ITY OF CALIFORNIA LIBRARY OF THE UNIVERSITY OF CALIFORNIA -^ a s 6 TY OF CALIFORNIA 5 LIBRARY OF THE UNIVERSITY OF CALIFORNIA 1 3 TY OF CALIFORNIA LIBRARY OF THE UNIVERSITY OF CALIFORNIA ) ^ rff&L UNIVERSITY OF CALIFORNIA LIBRARY OF THE UNIVERSITY OF CAI = "~~~ ~ UNIVERSITY OF CALIFORNIA LIBRARY OF THE UNIVERSITY OF CAI THE WATTHOUR METER BY WILLIAM M. SHEPARD '! AND ALLEN G. JONES TECHNICAL PUBLISHING COMPANY 604 MISSION STREET. SAN FRANCISCO 1910 Copyright 1910 BY TECHNICAL PUBLISHING COMPANY P R E F A C K Considerable information may be derived from various sources relative to the watthour meter. Real- izing the desirability and advantage of collecting and publishing such information in concrete form, the authors have endeavored to describe the prominent types and the best usage of modern domestic watt- hour meters. It has been the intention to prepare the facts in a form which will impart to the central station manager, the practical meter man and to the student alike, in- formation which will be edifying and serviceable for reference and as a guide for the proper installation, connection, testing and maintenance of that most vital factor in the distributing system the watthour meter. Especial attention has been given the induction type and a brief but concise explanation of its theory and operation has been made without the use of higher mathematics. Maintenance and testing are also treated in detail with the dominant idea of giving the practical man assistance in modern, effective and quick methods of obtaining efficient results. Comprehensive tables of testing constants and formula are incorporated in Chapter VIII. Where specific make of meters is mentioned, such reference should not be construed as indicating the superiority of that particular type over others, but iv PREFACE should be considered from the view-point of uniformity of nomenclature in order that comparisons may be made briefly and intelligently. In preparing the contents of this book, the details of electrical design have been intentionally omitted. The authors did not feel that those interested in the general and practical phases of the subject would desire to go deeply into such matters. It has been the urgent endeavor to cover the field thoroughly. Supplementary information pertinent to the subject will be gladly furnished by addressing the authors in care of the publishers of this book. We desire to avail ourselves of the opportunity to thank manufacturers of meters referred to in this pub- lication for their generous and able co-operation. We are also indebted to Mr. F. G. Vaughen, Mr. O. A. Knopp and Mr. F. E. Geibel and others for their liberal advice and assistance. THE AUTHORS. San Francisco, June, 1910 TABLE OF CONTENTS CHAPTER I GENERAL PAGE Relation of the Meter to the Central Station 1 The Selection of Meters 3 Factors Affecting the Meter's Accuracy 3 General Construction 7 CHAPTER II MEASUREMENT OF POWER Graphical Representation of Alternating Currents 15 Connections of Indicating Instruments 18 Equations of Power in Alternating Current Circuits 17 CHAPTER III THE INDUCTION MKTKR Reasons for Its Extensive Use 24 Principle of Operation 24 Lagging for Low Power Factor 28 Light Load Adjustment 33 Effect of Frequency Variation 39 Calibration Curves 42 Connections of Single Phase Meters 44 Determination of Power Factor by Means of Two Single Phase Meters , 55 Polyphase Meters Adjustment of Elements 61 Metering High Potential Circuits 63 CHAPTER IV THE COM MUTATING METER Principle of Operation 66 Comparison to a Shunt Motor ; 66 General Construction 68 Use of Commutating Meters on Alternating Currents 75 Three- Wire Meters 77 Switchboard Meters 80 Connections . 83 vi TABLE OF CONTENTS CHAPTER V. MERCURY FLOTATION METER PAGE Principle of Operation 87 Diagrammatic Illustration D. C. Type 89 Diagrammatic Illustration A. C. Type 92 Ampere Hour Meter 93 Connections 94 CHAPTER VI MISCELLANEOUS The Prepayment Meter 96 Maximum Demand Indicators 100 CHAPTER VII MAINTENANCE AND TESTING Reading Meters and Keeping of Records 107 Forms of Record Cards 107 to 113 Installation of Meters 114 Testing With Indicating Instruments 118 Constants and Testing Formulae 119 to 125 Testing With Rotating Standard 125 Testing With Phantom Loads 130 Knopp Method of Testing 132 Special Testing Set for D. C. Meters 137 Shop Methods of Testing 140 Testing Polyphase Meters 144 The Use of Current and Potential Transformers 145 Meter Troubles 150 CHAPTER VIII RATES Commonwealth Edison Company, Chicago 157 Edison Illuminating Company, Boston 165 Birmingham (Alabama) Railway, Light and Power Company. 170 San Francisco (California) Gas and Electric Company 171 APPENDIX Definitions 173 Determination of Temperature Rise by Resistance Method.. 175 Adjusting Meters for Use With Current and Potential Transformers . . 175 THE WATTHOUR METER CHAPTER I. GENERAL. Definition. The name "recording wattmeter" or "integrating wattmeter," is often erroneously applied. The true name for the instrument commonly used for recording the energy flowing in an electrical circuit for a certain period of time is the watt-hour meter, since it records the product of the watts and the time. The "recording wattmeter" in the true sense of the word is the instru- ment which is ordinarily known as the graphic, or "curve-drawing wattmeter/' which records the watts for any given instant without taking into consideration the time element. Relation of the Meter to the Central Station. The relation of the meter and the meter system to the distributing station is a factor of great im- portance, the gravity of which, as a rule, is not fully realized; especially is this true with the small and the medium-sized lighting and power companies. The rev- enue of the distributing company depends on the meter in more ways than are at first apparent,and the con- tinued accuracy of its meters is a matter materially affecting its financial success. Inaccurate meters are eventually detrimental to the interests of the company selling current, regardless of whether the meter runs fast or slow. A fast meter furnishes the consumer a very just cause for complaint, and when detected usually reacts strongly against the company in produc- ing mistrust of its methods and a general feeling among its customers that they are paying for some- thing that they never receive. Such a feeling is to be 2 THE WATTHOUR METER avoided by every possible means, as it causes endless complaints and in many cases the loss of customers with the resulting loss in revenue. Slow meters, of course, act directly on the com- pany's revenue, failing to record the power which is actually being delivered. This is often a very serious source of loss, especially where meters are operating at light load for a considerable portion of the time, as is almost always the case under commercial conditions. It is this inaccuracy in meters at light loads that con- stitutes, in the majority of cases, the chief source of loss to the distributing company, and especially is this true where there is no attempt made to periodically test the meters and make any minor adjustments that may be necessary. Meters are often installed under conditions that are by no means the most favorable for a delicate piece of apparatus ; this however, is fre- quently unavoidable, as the meter must be installed wherever power is sold. It is often installed in places which are inaccessable, allowed to become covered with dust and dirt, and in some cases it is placed where it is subjected to severe and continual vibrations; it is usually then left to take care of itself, receiving no fur- ther attention than to be read once a month. Under such conditions it is almost inevitable that the meter will eventually run slow, especially on light loads. In carefully managed and well designed direct cur- rent systems, the energy lost in line drop and other- wise unaccounted for between the station bus-bars and the consumer's meters may be as low as 15 per cent, but on alternating current systems, having- many small transformers connected to the lines which are contin- ually consuming power in the form of core loss, and with meters which are poorly maintained or entirely neglected, the loss shown by the comparison of the reading of the station meters and the consumer's meters may be as high as 70%. From 15% to 20% represents very good practice on direct current systems, and from 20% to %3O on alternating current systems. GENERAL 3 The Selection of Meters. The selection of meters is a question which should be thoroughly investigated. While there are several excellent makes of watt-hour meters on the American market, there are still others which may be disastrous to the revenue of the distributing company. It is not always the meter which when new shows itself capa- ble of finer adjustments and consequent high initial accuracy that will prove the most satisfactory or the most accurate after a period of service under average commercial conditions. Of course initial accuracy is an important factor, but it should not be sought at the expense of continued accuracy. The meter should be of as substantial and rugged construction as is con- sistent with efficient design. Such a meter will prove to be more satisfactory and will show less error after a period of service than will a meter of more deli- cate construction, although when new it can be ad- justed to a finer degree. This question of continued accuracy is of paramount importance and should always be borne in mind while selecting the instru- ment upon which the revenue of the company is to depend. Factors Affecting a Meter's Accuracy. The factors affecting the accuracy of a watt-hour meter are various, but the two principle ones are fric- tion and the weakening of the permanent magnets. If these two factors could be eliminated, a meter once accurately adjusted would remain so indefinitely. Un- fortunately, however, these two factors do play a very serious part in the performance of the meter, the most serious being friction. If the friction component was a constant quantity it could be compensated for by the light load adjustment device and thus permanently eliminated as regards the meter's accuracy. It has been found though that friction is an extremely varia- ble quantity and in the case of any motor-meter it may vary by quite a large amount, even