractiveness; it is also fireproof and an insulator, but the high cost of such a structure for an electrical station usually prohibits its use except in private plants or in electrical stations located in large cities.

Fig. 2,735.—View showing part of template for locating anchor bolt centers, pipes through which the bolts pass and bolt boxes at lower end of bolts. The completed foundation is shown in [fig. 2,734], with template removed. The template is made of plain boards upon which the center lines are drawn, and bolt center located. Holes are bored at the bolt centers to permit insertion of the pipes as shown.

Brick is a good material and is readily obtained in nearly all parts of the country; it is comparatively cheap, and is also an insulating and fireproof material. The bricks selected for this purpose should possess true sharp edges, and be hard burned.

Ques. What are the features of wood?

Ans. Wood forms the cheapest material that can be used for the walls of electrical stations, and it usually affords satisfaction, but has the disadvantage of high fire risk.

Roofs.—In [fig. 2,736] is shown one form of construction for the roof of an electrical station. The end view here presented shows the upper portion of the walls at B and D; these support the iron trusses C, and the roof proper MN. In many stations there is provided throughout the length of the building, a monitor or raised structure on the peak of the roof for ventilation and light. The end view of the monitor is shown at S in the figure; its sides should be fitted with windows adjustable from the floor.

Fig. 2,736.—One form of roof construction.

Floors.—The floor of the station should be so designed that it will be capable of supporting a reasonable weight, but as the weights of the machines are borne entirely by their respective foundations the normal weight upon the floor will not be great; for short periods, however, it may be called upon to support one or two machines while they are being placed in position or interchanged, and due allowance must be made for such occurrences.

Station floors for engine and dynamo rooms are, as a rule, constructed of wood. Where very high currents are generated, however, insulated floors of special construction mounted on glass are necessary as a protection from injurious shocks. Brick, concrete, cement, and other substances of a similar nature are objectionable as a floor material for engine and dynamo rooms on account of the grit from them, caused by constant wear, being liable to get into the bearings of the machines.

Where there are no moving parts, however, as in the boiler room, the materials just mentioned possess no disadvantages and are preferable to wood on account of being fireproof.

Chimneys.—These are generally constructed of brick and iron, sometimes of concrete. Iron chimneys cost less than brick chimneys, necessitate less substantial foundations, and are free from the liability of cracking. They must be painted to prevent corrosion, are less substantial, and lose considerably more heat by radiation than do brick chimneys.

Fig. 2,737.—An example of direct connected unit with gas engine power. The view shows a Westinghouse 200 kva., 4,000 volt, three phase, 60 cycle alternator direct connected to a gas engine.

Fig. 2,738.—Curves showing comparative costs of chimney and mechanical draft. In certain of these, the cost of the existing chimney is known, and that of the complete mechanical draft plant is estimated, while in others, the cost of mechanical draft installation is determined from the contract price, and the expense of a chimney to produce equivalent results is calculated. Costs are shown for both single, forced and induced engine driven fans and for duplex engine driven plants, in which either fan may serve as a relay. An apparatus of the latter type is the most expensive, and finds its greatest use where economizers are employed.

Both brick and iron chimneys, require an inner wall or lining of brick, which forms the flue proper, and in order that this wall be not cracked by sudden cooling an air space is left between it and the outer wall. In a brick chimney the inner wall need not extend much beyond half the height of the chimney, but when iron is used it should reach to the top.

Ques. Upon what does the force of natural draught in a chimney depend?

Ans. It depends upon the difference between the weight of the column of hot gases inside the chimney and the weight of a like column of the cold external air.

Figs. 2,739 and 2,740.—Substituting mechanical draught in place of chimney. The relative proportions of a brick chimney, and of the smoke pipe required when mechanical draft is introduced are forcibly shown in the illustrations, which show the works of the B.F. Sturtevant Co., at Jamaica Plain, Mass. The removal of the boilers to a position too far distant from the existing chimney to permit of its longer fulfilling its office, led to the substitution of an induced draft fan and the subsequent removal of the chimney. The present stack or smoke pipe, barely visible in fig. 2,740, extends only 31 feet above the ground, and no trouble is experienced from smoke.

Ques. How is the intensity of the draught expressed?

Ans. In terms of the number of inches of a water column sustained by the pressure produced.