under very favorable conditions. For this reason a high value 4 THE WATTHOUR METER of the torque, or turning effort is very desirable, since with a high torque the percentage of this torque re- quired to overcome any increase in friction is rela- tively small, and the percentage increase of effective torque due to any decrease in friction is also corres- pondingly small. Thus it will be seen that a meter having a high torque will not suffer in accuracy nearly so much for the same amount of change in friction as will a meter of low torque. There is, however, a value for the torque, which if exceeded, will result in poor economy, because by increasing the losses a higher value of torque can be produced. It can therefore be readily seen that the design of a meter should be such that this ratio of torque to watts loss will be at the most economical point. Since friction is the most serious factor affect- ing the accuracy of a watt-hour meter, it is essential that every care and precaution be taken both in the design and the manufacture to insure low initial fric- tion, and to insure as far as possible against changes in friction after the meter has been in service for some length of time. Friction will develop in the lower jewel bearing, in the upper bearing and in the record- ing mechanism. The Jewel Bearing. In order to obtain low friction in the jewel bear- ing the revolving element should be light in weight; only the highest grade of jewels should be used, and they should be carefully selected and ground. The pivots, or the bearing points, should be of the finest grain of glass-hardened steel. It is usual practice of manufacturers to mount the jewel on a spring sup- port, thus taking up any sudden vibrations and thereby preventing excessive pressure between the jewel and the bearing point, therefore prolonging the life of each. Although the actual weight supported by the lower bearing* is small, the pressure between the jewel and the pivot in a meter is great, since the actual contact area is exceedingly small, being as it is, almost a GENERAL 5 "point" contact, so that the pressure per square inch of contact reaches an extremely high value. It is for this reason that jewels of the best quality and "glass- hardened" steel pivots are necessary in the construc- tion of the lower bearing, as any other material would quickly break down and develop excessive friction. The otherwise objectionable "point" contact between the jewel and the pivot is necessary in order that low initial friction may be secured. Recording Mechanism. When properly and carefully made, the record- ing mechanism is not subject to the variations in fric- tion which occur in the bearing, and it can therefore be much more completely compensated for by means of the light load adjustment device of the meter. Only machine cut gears should be used in the construction of the recording mechanism, and during the course of manufacture every precaution should be taken to see that the gears and their bearings are in perfect condition and free from all burrs ; even the slightest burr or imperfection in the individual gears will prove to be a source of friction variation, and as some of the gears move very slowly and as friction variation would only appear when the imperfect portion was in mesh, the only feasible way of detecting and preventing this source of future error in the meter is by rigid factory inspection of all parts which enter into the construc- tion of the recording mechanism. Weakening of the Permanent Magnets. The next important factor affecting the contin- ued accuracy of the watthour meter, and one of a very serious nature, is the weakening of the permanent mag- nets, often called the "retarding magnets." The only insurance which the purchaser has against a poor grade of permanent magnet is the ability .and the experience of the manufacturer. It sometimes hap- pens that meters, especially for switchboard service, are installed where they are subjected to the influence 6 THE WATTHOUR METER of powerful "stray fields" which may be set up, due to the proximity of wires or bus-bars carrying heavy currents. To nullify the effects of such stray fields on the retarding magnets, some manufacturers arrange the magnets astatically, that is, they are placed so that any stray field which will tend to weaken one magnet will correspondingly strengthen another, and vice versa. Creeping. Under the same category as fast meters comes the "creeping" of meters. It is sometimes found that meters will run slowly, or "creep" when there is no current flowing in the series fields, the potential circuit alone, being energized. This is due to the light load adjustment exerting more than enough torque to over- come the friction, and may be due to one or more causes. The light load adjustment may be so set that it just compensates for the initial friction and the meter then installed where it is subject to continual vibra- tion, under which condition, the "friction torque" is reduced and the meter will creep. Again, the meter may be on a circuit where the voltage is above normal, which will tend to produce creeping. As a general rule, meters are so adjusted at the factory as to allow for a range of several per cent in voltage without causing creeping; such practice is to be recommended, as the slight benefit to be derived from having the friction completely compensated for is more than counterbal- anced by the trouble due to creeping when such fine adjustments are made. Overmetering. Another frequent and easily avoidable source of loss to the distributing company is "over-metering." It often happens that in the case of public buildings, theaters and other places where there is a large "con- nected" load, and where for a greater part of the time only a small part of this connected load is actually tak- ing current, that one large meter of sufficient capacity GENERAL 7 to take care of the entire installation is employed. When this is done the large meter will operate the majority of the time on light load, and for a consid- erable portion of the time it may be operating on very light loads, and since no commercial meter can be re- lied upon to continuously record such light loads with the same accuracy as at or near full load, there will re- sult a considerable loss from the practice of "over- metering." It is often much better to install a smaller meter to take care of such loads, even at the risk of an occasional burn-out. Practically all standard meters will carry a considerable overload for short periods, and will carry as much or more than 25% overload continuously when located in cool, dry places. It is not recommended that meters be worked at overloads continuously, but in many instances it will be economy to have them work at overloads during the period of maximum demand. By exercising a little judgment a meter can be so selected that it will never be excess- ively overloaded, but which will be small enough to give a fair degree of accuracy during the light load period. Where the ratio of the connected load to the aver- age actual load is large, it is better to subdivide the circuits and install two or more meters than to at- tempt to handle the entire load on one meter. In this way it can be so arranged that while there is no dan- ger of a meter being severely overloaded, it will still be small enough to accurately record the power during the light load period. General Construction. The different types of meters will be dealt with separately hereafter, therefore we will take up at this point the various parts which are common to all types. Frames and Covers: The supporting frames to which the mechanism is secured should be rigidly con- structed from a mechanical standpoint, and the ma- terial used should be non-magnetic. The covers should be of sufficient rigidity to protect the meter from ordi- 8 THE WATTHOUR METER nary mechanical injury, and should also be light ; the composition known as "white metal" is a good material, being used either in its natural finish or with a coat- ing of dull black japan. In some cases glass is used for the covers so that all working parts of the meter may be superficially inspected without removing the O Fig. 1. Supporting Lugs of Base Frame. cover. For switchboard meters, glass is a sat- isfactory material for the covers, but for ordinary house type meters the metal covers are to be recom- mended ; glass covers, exposing all parts of the interior to view, may tend to invite tampering by unauthor- ized persons. The internal frame which actually sup- ports the bearings and other parts of the meter proper GENERAL 9 should also be made of non-magnetic material. In the ordinary type of "commutating" meter the construction of the internal frame is such that when used on alter- nating currents it is often the case that heavy eddy currents may be induced in the frame by "the rapid reversals of the "projected" field. Such eddy currents cause undue heating, and to obviate this some manu- facturers split the frame and insert a piece of fibre or other insulating material. The base frame is usually furnished with three supporting lugs as shown in Fig. i, the top lug being key-holed and the lower right hand one slotted, thus allowing the meter to be rapidly hung in place. It can then be properly levelled and set and the sup- porting screws driven home. The removable covers are usually held in position by two or more studs which are fastened to the base frame and which project up through the covers ; wing- nuts having holes through their bodies, through which seal wires may be passed, are used on the studs to securely hold the covers in place. The groove in the base frame into which the covers fit, should be pro- vided with felt gaskets to exclude dust, moisture and insects from the interior of the meter. The holes for the entrance and exit of service wires should also be provided with a dust proof feature, and the dial win- dow should be set in putty or other suitable material. The Top Bearing: The top bearing of a meter is necessarily simple, as it does not have to support any weight, but simply acts as a guide bearing, and may be the same as or similar to either of the two types shown in Fig. 2. The Shaft: The shaft should be made as light as is consistent with good design, and is usually made of steel approximately y%-'m. in diameter, some manu- facturers using a solid shaft, and others a tubular form. At the top of the shaft is mounted the "worm" gear which transmits the motion of the shaft to the record- ing mechanism. There are two general methods of constructing the worm; one consists of cutting it 10 THE WATTHOUR METER directly into the steel shaft, the other method being to mount a worm of composition material in the end of the shaft. This latter method possesses the ad- vantage of allowing the use of a non-rusting material. On the lower extremity of the shaft is mounted the removable pivot. Discs: Until several years ago, the meter discs were made almost exclusively of copper on account of its high conductivity, but aluminum has practically superseded copper for this purpose, due to its lighter weight. An aluminum disc having the same conduc- tivity as a copper disc will weigh only about 48% as Discs of Billiard Cloth oiked In Jewelers Oil Fig. 2. Types of Top Bearing. much as the copper. Therefore the aluminum, though of greater thickness, is much more desirable. The question is often asked : "Why are most meter discs roughened or covered with little holes which resemble prick-punch marks?" This has nothing whatever to do with the electrical characteristics of the meter, as is sometimes supposed, but simply results from a fac- tory method of producing a plane surface. The disc is placed on a heavy metal block, and a weight having a roughened surface is allowed to fall upon the disc, thus producing the peculiar marking. It has been found that this process eliminates any trouble which may be due to the warping of the disc. The Lower Bearing: There are at the present time two general types of lower bearings in use ; the pivot and jewel type as shown in Fig. 3, and the ball and jewel type as shown in Fig. 4. The ball bearing GENERAL 11 is relatively a new departure, but in reality it is essen- tially a "pivot" bearing also, and as far as the com- parative friction is concerned, they are, it is safe to say, about equal. So long as the ball remains perfectly smooth and free from rus+ it serves its purpose ad- mirably. Fig. 3. Pivot and Jewel Type of Lower Bearing. Jewels: It has been found that there are but two kinds of jewels which are satisfactory for use in the lower bearings of meters, they being the diamond and the selected eastern sapphire. In self-contained meters, up to and including 50 k. w. capacity, the sapphire is generally used to the best advantage ; above 12 THE WATTHOUR METER this value, it is advisable to use the diamond, because of its unequalled hardness. Where great accuracy is desired in the case of switchboard meters in central stations, it is often desirable to use diamond jewels in meters of as small a capacity as 5 amperes. In all cases, the jewel should be carefully selected, ground and polished, and should be free from all flaws. It has been noted that under normal conditions, the average sapphire jewel will stand as much or more than 600,000 revolutions of the shaft, and in some Figr, 4. Ball and Jewel Type of Lower Bearing. cases the diamond jewel has lasted for as many as 35,000,000 revolutions of the shaft. These values, how- ever, are extremely variable and depend to a great ex- tent upon the conditions and care under which the meter operates. The Retarding Magnets: It is of the utmost im- portance that the strength of the retarding magnets be as permanent as is possible to make them, since their retarding or "dragging" effect is proportional to the square of their magnetic strength. Therefore a slight change in the strength will have an appreciable GENERAL 13 effect upon the speed of the disc. Much depends upon the physical properties of the steel from which the magnets are manufactured, and the most rigid inspection, by both chemical and physical analyses should be made of each lot of steel before it is treated for use as meter magnets. The manufacturer, after he has given the steel a special process of treatment, hardens, forms and magnetizes the product. The com- pleted magnet is then subjected to hammer blows to detect any mechanical imperfections, and if it should fail to "ring true" is rejected. It then undergoes an artificial aging process ; accurate measurements of magnetic strength being made at frequent intervals. It is then laid away for several months after which the strength is again measured and if this latter meas- urement differs in the least from its strength when first laid away it is discarded. A very successful process of magnetizing meter magnets consists in slipping the completed form over a copper bar, through which a heavy current of elec- tricity (many thousands of amperes) is passed mo- mentarily. In this way a great number of forms can be magnetized at the same time, and a uniform strength produced. Great care is taken in the manu- ture of the permanent magnets, and as a rule the re- sults are very satisfactory. The Recording Mechanism : As previously pointed out, the recording mechanism should be manufactured with the greatest care, and rigid factory inspection is practically the only safeguard against imperfections. Meters are often placed in such positions that the meter reader will encounter reflected light, and for this reason it will generally be found that a dial of unglazed material will be less difficult to read under all conditions. The recording mechanism should be so constructed and provided with such dowel pins that it can be removed from the meter at any time and then replaced in the exact position from which it was orig- inally taken without disturbing in the least the mesh of the worm with the first gear. 14 THE WATTHOUR METER From the foregoing description of the general con- struction, a very good idea can be gathered as to the mechanical requirements of a good meter; a study of the subsequent chapters treat of the electrical char- acteristics. There are still to be found throughout the coun- try a number of the very old type "low efficiency" meters, and a great many of the ampere-hour type ; it is the recommendation of the authors to replace such old metors with some make of good modern watt- hour meter, as in the majority of cases the increased revenue to be derived will pay many times for the in- terest on the cost of the exchange. CHAPTER II. THE MEASUREMENT OF POWER. The power in a direct current circuit is equal to the product of the electro motiveforce and the current ; in other words, if I represents the current in amperes, and E, the e.m.f. in volts, then the El Watts, W = El, or the kilowatts 1000 Fig. 5. The power flowing in a direct current circuit can therefore be determined by the use of a voltmeter and an ammeter, or by one instrument, an indicating watt- meter, which will indicate the product of the volts and amperes. The power flowing in an alternating current cir- cuit is dependent not only upon the e.m.f. and the current, but also upon the power factor of the circuit. This is evident as is illustrated by Fig. 5, which shows 16 THE WATTHOUR METER a sine wave of e.m.f and current at unity power fac- tor. In Fig. 6 is shown the same current and e.m.f. but with a power factor of 50 per cent instead of unity. The instantaneous value of the power flowing in any circuit is equal to the product of the instantaneous value of the e.m.f., and the instantaneous value of the current. The curve P represents these instantaneous values of the power. It will be noted that in the case of unity power factor (Fig. 5), the curve P is entirely above the axis, that is the line of zero value ; this indi- cates that the power is all flowing in one direction. It Fig. 6. will also be noted that the maximum value of the e. m. f. occurs at the same instant as the maximum current, which condition gives the maximum value of the power for these values of the current and the e.m.f., as can be seen from the figure. Referring to Fig. 6, it will be noted that for a power factor of 50 per cent, part of the curve, P, ?.s below the axis, which indicates that the power is not all flowing in the same direction, but that during a part of the cycle a portion of the power is actually being "pumped back" into the circuit. The net value of the power supplied is equal to the difference be- tween that represented by the area enclosed by the THE MEASUREMENT OF POWER 17 curve, P, which is above the axis and the area en- closed by that part of the curve which is below the axis. Assuming a sine wave of e. m. f., and of current (modern commercial alternating current generators give waves closely approximating a sine wave), and. denoting the maximum value of the e.m.f. by E, the maximum value of the current by I, and the instan- taneous value of the e.m.f. by e, we have e = E sin , where < = w t, in which 6), where 6 = the angle of phase displacement between the current and the e. m. f. The instanta- neous power, p, is equal to the product of the in- stantaneous e. m. f. and the instantaneous current, or p = e i, = E sin I sin (< 6) or p = El cos sin 2 < El sin 6 sin cos . Let P = the average value of p, then P = av'g (El cos 6 sin 2 < El sin sin $ cos ) = El cos 6 (av'g sin *) El sin 6 av'g (sin cos <). The average value of sin ^ = ^2, and the aver- age value of sin < cos = O, substituting these average values in the above equation, we have __ El cos 6 2 But E, the maximum value of the e.m.f. wave=V 2 E, where E is the effective value of the e.m.f. Also, if I denotes the effective current, the maximum current, I = V 2 I- Therefore, we have the fundamental for- mula : 18 THE WATTHOUR METER P = El cos 0; the cos being the power factor of the circuit. If the power factor is unity, then cos 0= i, and hence the above equation becomes P = El, which, as will be noted, is the same as for direct current. The power factor is very seldom as high as unity, and it is therefore almost always necessary to use a watt- meter rather than a voltmeter and ammeter; a prop- erly constructed arid accurately calibrated wattmeter m "SET B 1 i I i 1 1 Sot Fig. 7. will measure power correctly regardless of the value, of the power factor. The power factor of a single phase alternating current can be easily obtained by taking the product of the volts and the amperes as indicated by a voltmeter and ammeter and dividing this result into the actual power reading as indicated by a wattmeter. The actuating