Ques. Are high chimneys necessary?

Ans. No.

Chimneys above 150 feet in height are very costly, and their increased cost is not justified by increased efficiency.

Figs. 2,741 to 2,744.—Installation of forced draft system to old boiler plant. The figures illustrate the simplest method. The fan which is of steel plate with direct connected double cylinder engine, is placed immediately over the end of a brick duct into which the air is discharged. This duct is carried under ground across the front of the boilers, to the ash pits of each of which connection is made through branch ducts. Each branch duct opening is provided with special ash pit damper, operated by notched handle bar, as illustrated in the detail. This method of introduction serves to distribute the air within the ash pit, and to secure even flow through the fuel upon the grate above. Of course, the ash pit doors must remain closed in order to bring about this result. A chimney of sufficient height to merely discharge the gases above objectionable level is all that is absolutely necessary with this arrangement. Although the introduction of a fan in an old plant is usually evidence of the insufficiency of the existing chimney to meet the requirements, such a chimney, will, however, usually serve as a discharge pipe for the gases when the fan is employed. The fan thus becomes more than a mere auxiliary to the chimney; it practically supplants it so far as the method of draught production is concerned.

The latest chimney practice is to build two or more small chimneys instead of one large one. A notable example is the Spreckels Sugar Refinery in Philadelphia, where three separate chimneys are used for one boiler plant of 7,500 horse power. The three chimneys are said to have cost several thousand dollars less than an equivalent single chimney.

Very tall chimneys have been characterized by one writer as "monuments to the folly of their builders."

Figs. 2,745 and 2,746.—Comparison of chimney draft and mechanical draft. The illustrations show a plant of 2,400 H.P. of modern water tube boilers, 12 in number, set in pairs and equipped with economizers. Fig. 2,745 indicates the location of a chimney, 9 feet in internal diameter by 180 feet high, designed to furnish the necessary draft; fig. 2,746 represents the same plant with a complete duplex induced draught apparatus substituted for the chimney, and placed above the economizer connections. Each of the two fans is driven by a special engine, direct connected to the fan shaft, and each is capable of producing draft for the entire plant. A short steel plate stack unites the two fan outlets and discharges the gases just above the boiler house roof. All of the room necessary for the chimney is saved, and no valuable space is required for the fans.

Ques. How is mechanical draft secured?

Ans. In two ways, known respectively as induced draught and forced draught.

Ques. Describe the method of induced draft.

Ans. A fan is located in the smoke flue, and which in operation draws the gases through the furnace and discharges them into a short chimney.

Ques. Describe the method of forced draft.

Ans. In this method, air is forced into the furnace underneath the grate bars by means of a fan or a steam jet blower.

Fig. 2,747.—Forced draft plant with hollow bridge wall at the Crystal Water Co., Buffalo, N. Y. The air is delivered to the ash pit via the hollow bridge wall, being supplied under pressure by the blower seen at the side of the boiler setting. As shown, the blower is operated by a small reciprocating engine; however, compact blowing units with steam turbine drive can be had and which are designed to be placed in the boiler setting.

Ques. What is the application of the two systems?

Ans. Induced draft is installed mostly in new plants, while forced draft is better adapted to old plants.

Steam Turbines.—It is not the author's intention to discuss at length the steam end of the electric plant, because too much space would be required, and also because the subject belongs properly to the field of mechanical engineering rather than electrical engineering. However, because of the recent introduction of the steam turbine for the direct driving of large generators, and the fact that it is now almost universally used in large central stations, a detailed explanation of its principles and construction may not be out of place.