force in an indicating wattmeter is derived from two sets of coils, one being connected in multiple, and the other in series (as in the case of THE MEASUREMENT OF POWER 19 the watthour meter) with the load to be measured. The reaction between these two coils is at each instant proportional to the instantaneous values of the current and the e. m. f., so that the total deflecting force acting on the pointer of the instrument is at all times propor- tional to the true power. A two-phase system (often called "quarter phase"), can be considered as two single phase sys- tems, and the power being supplied by such a system is simply the sum of the power flowing in the two Sou fee J_ ri i i - =2 J Fig. 8. equivalent single-phase systems, and can be measured by a single-phase wattmeter in each system, or by one polyphase wattmeter as shown in Fig. 7, in which lines i and 3 constitute one-phase and 2 and 4 the other phase. The power in a two-phase three-wire system can also be measured by two single-phase wattmeters or by one polyphase wattmeter, the connections being made as shown in Fig. 8, in which line number 2 car- ries the resultant current. Fig. 9 shows the connec- tions used when measuring power in a balanced two- 20 THE WATTHOUR METER phase three-wire system with one single-phase meter. In this case the voltage impressed on the meter will be V 2 or 1.41 times the voltage of either phase, and when the system is balanced the current flowing in the line, 2, will also be V 2 , or 1.41 times the current Scrasce Fig. 9. Fig. 10. Fig. 11. in either phase. The one wattmeter method will meas- ure the true power only when the phases are per- fectly balanced, and is therefore very seldom used. The power flowing in a three-phase system can be measured by two single-phase meters connected as shown in Fig. 10, or by one polyphase meter con- nected as shown in Fig. n. THE MEASUREMENT OF POWER 21 The power in a three-phase four-wire system can be measured by three single-phase wattmeters con- nected as shown in Fig. 12; the three-phase four-wire system being virtually three single-phase systems. The Fig. 12. total power will be the sum of the indications of the three meters. The power in a three-phase four-wire system can also be measured by two single-phase meters connected as shown at (a) in Fig. 13, or with one polyphase meter in conjunction with current (or series) transformers connected as shown at (b) in Fig. 13- Fig. 13a. The power flowing in a three-phase system is ex- pressed by the equation, P = V 3 El, cos B, where E is the voltage between the phases, I the current per leg and cos (9, the power factor of the circuit, When the system is not balanced the average values 22 THE WATTHOUR METER of the current, the voltage and the power factor should be used in the above equation, remembering' that 6 is the angular displacement between the line current and the voltage between line and neutral. Fig. 13b. Fig. 14 shows the method of connecting one single phase wattmeter for measuring the power in a bal anced three-phase three-wire system. Sot/s-ce Fig. 14. Power is the rate at which energy is supplied. Electrical power is measured in watts and kilowatts, and electrical energy is measured in watt-hours and kilowatt-hours. "Purchasers of power" are in reality purchasers of energy, and in order to determine the energy flowing in a circuit it is necessary that the THE MEASUREMENT OF POWER 23 power be multiplied by the time. If w = the power in kilowatts, and t = the time in hours during which the power is flowing, then the energy = w t = kilo- watt-hours. Since it is energy and not power which is bought, it is necessary to have an instrument which will take into consideration the time element ; such an instrument is the watthour meter. CHAPTER III. THE INDUCTION METER. At the present time the induction type of watt- hour meter is used almost exclusively where alternat- ing currents are concerned, and as alternating current is much more extensively used for general lighting and power distribution than is direct current, there are considerably more induction meters being manufac- tured than there are of any other type. Reasons for Its Extensive Use. Some of the principle reasons for this almost ex- clusive use of the induction meter on alternating cur- rent circuits are as follows : The induction meter is more rugged in design, having no brushes, no com- mutator, or other moving contacts. The revolving ele- ment consists simply of the shaft and revolving disc, all windings being on the stationary element. The weight of the moving element being less than that of the commutating type of meter, and the fact that it has no commutator with its resulting friction, necessarily eliminates an appreciable amount of fric- tion and also results in less jewel wear. For the above reasons, the induction meter will maintain its accuracy better with the same amount of attention than will other types of "motor" meters. The induction meter is entirely free from commu- tator and brush troubles, having neither brushes nor commutator. It is cheaper in first cost than any other type of meter, suitable for use on alternating currents, which can compare with it in continued accuracy. Principle of Operation. The induction meter consists essentially of the stationary element, the rotating element (consisting THE INDUCTION METER 25 merely of the shaft and disc), the recording mechanism, the jewel bearing and the retarding magnets. The stationary element consists of the magnetic circuit, A, Fig. 15, which is built up of laminated steel punchings ; the current coils, B ; the potential coil, C ; the light load adjustment, D ; and the lagging coil, E. The current and potential coils are mounted as shown in the figure, in such a way that the magnetic flux set Fig. 15. up by each of these coils will pass through the meter disc, and this alternating flux passing through the me- tallic disc will set up currents therein which will flow as indicated in Fig. 16, the disc acting virtually as the short-circuited secondary of a transformer. It will be seen from Fig. 16 that the currents set up in the disc by the potential coil P flow past the poles of the current coils, P 1 , and that the currents set up by the 26 THE WATTHOUR METER current coils flow past the pole of the potential coil. These currents set up in the disc are in phase with the voltages producing them, since the circuit offered by the disc itself is non-inductive. The voltages jn the disc which produce these currents, however, lag 90 degrees behind the fluxes set up by the coils on the stationary element, as an induced voltage is always 90 degrees behind the inducing flux. The flux is in phase with the current which produces it, the angle of hysteretic lag being negligible, so that we have currents flowing in the disc lagging 90 degrees behind the currents flowing in the meter windings. Fig, 16. The potential coil is wound with many turns of fine wire, and is therefore highly inductive, so that the current flowing in this coil is practically 90 degrees be- hind the impressed e.m.f, and the flux from the pole of the potential coil is brought to exactly 90 degrees behind the impressed e.m.f. by use of the lagging coil, as will be explained later. The flux from the poles of the current coils will be in phase with the current, and therefore in the case of a load of unity power factor will be in phase with the impressed e.m.f. It can be readily seen from this that in the case of unity power factor the current set up in the disc by the~~ptential coil (which lags 90 degrees behind the flux from the potential coil), will be in phase with the flux from the THE INDUCTION METER 27 current coils, and also that the current set up in the disc by the current coils will be in phase with the flux from the potential coil. It will further be seen by referring to Fig. 16 that the disc currents set up by the potential coil will flow past the center of the cur- rent coil poles, and that the disc current set up by the current coils will flow past the center of the potential coil pole. This will give rise to a mechanical force tending to cause the disc to revolve, since any con- ductor carrying current at right angles to a magnetic field is subjected to a force which tends to move the conductor out of such field. Furthermore, this force is proportional to both the current flowing in the disc and to the field strength or to the product of these two factors. In the case of the meter the current flow- ing under the pole of the potential coil is proportional to the line current, and the flux is proportional to the impressed e.m.f. Similarly, the current flowing under the poles of the current coils is proportional to the im- pressed e.m.f. and the flux from the current coil poles is proportional to the line current. The force tending to revolve the disc is therefore proportional to twice the product of the current and the voltage, or what is the same thing, it is proportional to the product of the current and voltage, or to the watts. The principle of the induction meter's operation may be explained in a somewhat different way, which is perhaps more clearly understood ; that is, the electri- cal element may be considered as the stator of an in- duction motor and the disc as the rotor. The "shifting" magnetic field in the case of a meter (which corre- sponds to the "revolving" magnetic field of the motor), is supplied by the current coil and the potential coil poles, the flux from the potential coil pole being 90 degrees out of phase with the flux from the current coil poles, as previously explained. This "shifting" magnetic field sets up currents in the meter disc, which reacting with the magnetic field produces a force tend- ing to rotate the disc, exactly as the "revolving" field of an induction motor sets up currents in the rotor, 28 THE WATTHOUR METER which reacting with the "revolving" magnetic field pro- duces a torque which causes the motor to run. In the case of power factors which are other than unity, the flux produced by the current coils (which flux is in phase with the current), will no longer be 90 degrees out of phase with the flux from the potential coil, but will be 90 degrees plus or minus the angle of current displacement or the angle by which the cur- rent is out -of phase. This being the case, the disc currents set up by these coils will no longer be in phase with the flux from the poles under which they flow, but will be out of phase by the angle of current dis- placement. The force tending to turn the disc will therefore no longer be directly proportional to the product of the current and the flux, but it will now be proportional to the product of the current, the flux and the cosine of the angle of current displacement, which is the power factor. Therefore the meter will still register the true watt-hours. Another way of expressing this is to consider that the force acting on the disc will be proportional to the product of the flux and the component of the disc current which is in phase with the flux. Since the disc current is out of phase with the flux by the angle of current displacement, the component of the disc cur- rent in phase with the flux is equal to the total disc current multiplied by the cosine of the angle of dis- placement, or the power factor. Lagging for Low Power Factor. In order for the meter to register correctly on low .power factors it is necessary for the flux from the pole of the potential coil to be exactly 90 degrees behind the impressed e.m.f. If the flux from the potential pole is less than 90 degrees behind the impressed e.m.f. the meter will run slow on lagging and fast on leading cur- rents, while if the flux lags more than 90 degrees it will run fast on lagging and slow on leading currents. This condition is obtained by a method known as lagging,. and is accomplished as follows: In figure 15 THE INDUCTION METER 29 C is the potential coil, and E is the lagging coil which is mounted over the pole tip of the potential coil. The current in the potential coil will be not quite 90 degrees behind the impressed e.m.f., due to fhe RI 2 losses in the winding and the losses in the iron which give rise to an energy component of the current. The flux will be in phase with the current, and will therefore be not quite 90 degrees behind the impressed e.m.f. A part of this Fig. 17. flux will pass through the lagging coil and on through the meter disc. This flux induces an e.m.f. in the lag- ging coil which is 90 degrees behind it in phase. This is shown by the vector diagram, Fig. 17. In this diagram OE represents the impressed e.m.f. and OI the current in the potential coil which lags not quite 90 degrees behind this e.m.f. O< represents the flux set up by the current, which passes through the lagging 30 THE WATTHOUR METER coil and meter disc. This flux induces the e.m.f., OE', in the lagging coil, which is 90 degrees behind it in phase. The circuit of the lagging coil is closed through a resistance, the amount of which can be varied and therefore the amount of current flowing in this circuit can be varied. This current is represented in the diagram by OI'. The current OF will set up a flux O^> in phase with itself, and this will combine with the flux O<, producing the resultant flux, O<-% which will pass through the meter disc. It can be readily seen by reference to the figure that if the current, OI' is of the proper value, that this resultant flux Ofa will be exactly 90 degrees behind the impressed e.m.f., OE. By adjusting the amount of non-inductive resistance in the circuit of the lagging coil, this condition can be very easily produced, which process is known as "lag- ging." A properly lagged meter will register with accuracy on low power factor. The method of lagging above described is used in meters manufactured by the General Electric Company. The method of lagging which is employed in meters manufactured by the Westinghouse Electric and Manufacturing Company is somewhat different from that which has just been explained, though the principle is essentially the same. In the Westinghouse meter the lagging coil consists of an adjustable short- circuited turn, placed on the pole tip of the potential coil. The position of this turn can be adjusted so as to obtain the required flux component to bring the resultant flux 90 degrees behind the impressed e.m.f. By referring to Fig. 18 it will be seen how this is accomplished. OE represents the impressed e.m.f., OI, the current in the potential coil, OE', the voltage induced in the short-circuited lagging turn, OF, the corresponding current, and O' the flux set up by this current. Ofa represents the resultant flux which lags 90 degrees behind the line e.m.f. The proper value of the flux, O(f>' can be obtained by varying the position of the short-circuited turn. In the induction meter manufactured by the Fort THE INDUCTION METER 31 Fits. 18. \ V Fiu. 19. 32 THE WATTHOUR METER Wayne Electric Works, the lagging device consists of two elements, one being wound on the light load adjusting arm (shown at G, Fig. 19), and is connected in series with the lagging resistance, H. This coil and resistance is shunted across a portion of the potential winding as shown. The other coil, E, is wound on the potential pole tip and is short-circuited through a resistance, L. In the vector diagram, Fig. 20, OE is the impressed e.m.f., OI, the current flowing in the poten- tial circuit, which lags not quite 90 degrees behind the voltage, and O< is the flux produced by this current. Fig, 20. This flux produces the voltage OEa in the lagging coil, E, which in turn sets up the current Ok, and the flux, O$2. The current induced in the meter disc by the potential coil also sets up a flux, O< d , in phase with itself. The resultant of Ofa and O< d , which is repre- sented by O/\sc. Fig. 24. described for the General Electric meter. Figure 23 is an illustration of the Fort Wayne company's single THE INDUCTION METER 37 phase meter, and, as will be noted, the disc has a peculiar "cup-shaped" form. The illustration at (b) shows the ease with which the disc may be removed without disturbing other parts of the meter. The light load adjusting device of the Westing- house induction meter consists of two adjustable short- circuited turns so mounted that they may be rotated Fiy. 25a. r through a small angle. One side of each of these short- circuited turns is in an air-gap in the magnetic circuit of the potential winding and by partially rotating the turn it can be made to enclose more or less lines of magnetism, as can be readily seen from the diagram in Fig. 24 at A. The lines of magnetism, in passing through the short-circuited turns, induce currents therein, which currents set up an auxiliary field. This 38 THE WATTHOUR METER auxiliary field is out of phase with the main field from the potential pole, and the two, acting in conjunction, produce a torque on the meter disc, the amount of which can be varied by moving the short-circuited turns so that they will embrace more or less of the flux passing through the air-gap. In Fig. 25 (a and b) are shown two views of the single phase type of induction meter as manufactured by the Westinghouse company. The object of the light load adjusting device is to Fisr. 25b produce a torque from the potential circuit alone (inde- pendent of the load on the meter), the magnitude of which will be just enough to overcome the friction of the meter, therefore rendering it accurate on light loads. Creeping. If the light load adjustment is set so as to exert a torque greater than is actually necessary to^overcome the' friction it will cause "creeping" on no load. THE INDUCTION METER 39 Creeping will also result if the light load adjustment is properly set for operation at normal voltage and then the meter installed on a circuit where the voltage is considerably above the normal voltage rating of the meter. A higher voltage will produce a higher flux from the potential pole, which in turn will induce a higher current in the light load adjusting coil, and this higher current and higher flux will mutually react and produce a higher no load torque, thereby causing the meter to creep. Effect of Frequency Variations. When an induction meter is operated on a fre- quency other than that for which it is adjusted, the lagging coil will no longer set up just the necessary flux to bring the resultant flux from the potential pole exactly 90 degrees behind the impressed e.m.f. ; it will either be ahead or behind this correct 90 degree posi- tion, depending upon whether the frequency is below or above the normal value. Errors from this source will be inappreciable so long as the frequency is within 10% (approximately) of the normal value. It is at the present time the practice of the leading meter manufacturers to design their 125 cycle and their 133 cycle meters so that by a simple connection or adjustment, which can be easily made, they may be used with accuracy on 60 cycle circuits. This is on account of the fact that 60 cycles is the standard light- ing and power frequency, and as the majority of the higher frequency plants will sooner or later be changed over to 60 cycles, it will evidently be a great saving to them if they can use their old meters rather than have to purchase new 60 cycle meters when such a change may be made. Meters so constructed are known as "double lagged" meters, since they are lagged at the factory for two different frequencies. The effect of a frequency other than normal can be best shown by reference to the diagram shown in Fig. 26, in which 40 THE WATTHOUR METER OE=the impressed e.m.f., OI=current in potential coil at normal frequency, OIi=current in potential coil at low frequency, Ol2=current in potential coil at high frequency, OI L , OI L -i and OI L -2 = the currents in the lagging coil for these different frequencies. Fig. 26. Now suppose that the meter is properly lagged for a frequency, f, the current in the potential winding being OI, and the flux therefrom being O<. The cur- rent in the lagging coil will be OI L , and the flux there- THE INDUCTION METER 41 from will be O< L ; these two fluxes, O andO< L com- bine to produce the resultant flux Oc/>R, which is exactly 90 degrees behind the impressed e.m.f. Now suppose that the meter is used on a frequency, fi, which is below normal. With a lower frequency the flux set up by the potential coil will be greater, as the rate of change of flux must remain the same. The magnetizing current, OL, will therefore be greater, the core loss and the RP losses will be higher, so that there will be a larger energy component of the current, and it will therefore not lag by as