Fig. 2,748.—Longitudinal section of elementary Parsons type steam turbine. The turbine consists essentially of a fixed casing, or cylinder, and a revolving spindle or drum. The ends of the spindle are extended in the form of a shaft, carried in two bearings A and B, and, excepting the small parts of the governing mechanism and the oil pump, these bearings are the only rubbing parts in the entire turbine. Steam enters from the steam pipe at C and passes through the main throttle or regulating valve D, which, as actually constructed, is a balanced valve. This valve is operated by the governor through suitable controlling mechanism. The steam enters the cylinder through the passage E and, turning to the left passes through alternate stationary and revolving rows of blades, finally emerging from them at F and flowing through the connection G to the condenser or to the atmosphere, depending upon whether the turbine is condensing or non-condensing. Each row of blades, both stationary and revolving, extends completely around the turbine and the steam flows through the full annulus between the spindle and the cylinder. In an ideal turbine the lengths of the blades and the diameter of the spindle which carries them would continuously and gradually increase from the steam inlet to the exhaust. Practically, however, the desired effect is produced by making the spindle in steps, there being generally three such steps or stages, H, J and K. The blades in each step are arranged in groups of increasing length. At the beginning of each of the larger steps, the blades are usually shorter than at the end of the preceding smaller step, the change being made in such a way that the correct relation of blade length to spindle diameter is secured. The steam, acting as previously described, produces a thrust tending to force the spindle toward the left, as seen in the cut. This thrust, however, is counteracted by the "balance pistons," L, M and N, which are of the necessary diameter to neutralize the thrust on the spindle steps, H, J and K, respectively. These elements are called "pistons" for convenience, although they do not come in contact with the cylinder, but both the pistons and the cylinder are provided with alternate rings which form a labyrinth packing to retard the leakage of steam. In order that each balance piston may have the proper pressure on both sides, equalizing passages O, P and Q are provided connecting the balance pistons with the corresponding stages of the blading. The end thrust being thus practically neutralized by means of the balance pistons, the spindle "floats" so that it can be easily moved in one direction or the other. In order to definitely fix the position of the spindle, a small adjustable collar bearing is provided at R, inside the housing of the main bearing B. This collar bearing is adjustable so as to locate and hold the spindle in such position so that there will be such a clearance between the rings of the balance piston and those of the cylinder, that the leakage of steam will be reduced to a minimum and, at the same time, prevent actual contact under varying conditions of temperature. Where the shaft passes out of the cylinder, at S and T, it is necessary to provide against in-leakage of air or out-leakage of steam by means of glands. These glands are made tight by water packing without metallic contact. The shaft of the turbine is extended at U and coupled to the shaft of the alternator by means of a flexible coupling. The high pressure turbines are so proportioned that, when using steam as previously described, they have enough capacity to take care of the ordinary fluctuations of load when controlled by the governor through the valve D, thus insuring maximum economy of steam consumption at approximately the rated load. To provide for overloads, the valve V is supplied to admit steam to an intermediate stage of the turbine. This valve shown diagrammatically in the illustration, is arranged to be operated by the governor and is, according to circumstances, located either as shown by the illustration, or at another stage of the turbine.

Fig. 2,749.—Arrangement of blading in Parsons type turbine, consisting of alternate moving and stationary blades. The path taken by the steam is indicated by the arrows.

A turbine is a machine in which a rotary motion is obtained by transference of the momentum of a fluid or gas. In general the fluid is guided by fixed blades, attached to a casing, and, impinging on other blades mounted on a drum or shaft, causing the latter to revolve.

Turbines are classed in various ways as: 1, radial flow, when the steam enters near the center and escapes toward the circumference; and 2, parallel flow, when the steam travels axially or parallel to the length of the turning body.

Turbines are commonly, yet erroneously classed as:
1. Impulse;
2. Reaction.

Ques. What is the distinction between these two types?

Ans. In the so called impulse type, steam enters and leaves the passages between the vanes at the same pressure. In the so called reaction type, the pressure is less on the exit side of the vanes than on the entrance side.

Fig. 2,750.—Sectional view of Parsons-Westinghouse turbine, showing rotor and governor.

[Fig. 2,750] is a sectional view of the Parsons-Westinghouse parallel flow turbine. Steam from the boiler enters first a receiver in which are the governor controlled admission valves. These valves are actuated by a centrifugal governor.

Steam does not enter the turbine in a continuous blast, but intermittently, or in puffs. The speed regulation is therefore accomplished by proportioning the duration of these puffs to the load of the engine, this being effected by the governor, [fig. 2,752].

The governor of the turbine has only to move a small pilot valve, or slide, E, which admits steam under the piston F, and lifts the throttle valve proper off its seat.

As soon as the pilot valve closes, the spring shifts the main throttle valve. Thus, at light loads, the main throttle or admission valve is continually opening and shutting at uniform intervals, the length of time during which it remains open depending upon the load.

As the load increases, the duration of the valve opening also increases, until at full load the valve does not reach its seat at all and the steam flows steadily through the turbine. The steam thus admitted flows into the annular passage A, [fig. 2,750], by the opening S, and then past the blades, revolving the rotor.

When the load increases above the normal rated amount a secondary pilot valve is moved by the same means, this in turn admitting steam to a piston, similar to F, which lifts another throttle valve. This admits steam into the annular space I, so that it acts upon the larger diameter of the drum or rotor, giving largely increased power for the time being.

The levers or arms of the governor are mounted upon knife edges instead of pins, making it extremely sensitive. The tension spring may be adjusted by hand while the turbine is running.

Fig. 2,751.—Sectional view of a combination impulse and reaction single flow turbine. This is a modification of the single flow type, in which the smallest barrel of reaction blading is replaced by an impulse wheel. Steam is admitted to the nozzle block A, is expanded in the nozzles and discharged against a portion of the periphery of the impulse wheel. The intermediate and low pressure stages are identical with the corresponding stages in the single flow type. The substitution of the impulse element for the high pressure section of reaction blading has no influence one way or another on the efficiency. That is to say the efficiency of an impulse wheel is about the same at the least efficient section of reaction blading. This design is attractive, however, in that it shortens the machine materially, and gives a stiffer design of rotor. The entering steam is confined in the nozzle chamber until its pressure and temperature have been materially reduced by expanding through the nozzles. As the nozzle chamber is cast separately from the main cylinder, the temperature and pressure differences to which the cylinder is subjected are correspondingly lessened. However, probably on account of its small diameter at the high pressure section, the straight Parsons type has always shown itself to be adequate for all of the steam pressures and temperatures encountered in ordinary practice.

The governor does not actually move the pilot valve, but shifts the point L in [fig. 2,752]. A reciprocating motion is given to the rod I by a small eccentric on the governor shaft; this is driven by worm gearing shown near O in [fig. 2,750], so that the eccentric makes one revolution to about eight of the turbine. Thus, with a turbine running 1,200 revolutions, the rod I would be moved up and down 150 times per minute. As the points A and H are fixed, the motion is conveyed to the small pilot valve E, thus giving 150 puffs a minute. The governor in shifting the point L brings the edge of the pilot valve nearer the port and so cuts off the steam earlier.

The annular diameter or space between the rotor and the stator is gradually increased from inlet to exhaust, the blades being made longer in each ring. When the mechanical limit is reached, the diameter of the rotor is increased as at I and D so as to keep the length of blade within bound.

Balance pistons as at B, C, F are attached to the rotor, their office being to oppose end thrust upon those blades in corresponding diameter of the rotor. Communication is established through the passage V and pipe M between the eduction pipe and the back of these pistons, thus increasing the efficiency of their balancing and also taking care of any leakage past them.

A small thrust bearing T prevents end play of the rotor, and is adjustable to maintain the proper clearance between the rings of blades; this varies from ⅛ inch at the admission to 1 inch at the exhaust. This bearing also takes up any extra unbalanced thrust. A turbine should operate with a high vacuum, because without this it does not compare favorably with an ordinary reciprocating engine from the point of economy.

Fig. 2,752.—Sectional view of governor of the Parsons-Westinghouse turbine.

Separate air pumps are provided to create the vacuum.

Where the ordinary type of vertical air pump is employed, a booster or vacuum increaser is added, as nothing below 26 inches is advisable, 28 and 29 inches being always striven for. It is also preferable to use a certain amount of super-heat with steam turbines.

To assist in producing the high vacuum, exhaust passages are made large, the eduction passage E in [fig. 2,750] being nearly twenty-three times the area of the steam pipe.

Among other details, a noteworthy feature is a small oil pump K, which circulates oil through bearings of the machinery, the oil being drawn from the tank under the governor shaft and gravitating there after use. No pressure of oil is employed. Stuffing rings prevent leakage; these consist of alternate grooves and collars in shaft and bearing, like the grooves in an indicator piston.

Ques. Why is a high vacuum desirable?

Ans. Because the turbine is capable of expanding the steam to a very low terminal pressure, and this is necessary for economy.

Ques. What may be said of the working pressures for turbines?

Ans. To meet the varied conditions of service, turbines are designed to operate with: 1, high pressure, 2, low pressure, or 3, mixed pressure.