great an angle as with normal frequency ; the current OL sets up the flux OR-z, which lags too much, being beyond the 90 degree position. In order for the resultant flux to lag to the correct position, it will therefore be necessary for the lagging coil to set up a flux less than O< L - 2 ; in other words, the meter would have to be relagged for this higher fre- quency, fa. Obviously, for power factors other than unity, serious errors would be introduced by using a meter adjusted for a frequency different from that of the circuit on which it operates ; the meter might either 42 THE WATTHOUR METER run fast' or slow, depending upon whether it is adjusted for a higher or lower frequency than that of the cir- cuit on which it operates, and upon whether the cur- rent is lagging or leading. The effect of a frequency above normal will be to make the meter run fast on lagging currents and slow on leading currents ; a frequency below normal will cause the meter to run slow on lagging currents and fast on leading currents. For unity power factor there would also be an error introduced, although it would not be so pronounced as in the case of power factors other than unity. In this case only that component of the flux from the potential pole which is in the correct 90 degree position will be effective, so that the phase dis- placement of the resultant flux will tend to make the meter run slow- on any frequency other than that for which the meter is adjusted. The values of the result- ant fluxes are not strictly proportional to the fre- quencies, however, since the component supplied by the lagging coil is not proportional to the frequency and its angular relation to the main component is different for the different frequencies ; also for lower frequencies, the energy component of the voltage is greater and the reactive component is less, due to the increased shunt current which tends to make the meter run slow, and vice versa for higher frequencies. The currents induced in the meter disc by the cur- rent coils should be directly proportional to the fre- quency, but due to the demagnetizing effect of these THE INDUCTION METER 43 currents on the current coil poles, this condition is not strictly fulfilled, which causes the meter to have a tendency to run slow on frequencies above normal and fast on frequencies below normal. The resultant effect of the different disturbing factors above mentioned will affect the meter to an extent dependent largely upon the design. Figure 27 is a curve showing the accuracy of a Fig. 28. standard make of induction meter on different fre- quencies at unity power factor, and Fig. 28 shows the load and voltage curves of a standard 5 ampere induc- tion meter operating at normal frequency. Connections of Single Phase Meters. SOL// Fig. 29. Fig. 29 shows the diagram of connections for a 44 THE WATTHOUR METER single phase watt-hour meter when used on a single phase two wire circuit. Fig. 30 shows the connections of two single phase meters connected so as to register the power flowing in a single phase three wire circuit, and Fig. 31 shows one three wire single phase meter connected for the same conditions. The three wire meter in effect is Fig. 30. really two meters with but one disc and one potential coil, but with a current coil in each side of the line. Source Fig. 31. The fact that the three wire meter has but one poten- tial winding will cause an error in its registration if the voltage between each line and the neutral is not the same. The polyphase meter can also be used as a single phase three wire meter, and is not subject THE INDUCTION METER 45 to the error just mentioned, but owing to its greater cost, it is seldom used for this purpose. One single phase two wire meter can also be used to register the power flowing in a single phase three wire system by using it in connection with a special "three wire" current transformer. Such a transformer has two primary windings and one secondary winding. The two primary windings are connected respectively in series with each side of the line, the current in the secondary being proportional to the vector sum of the Fig. 32. currents in the two primaries. The connections of a single phase meter when used with such a transformer are shown in Fig. 32. Single Phase Meters on Polyphase Circuits. Two single phase meters may be used to register the power being supplied by either a two phase or a three phase system. For a two phase four wire sys- tem, one meter should be connected in each phase as 46 THE WATTHOUR METER shown in Fig. 33, and when so connected, each meter will register the power in its respective phase ; the algebraic sum of the readings of the two meters will then be the total power supplied by the two phase system. 4- Source Fig. 33. In the case of a three phase system, the two single phase meters should be connected as shown in 3 Source. Fit. 34. Fig. 34. (Similar connections for potential and cur- rent transformers are shown in the appendix, Figs. 33a, 34a.) The action of the two meters thus connected can best be explained by reference to the vector dia- gram, Fig. 35, in which AC, CB and AB represent the THE INDUCTION METER 47 voltages between the phases 2 and i, i and 3, and 3 and 2 respectively ; also let CO, BO and AO represent the currents in the phases i, 3 and 2 respectively, for the condition of unity power factor, and C'O, B'O and A'O the currents for a power factor other than unity. t Fig. 35. We will first consider the case of a balanced sys- tem. In this case let e represent the e.m.f. between a line and neutral, then e = i/^T when E is the voltage between lines; also let I represent the current per phase in a balanced system. A three phase system may be considered as consisting of three single phase systems with the neutral as a common return; the voltages of each of the single phase systems being 48 THE WATTHOUR METER represented by e, and the current by I. The power in each single phase system will be = (e I) cos 0, where is the angle by which the current is displaced in phase from the voltage, (Fig. 35) ; the power in the three phase system will therefore be the sum of the power in the three equivalent single phase systems, or, numerically, ip P = 3 e I cos 0; and since e = . , we have P = \/3 E I cos 0, which is the funda- mental equation for the power flowing in a three phase system. The two meters connected as shown in Fig. 34 will each have a current I flowing through it, and a voltage E impressed upon its potential winding. In meter No. I, the current is represented by the line CO (Fig. 35), and the voltage by CA; and in meter No. 2 the current is represented by B'O, and the volt- age by AB ; the current being represented as being out of phase by the angle 0. The angle OCA = angle OBA = 30 degrees, which is the angular displacement between the impressed voltage and the line current for unity power factor. For power factors other than unity, this angular displacement is equal to 30 degrees plus or minus the angle 9, and as can be readily seen from the diagram it will be (30 6) for one meter and (30 + 6) for the other meter. The power p', registered by one meter will there- fore be p' = E I cos (30 + 0), and the power p", regis- tered by the other meter will be p" = E I cos (30 6), from which p' + p " = El cos (30 + 0) + El cos (30 0) = El [cos (30 + 0) + cos (30 6} ] = El [2 cos (30 cos 0) ] and since cos 30= I/2\/3, we have p' + P" = El V3 cos 0, which, as shown above, is the THE INDUCTION METER 49 equation for the power flowing in a three phase sys- tem. It is therefore seen that two single phase meters will register correctly the power in a balanced three phase system. An unbalanced three phase system may be con- sidered as consisting of a balanced system with the addition of an unbalancing component of either cur- rent, voltage or both. When using two single phase meters on an unbalanced three phase system, the unbalanced component will be taken care of as follows : Suppose that in addition to the balanced current, there is a current flowing between the phases 2 and 3 (Fig. 35), or between 2 and i ; this current would flow either through meter No. i or meter No. 2, and as the meter through which it would flow has impressed upon it the voltage of the phases between which this current is flowing, the meter would register the power correctly. In the case of an unbalanced current pass- ing between phases 3 and i, such current would flow through both meters, and if this unbalanced current is in phase with the voltage BC, between phases 3 and i, it will be 60 out of phase with the voltage im- pressed upon each meter, and as the cosine of 60 is 1/2, the correct amount of power will be registered, one-half being registered by each meter. If this cur- rent is not in phase with BC, it will be out of phase more than 60 in one meter, and less than 60 in the other meter ; the correct amount of power will still be registered, but it will not still be equally divided between the two meters. The angle by which this unbalanced current will be out of phase in one meter will be (60 + 6), and in the other it will be (60 0), where is the angle of displacement between the un- balanced current and the voltage BC. The power registered by one meter will be = E i cos (60 + 0), where i = the unbalanced current, and that registered by the other would be = E i cos (60 0), and the total unbalanced power would be, 50 THE WATTHOUR METER p E i cos (60 + 0) + E i cos (60 - - 0), = E i (2 cos 60 cos 0),and since cos 6o=i/2, we have p = E i cos 6, which shows that the power would be correctly registered in the case of an unbal- anced current. Unbalanced voltages would be taken care of in a similar manner. An unbalanced voltage across phases i and 2, or across 2 and 3, would directly affect the potential winding of one or the other of the single phase meters. An unbalancing of the voltage across phases 3 and i would affect both meters by distorting the voltage triangle so that the power transmitted would still be correctly registered. Fig. 36. Another method of connecting two single phase meters to register the power in a three phase system in conjunction with current and potential trans- formers is shown in Fig. 36; the relations of the cur- rents and voltages being shown in the vector diagram, Fig. 37. Let I, I' and I" represent the currents in the three legs of a three phase system ; E being the volt- age between lines and e the voltage between any line and neutral. Also let 0, 0', and 6" represent the angles by which the currents are displaced from the position of unity power factor; we will assume the voltage to THE INDUCTION METER 51 be balanced, since this makes the explanation some- what simpler. The true power is P = e I cos + e I' cos 0' -f- e I" cos 0". Meter No. I has currents I and I' flowing through its winding (that is, the resultant of these currents), and the voltage, E, CA, impressed upon it ; it is used with a multiplier of 1/2. Meter No. 2 has the current I" flowing through it, and the voltage Bn, impressed upon it ; Bn = (3/2) e. Let P represent the total power, P' the power regis- tered by meter No. i and P" the power registered by meter No. 2, Fig. 37. P' = El' cos (30 0') + El cos (30 + 0). cos P = y+ P' El' cos (30 0')+ El cos (30 + 0) , 3 T ,, ^ *~2~ C and cos (30 0') = cos 30 cos 0' + sin 30 sin 0' cos (30 + 0) = cos 30 cos sin 30 sin 0; 52 THE WATTHOUR METER - y- * also, cos 30 = ^?