Fig. 2,753.—Sectional view of a double flow turbine. The maximum economical capacity of a single flow turbine is limited by the rotative speed. The economical velocity at which the steam may pass through the blades of the turbine depends on the velocity of the moving blades. The capacity of the turbine depends on the weight of the steam passed per unit of time, which in turn depends on the mean velocity and the height of the blades. For a given rotative speed, the mean diameter of blade ring practicable is limited by the allowable stresses due to centrifugal force, and there is a practical limit for the height of the blades. Now if the rotative speed be taken only half as great, the maximum diameter of the rotor may be doubled and, without increasing the height of the blades, the capacity of the turbine will be doubled. So with the single flow steam turbine as well as with the single crank reciprocating engine, there is a practical limiting economical capacity for any given speed. If this limit be reached with a single crank reciprocating engine, a unit of double the power may be produced at the same speed by coupling two single crank engines to one shaft. Similar results are secured making a double flow turbine which is in effect, as will be seen from the figure, two single flow turbines made up in a single rotor in a single casing with a common inlet and two exhausts. Steam enters the nozzle block, acts on the impulse element, and then the current divides, one-half of the steam going through the reaction blading at the left of the impulse wheel; the remainder passes over the top of the impulse wheel and through the impulse blading at the right.

High pressure turbines operate at about the same initial pressure as triple expansion engines.

Low pressure, as here applied, means the exhaust pressure of the reciprocating engine from which the exhaust steam passes through the turbine before entering the condenser.

Mixed pressure implies that the exhaust steam is supplemented, for heavy loads, by the admission of live steam.

Ques. What determines the working pressure?

Ans. When all the power is furnished by the turbine, it is designed for high pressure; when operated in combination with a reciprocating engine, low pressure is used for constant load, and mixed pressure for variable load.

Fig. 2,754.—Sectional view of a semi-double flow turbine. This is a modification in which the intermediate section of reaction blading is single flow, and the low pressure section only is double flow. This would be analogous to a four cylinder triple expansion engine, that is, one with one high pressure, one intermediate pressure and two low pressure cylinders—a design not at all uncommon in very large engines in which the required dimensions of a single low pressure cylinder would be prohibitive. Such turbines are useful for capacities greater than is desirable for a single flow turbine, and which are still below the maximum possibilities of a double flow turbine of the same speed. In such machines the best efficiency is secured by making the intermediate blading in a single section large enough to pass the entire quantity of steam. A "dummy" similar to those used on the single flow Parsons type, shown at the right of the impulse wheel, compels all of the steam to pass through the single intermediate section of the reaction blading, and balances the end thrust due to this section. When the steam issues from the intermediate section, the current is divided, one-half passing directly to the adjacent low pressure section, while the other half passes through the holes shown in the periphery of the hollow rotor and through the rotor itself, beyond the dummy ring, into the other low pressure section at the left hand end of the turbine.

NOTE.—There are logical engineering reasons for the existence of the several types of turbine, viz., single flow, double flow, and semi-double flow. The double flow turbine is not inherently superior to the single flow design, but is used under conditions for which the single flow machine is unsuitable. Similarly, the semi-double flow is recommended only for conditions which it can meet more satisfactorily than either of the other types.

NOTE.—Low pressure turbines use exhaust steam from non-condensing engines and are valuable as an adjunct to existing plants for the purpose of increasing economy and capacity with a minimum outlay for new equipment.

NOTE.—Bleeder turbines are for use in plants which are required to furnish, not only power, but also considerable and varying quantities of low pressure steam for heating purposes. In these turbines a part of the steam after it has done work in the high pressure stages may be diverted to the heating system, and the remainder expanded through the low pressure blading and exhausted into the condenser. In this way none of the energy of the heating steam, due to the difference of pressure between the boiler and the heating system is wasted. On the other hand if no steam is required for heating purposes, the turbine operates just as efficiently as though the bleeder feature were absent.

Fig. 2,755.—Westinghouse valve gear with steam relay. In the smaller turbines, the governor acts directly on the steam admission valves, opening first the primary valve, and then, if necessary, the secondary valve, after the primary is fully open. In turbines of the single flow Parsons type, the governor actuates two small valves controlling ports leading to steam relay cylinders which operate the admission valves. The little valve controlling the relay cylinder for the secondary valve has more lap than the other and consequently does not come into action until the primary valve has attained its maximum effective opening. The figure shows the general design of this type of valve gear.