-, sin 30 = ^ and e = ^7= 2 1/3 whence?- -^- cos - ^-+-^-cos r 2 2 = V 3 E (I cos + I' cos 0' ) + ~~ (I' sin 0' cos 0"r. The vector sum of the currents in a three phase three wire system is zero, therefore I" cos 6" = I cos (60 0) + T cos (60 + 0'), reducing this we get I'sin Q' I sin 0= p=[ l/ 2 (I cos + J' cos 0') I" r i) cos ^",] substituting this for I' sin - - I sin in the above and substituting K 3 e for E, we derive, P= : ^-e (I COS0 + I' cos 0') + -| - (ycos0 + - /!/ T" /I" \ I ^ C T // /V/ cos - - I cos ) -+- I cos V . whence P = e I cos + e I' cos 0' -f- e I" cos 0". The particular feature of this connection is that it gives an indication of how well the system is balanced; if the system is perfectly balanced the two meters will register the same power, taking into con- sideration, of course, the multiplier of 1/2. This is true with the connection previously described and most often used, only when the power factor is unity. If the system is perfectly balanced, either of the meters can be relied upon to record the total power, THE INDUCTION METER 53 regardless of the value of the power factor, in which case the dials of meter No. I would be read without a multiplier, and meter No. 2 would have a dial multi- plier of 2. Fig. 38. In Fig. 38 is shown the connections of three, single phase meters for measuring the power in three phase, three and four wire systems. The three phase circuit is metered in this case simply as three single phase circuits, the current in each phase being the current in one of the single phase circuits, and the voltage of each single phase circuit being the voltage from the corresponding line to the neutral. The sum of the readings of the three meters will be the kilowatt- hours supplied by the three phase system. The advantage of this connection for the three wire system is that the meters operate under better power factor conditions than with the usual two meter method. With this method the current and e.m.f. of each meter will be in phase when the power factor of the load being metered is unity, while with the two 54 THE WATTHOUR METER meter method the current and e.m.f. are 30 out of phase. Figure 38 (a) shows another method of connect- ing the potential transformers for measuring the power flowing in a three phase three wire system by means of three single phase meters. This connection, with certain primary voltages, permits the use of standard ratio potential transformers, where the connection shown in Fig. 38 will require ratios other than the standard. This method is especially applicable to 2300 volt circuits, the standard potential Fig. 38a. transformer used in this case would be rated 2200 volts primary, and 110/122 volts secondary, the higher secondary voltage (122) being used as is indicated in the figure. With a 2300 volt primary, a secondary voltage of approximately 220 volts would be obtained from the potential transformers, thereby permitting the use of standard 220 volt meters. When meters are R X R' used in this manner, a multiplying factor, z , is THE INDUCTION METER 55 employed in obtaining the total reading, in which R = the ratio of the current transformers and R' = ratio of the potential transformers (= 2200. 122 ' Determination of Power Factor by Means of Two Single Phase Meters. The method of metering with two single phase meters on a balanced three phase system has an ad- vantage over the polyphase meter when it is desired Fig. 38b. to obtain the average power factor of the load, which can be done by applying the following formula: Average Power Factor = 2 I/CP') 2 (P' X P") + (P") 2 where P' and P" represent the readings of the two single phase meters. The deduction of this formula is as follows: In the vector diagram, Fig. 38 (b) AB, AC and BC represent the voltages between the phases 56 THE WATTHOUR METER of a three phase circuit, and OI, OI' and OI" repre- sent the currents in the legs I, 2 and 3 respectively, and which are displaced from the position of unity power factor by the angle 6. Now suppose that the current coil of meter No. i is connected in leg No. 2, and that its potential coil is connected across AB ; also that the current coil of 'meter No. 2 is connected in leg No. 3 and its potential coil across AC. Then the power, P', being registered by meter No. i will be = AB. OF cos <, where < is the angle between AB and OF = (30 + 0), 'being the angle of current dis- placement. The angle OBA is of course 30. Denot- ing the voltage AB by E, and the current OI' by I, we have P' = El cos (30 + 6) and similarly, the power being registered by meter No. 2 will be P" = El cos (30 6), (assuming the system to be balanced). Then by trigonometry we have, cos (30 + 6) = cos 30 cos sin 30 sin cos (30 0) = cos 30 cos 6 -f- sin 30 sin But cos 30 = 1/2 V3^ and sin 30 = 1/2, Therefore cos (30 + 0) = 1/2 y~$cosO 1/2 sin and cos (30 9) = 1/2 V^ cos + 1/2 sin Substituting these values in the above equations for P' and P" we have : P' = El ( l / 2 > X Tcos 1/2 sin 0), hence FI _ = _ __ ( 1/2 V 3 cos 1/2 sin 0) P" = El ( !/ 2 T/Tcos + 1/2 sin 0), hence P" El = - 7= - or since El = El, (1/2 V 3 cos + l/ 2 sin 0) we have P' P" l/ 2 V 3 cos l/ 2 sin l/ 2 V 3 cos + V 2 sin THE INDUCTION METER 57 By trigonometry, sin == V \ cos 2 0, and sub- stituting this value cos I/TP' cos o + p' v \ cos 2 e == p" i/ - - P" V 1 cos 2 0, V 1 cos 2 (?' + P") - (P" P') /T cos 0. Squaring and transposing, we have, cos 2 [ 4 (P') 2 4 P' P" + 4 (P") 2 ] = (P' + P") 2 , Whence P' + P' ^average power factor The true instantaneous power factor can also be determined by this method, using two indicating watt- meters. 58 THE WATTHOUR METER Single Phase Meters for Six Phase Circuits. Three single phase meters can be used for meas- uring the power in a six phase system by connecting them as shown in Fig. 39. Polyphase Meters. The polyphase induction watthour meter for use on either a two or three phase system consists essen- tially of two single phase meter elements mounted one above the other on the same shaft, and having but one register. The principle of operation is iden- tically the same as that previously explained for two single phase meters, except that in the case of the two single phase meters, the algebraic sum of the two registers is taken to obtain the total power, while So. Fig. 40. in the case of the polyphase meter this is automatically accomplished by having the two discs connected to one shaft. The polyphase meter has the advantage of being easy to read and install. The polyphase meter for use on three phase four wire systems when used without current transformers is of somewhat different construction from the meter used on three phase three wire, or on two phase sys- tems. In the three phase four wire meter without cur- rent transformers, it is necessary to have a current winding in the meter for each phase. These wind- ings are arranged on the two elements in such a man- ner that the current in one phase passes through a THE INDUCTION METER 59 winding on each element; the current in each of the other two phases passing through a winding on one element. Fig. 40 shows the vector diagram and also the diagram of connections of this type of meter, in which I, I' and I" represent the currents in phases i, 2 and 3 respectively, and i-n, 2-n and 3~n represent the voltages between the legs i, 2 and 3 and the neutral wire, n. One element of the meter has the voltage 3-n impressed upon its potential winding, and the current I" passing through one set of current coils, and the current I' passing through the other set of current coils. The other element has the voltage i-n im- pressed upon its potential winding, and the current, I passing through one set of current coils, and the cur- rent I' passing through the other set. A three phase four wire system is, in effect, three single phase systems, the current in each system being the current in the corresponding phase, and the volt- age of each system being the voltage between the corresponding phase and the neutral. With the connections described above, the power being transmitted by each of the two single phase sys- tems, 3-n and i-n will be correctly recorded by the meter for both unity power factor and for power fac- tors other than unity, as each element of the meter will act as a single phase meter in recording this power. The power being transmitted by single phase, 2-n will be recorded partly by one meter element and partly by the other. The current I' passes through both meter elements, the connections to its coils being reversed so that for unity power factor it is 60 out of phase with the voltage impressed on each element. (If these coils are not reversed the current, I' will pass through the meter 120 out of phase with the voltages impressed on the two elements and will sub- tract instead of adding the power in phase 2-n.) Since the cosine of 60 = 1/2, one-half of the power will be recorded by each element. For power factors other than unity, one element will record more than half, and the other element will record less 60 THE WATTHOUR METER than half the power being transmitted; the sum of the power recorded by both elements will be equal to the total power. When current transformers are used with three phase four wire meters, the standard