The De Laval steam turbine is termed by its builders a high speed rotary steam engine. It has but a single wheel, fitted with vanes or buckets of such curvature as has been found to be best adapted for receiving the impulse of the steam jet. There are no stationary or guide blades, the angular position of the nozzles giving direction to the jet. The nozzles are placed at an angle of 20 degrees to the plane of motion of the buckets. The best energy in the steam is practically devoted to the production of velocity in the expanding or divergent nozzle, and the velocity thus attained by the issuing jet of steam is about 4,000 feet per second. To attain the maximum efficiency, the buckets attached to the periphery of the wheel against which this jet impinges should have a speed of about 1,900 feet per second, but, owing to the difficulty of producing a material for the wheel strong enough to withstand the strains induced by such a high speed, it has been found necessary to limit the peripheral speed to 1,200 or 1,300 feet per second.

It is well known that in a correctly designed nozzle the adiabatic expansion of the steam from maximum to minimum pressure will convert the entire static energy of the steam into kinetic energy. Theoretically this is what occurs in the De Laval nozzle. The expanding steam acquires great velocity, and the energy of the jet of steam issuing from the nozzle is equal to the amount of energy that would be developed if an equal volume of steam were allowed to adiabatically expand behind the piston of a reciprocating engine, a condition, however, which for obvious reasons has never yet been attained in practice with the reciprocating engine. But with the divergent nozzle the conditions are different.

The Curtis turbine is built by the General Electric Company at their works in Schenectady, N. Y., and Lynn, Mass. They are of the horizontal and vertical types. In the vertical type the revolving parts are set upon a vertical shaft, the diameter of the shaft corresponding to the size of the machine.

The shaft is supported by and runs upon a step bearing at the bottom. This step bearing consists of two cylindrical cast iron plates bearing upon each other and having a central recess between them into which lubricating oil is forced under pressure by a steam or electrically driven pump, the oil passing up from beneath.

Figs. 2,756 and 2,757.—Westinghouse valve gear with oil relay. Governors for the larger turbines, particularly those of the combination impulse and reaction double, or single double flow type, employ an oil relay mechanism, as shown in the figure, for operating the steam valves. In these turbines the lubricating oil circulating pump, maintains a higher pressure than is required for the lubricating system. The governor controls a small relay valve A which admits pressure oil to, or exhausts it from the operating cylinder. When oil is admitted to the operating cylinder raising the piston, the lever C lifts the primary valve E. The lever D moves simultaneously with C, but on account of the slotted connection with the stem of the secondary valve F, the latter does not begin to lift until the primary valve is raised to the point at which its effective opening ceases to be increased by further upward travel. In the Westinghouse designs, the operating valve, A is connected not only to the governor, but also to a vibrator, which gives it a slight but continuous reciprocating motion, while the governor controls its mean position. The effect of this is manifested in a slight pulsation throughout the entire relay system, which, so to speak, keeps it "alive" and ready to respond instantly, to the smallest change in the position of the governor. The oil relay can be made sufficiently powerful to operate valves of any size, and it is also in effect a safety device in that any failure of the lubricating oil supply will automatically and immediately shut off the steam and stop the turbine.

A weighted accumulator is sometimes installed in connection with the oil pipe as a convenient device for governing the step bearing pumps, and also as a safety device in case the pumps should fail, but it is seldom required for the latter purpose, as the step bearing pumps have proven after a long service in a number of cases, to be reliable. The vertical shaft is also held in place and kept steady by three sleeve bearings one just above the step, one between the turbine and generator, and the other near the top.

Fig. 2,758.—Elevation of new turbine central station erected by the Boston Edison Co. The turbine room is 68 feet, 4 inches wide and 650 feet long from outside to outside of the walls. The boiler room is 149 feet, 6 inches by 640 feet and equipped with twelve groups of boiler, one group consisting of eight 512 H.P. boilers for each turbine. The switching arrangements are located in a separate building as shown in the elevation. The total floor space covered by boiler room, turbine room and switchboard room is 2.64 square feet per kw. The boilers are all on the ground floor. See [fig. 2,714] for plan.

These guide bearings are lubricated by a standard gravity feed system. It is apparent that the amount of friction in the machine is very small, and as there is no end thrust caused by the action of the steam, the relation between the revolving and stationary blades may be maintained accurately. As a consequence, therefore, the clearances are reduced to the minimum.

The Curtis turbine is divided into two or more stages, and each stage has one, two or more sets of revolving blades bolted upon the peripheries of wheels keyed to the shaft. There are also the corresponding sets of stationary blades bolted to the inner walls of the cylinder or casing.

The governing of speed is accomplished in the first set of nozzles and the control of the admission valves here is effected by means of a centrifugal governor attached to the top end of the shaft. This governor, by a very slight movement, imparts motion to levers, which in turn work the valve mechanism.

The admission of steam to the nozzles is controlled by piston valves which are actuated by steam from small pilot valves which are in turn under the control of the governor.

Fig. 2,759.—Illustration of a weir. To make a weir, place a board across the stream at some point which will allow a pond to form above. The board should have a notch cut in it with both side edges and the bottom sharply beveled toward the intake, as shown in the above cut. The bottom of the notch, which is called the "crest" of the weir, should be perfectly level and the sides vertical. In the pond back of the weir, at a distance not less than the length of the notch, drive a stake near the bank, with its top precisely level with the crest. By means of a rule, or a graduated stake as shown, measure the depth of water over the top of stake, making allowance for capillary attraction of the water against the sides of the weir. For extreme accuracy this depth may be measured to thousandths of a foot by means of a "hook gauge," familiar to all engineers. Having ascertained the depth of water over the stake, refer to the accompanying table, from which may be calculated the amount of water flowing over the weir. There are certain proportions which must be observed in the dimensions of this notch. Its length, or width, should be between four and eight times the depth of water flowing over the crest of the weir. The pond back of the weir should be at least fifty per cent. wider than the notch and of sufficient width and depth that the velocity of flow or approach be not over one foot per second. In order to obtain these results it is advisable to experiment to some extent.

Speed regulation is effected by varying the number of nozzles in flow, that is, for light loads fewer nozzles are open and a smaller volume of steam is admitted to the turbine wheel, but the steam that is admitted impinges against the moving blades with the same velocity always, no matter whether the volume be large or small. With a full load and all the nozzle sections in flow, the steam passes to the wheel in a broad belt and steady flow.

NOTE.—The weir table on this page contains figures 1, 2, 3, etc., in the first vertical column which indicates the inches depth of water running over weir board notches. Frequently the depths measured represent also fractional inches, between 1 and 2, 2 and 3, etc. The horizontal line of fraction at the top represents these fractional parts, and can be applied between any of the numbers of inches depth, from 1 to 25. The body of the table shows the cubic feet, and the fractional parts of a cubic foot, which will pass each minute for each inch in depth, and for each fractional part of an inch by eighths for all depths from 1 to 25 inches. Each of these results is for only one inch width of weir. To estimate for any width of weir the result obtained for one inch width must be multiplied by the number of inches constituting the whole horizontal length of weir.

Figs. 2,760 and 2,761.—Samson vertical runner and shaft, and complete Samson vertical turbine. The runner is composed of two separate and distinct types of wheel, having thereby also two diameters. Each wheel or set of buckets receives its separate quantity of water from one and the same set of guides, but each set acts only once and singly upon the water used, and the water does not act twice upon the combined wheel, as some suppose. In construction, the lower or main set of buckets is made of flanged plate steel, and cast solidly into a heavy ring surrounding the outer and lower edges, and into a heavy diaphragm, separating the two sets of buckets.

Fig. 2,762.—Water discharging from a needle nozzle due to a pressure of 169 lbs. per sq. in.

Hydro-Electric Plants.—The economy with which electricity can be transmitted long distances by high tension alternating currents, has led to the development of a large number of water powers in more or less remote regions.

Fig. 2,763.—Photograph of an operating tangential water wheel equipped with Pelton buckets.

This economy is possible by the facility with which alternating current can be transformed up and down. Thus at the hydro-electro plant, the current generated by the water wheel driven alternator is transformed to very high pressure and transmitted with economy a long distance to the distributing point where it is transformed down to the proper pressure for distribution.

A water wheel or turbine is a machine in which a rotary motion is obtained by transference of the momentum of water; broadly speaking, the fluid is guided by fixed blades, attached with a casing, and impinging on other blades mounted on a drum or shaft, causing the latter to revolve.

There are two general classes of turbine:
1. Impulse turbines;
2. Reaction turbines.

Fig. 2,764.—Sectional elevation of one of the 5,000 horse power vertical Pelton-Francis turbines directly connected to generator, as installed for the Schenectady Power Co.

Ques. What is an impulse turbine?

Ans. One in which the fluid is directed by means of a series of nozzles against vanes which it drives.

Ques. What is a reaction turbine?

Ans. One in which the pressure or head of the water is employed rather than its velocity. The current is deflected upon the wheel by the action of suitably disposed guide blades, the passages being full of water. Rotary motion is obtained by the change in the direction and momentum of the fluid.

Figs. 2,765 to 2,768.—Cross sections of Lowel dam power house, and wheel pits containing sixteen Samson turbines: The section C-D gives an end view of the generator room showing the locations of the generators below the head level water. They are secure against flood water, or leakage, by well constructed stuffing boxes in the iron bulkheads, through which the turbine wheel shafts pass and connect to the generators. Section E-F gives an end view of one of these wheel rooms or penstocks, and shows the extension of the draft tube from wheel case into tail water. The section A-B shows the sub-structure of gravel and macadam under the controlling gates, this forming also a portion or extension of the dam proper. These gates turn on an axis made of two 15 inch I beams securely riveted together with plates and angle irons to which the wooden frame is attached. The radius of the gates is 14 feet. They are designed to allow the water to pass underneath the gate, thus controlling any height of head water. They are intended to take care of an excess of water at unusual stages of the river. The whole affair has been well designed and executed. This plant furnishes a good example of a secure, and level foundation, since the wheel houses and generator room are immediately on the rock. It is necessary in all tandem plants to provide a very secure, substantial super-structure so that the long line of turbines and shaft will always remain straight and in proper alignment with the generator and the turbine cases. Users cannot be reminded of this too often.

Ques. Name three classes of reaction turbines.

Ans. Parallel flow, inward flow, and outward flow.

Parallel flow turbines have an efficiency of about 70% and are suited for low falls not over 30 feet. Inward and outward flow turbines have an efficiency of about 85%. Impulse turbines are suitable for high heads.

Figs. 2,769 and 2,770.—Exterior and interior of hydro-electric plant at Harrisburg, Va. It is located on the south fork of the Shenandoah River, twelve and one-half miles distant. A dam 720 feet long and 15 feet high was built on a limestone ledge running across the river; which with a fall of 5 feet from the dam to the power house, a quarter of a mile distant, secured an effective head pressure of 20 feet. The power house, comprising the generator room and the wheel room, also the machinery room, are here shown. The wheel room, which is 20 × 40 feet, extends across the head race, and rests upon solid concrete walls, forming the sides and ends of the wheel pits. The end wall is 6 feet thick at the bottom, and 4½ feet at the top. It has three arched openings, each 8 feet wide and 9 feet high, through which the water escapes after leaving the turbines. The intake is protected by a wrought iron rack 40 feet long. The power is obtained by three 50 inch vertical shaft Samson turbines, with a 20 inch Samson for an exciter. The three large turbines have a rating of 1,350 horse power; and are connected to the main horizontal line shaft by bevel mortise gears 7 feet diameter and 15 inches face. The couplings on the main shaft have 48 inch friction clutch hubs, permitting either or each turbine being operated, or shut down independently of the others. The main shaft is 85 feet long and 6 inches diameter; making 280 revolutions. This shaft carries two pulleys 70 inches diameter and 38 inches face for driving the generators. The accompanying illustration shows the harness work, gears, pulleys, etc., furnished with the turbines. The 20 inch horizontal shaft Samson turbine of 72 horse power is direct connected to an exciter generator of 20 kw., running 700 rev. per min. The two large generators are driven 450 revolutions per minute by belts producing a three phase current of 60 cycles of 11,500 volts for the twelve and one-half miles transmission. The line consists of three strands of No. 4 bare copper wire. This current is used for lighting and power purposes, and the plant is of the latest improved design and construction.

Isolated Plants.—When electric power transmission from central stations first came into commercial use, the distance from the station at which current could be obtained at a reasonable cost was exceedingly limited.