polyphase meter as used on three phase three wire circuits may be used, the current transformers being so connected that the resultant current of phases 3 and 2 passes through the current coils of one element, the voltage, 3-n being impressed on this element; the resultant -current of phases I and 2 passes through the current C-r \ t | 2 -E i _ r-| 3 , - CT. J J_ 1 , j e j ! ! Fiff. 41. coil of the other element which has the voltage i-n impressed upon it. The action of this type of meter is the same as just described for the meter without current trans- formers. In the latter case, two sets of current coils are used on each meter element, the resultant effect of the currents in these two sets of coils being the same as the resultant of the currents from two current transformers passing through one set of current coils. Fig. 41 shows the proper connections for a poly- phase meter when used on a three phase four wire system in conjunction with current transformers. THE INDUCTION METER 61 Adjustment of Elements. Polyphase meters should be provided with some means of adjusting the torque of one of the elements without disturbing the other. This is necessary be- cause there is only one retarding system which is common to both elements, and it is therefore necessary that some means be provided so that the two electrical elements may be adjusted to give the same torque when the same amount of power is passing through each. This adjustment is readily accomplished by changing the number of turns on the potential wind- ing of one of the elements. By this means, the torque of that particular element can be adjusted to be the same as that of the other element. This is done in some meters by bringing out a number of taps from the potential winding, having a very few number of turns between taps, so that a fine adjustment can be accomplished. Other meters employ what is known as a "bal- ance loop,'' which is a short-circuited turn whose posi- tion can be so changed that it will introduce more or less reluctance in the path of that portion of the flux from the potential winding which does not pass through the meter disc. Increasing this reluctance will cause more of the flux to pass through the meter disc, while decreasing it will cause less flux to pass through the disc. After adjusting the "balance loop' r the meter should be "re-lagged." The balance between the elements of a polyphase meter can also be altered by changing the air gap be- tween the potential and the current coil poles of one element ; this can be accomplished by loosening the screws and prying the poles further apart or by using a light wooden mallet to drive them closer together. The adjustment obtained by this means is necessarily very rough, but it is sometimes useful (usually when putting a new potential coil in place) to bring the elements within the range of adjustment provided by the manufacturer. 62 THE WATTHOUR METER Interference of Elements. Polyphase meters are subject to one source of error from which two single phase meters used on polyphase circuits are entirely free, that being the "interference of elements," which is due to the inter- ference or reaction of the magnetic fields of one element with the fields of the other, and in some cases it introduces errors amounting to as much as 4% or 5%. For this reason the elements should not be placed too close together. In a well designed meter operat- ing near unity power factor, the error from this source should never amount to more than 0.5%. Polyphase Meters on Six Phase Circuits. Fig. 42. Fig. 42 shows the proper connections of two poly- phase meters when used to measure the power flowing in a six phase system. THE INDUCTION METER 63 Metering High Potential Circuits. When meters are used on high potential circuits, the secondaries of both the potential and the current transformers should be solidly grounded. This is not only a precaution for the safety of those who have to read and test the meter, but it also prevents undue strain between the windings of the meter. Both the potential and the current transformers act as con- Ground Source 17 Fig. 43. densers, and so do the windings of the meter them- selves. The voltage of the system is thus impressed across several condensers in series, the strain across each condenser being inversely proportional to its electro-static capacity. It is possible for the strain thus impressed to reach a value which will puncture the insulation from the winding to the core. Fig. 43 shows the connections for both a single 64 THE WATTHOUR METER and a polyphase meter when used with current and potential transformers, showing the ground connec- tion to be made when used on high potential circuits. When metering the high tension side of a "Y" connected three phase system, the current trans- formers can be relieved of a great part of the high tension strain by connecting them between the power transformers and the neutral or "Y" point. This will also protect the current transformers from lightning and high potential surges, as each current transformer will have a power transformer between itself and the line. Current and potential transformers used with watt-hour meters should never be operated under the condition of overloads, and it is best to have them operate considerably underloaded. Overloading the transformers will cause the meters to run slow. Current transformers are usually rated at so many watts, for instance, 40 watts. The sum of the volt- amperes taken by all the meter coils in series with such a transformer plus the volt-amperes consumed in the leads to the meters should never exceed this amount and should preferably be less. Potential transformers are usually rated at from 10 to 200 watts. The total load in volt-amperes should never exceed the rating and where a high degree of accuracy is required the total load should be con- siderably less than the rating. In, making connections of polyphase watthour meters care should be taken to see that the meter is connected exactly in accordance with the diagram fur- nished by the manufacturer; if this is not done, the meter may be connected so that it will run in the proper direction, but the interference between ele- ments will be high ; this will be the case if both the current and potential connections of one element are reversed. CHAPTER IV. THE COMMUTATING TYPE OF WATTHOUR METER. During the past few years the commutating type of watthour meter has practically been superseded by the induction type for use on alternating current sys- tems, and at the present time its use is principally in connection with direct current work. The commutating meter (as well as other types) is in reality a direct connected motor-generator, the Fig. 44. motor being of the shunt type, having its armature connected in multiple (or parallel) with the source of supply and with its field coils in series with the load to be measured. The revolving aluminum (formerly copper) disc and the retarding magnets comprise the generator. As the disc D, Fig. 44, revolves between the jaws of the retarding magnets M, it cuts the lines of magnetic force, thus producing "Foucault" or "eddy" currents in the disc.. 66 THE WATTHOUR METER Principle of Operation. The torque, or turning effort, of the motor is proportional to the product of the magnetic flux set up by the armature and that set up by the series field coils. The magnetic flux of the armature is propor- tional to the impressed e. m. f., and the magnetic flux of the series field coils is directly proportional to the current flowing through them. The product of the current and the voltage equals the power, therefore the turning effort of the armature is directly propor- tional to the power being expended in the circuit C. The power generated and expended in the disc itself depends directly upon the speed, since the eddy cur- rents generated depend upon the rate at which the magnetic lines are cut, therefore the drag on the disc will be directly proportional to the speed. We there- fore have an instrument in which the turning effort is proportional to the power passing through it, and in which the retardation, neglecting friction, is propor- tional to the speed. Since the speed will increase until the torque just balances the retardation, the revolving element will turn at a speed proportional to the power passing, which is the condition sought. The revolu- tions of the armature are transmitted through a suit- able train of gears to the dials which register in units of electrical work, such as the watthour or the kilo- watthour. Comparison to a Shunt Motor. There is one essential difference between the ordinary shunt motor and the motor of a commutating watthour meter, ana that is the fact that the latter has no iron or steel in its magnetic circuit. If iron were employed in the meter, its torque would no longer be strictly proportional to the current flowing in the series field coils, due to the "saturation" effect of the iron, which would result in a greater reluctance (or magnetic resistance), with an increase in current. Therefore, on light loads, the torque would be corre- spondingly greater than at full load, thereby causing the meter to over-register on light loads, provided, of COMMUTATING TYPE. 67 course, that it was adjusted to register correctly on full load, or vice versa. It is a well known fact that the ordinary shunt motor will increase in speed if the field current is decreased, because the armature will then have to run faster in order to generate the "back" or counter e. m. f., under the conditions of a weaker field. On the other hand, a watthour meter will decrease in speed with a decrease in field current, or vice versa. These two facts are apparently contradictory and may be accounted for as follows : The speed of a shunt motor is proportional to the impressed e. m. f., and inversely to the field strength, and must be such that the back e. m. f. is equal (plus the RI drop) to the impressed e. m. f. Any weakening of the field will therefore cause an increase in speed, since the armature con- ductors have to cut the decreased field at a higher rate in order to generate the same back e. m. f. (For a full explanation of this theory, see any text book on direct current motors.) In the case of the meter, however, the counter e. m. f. is inappreciable, the impressed e. m. f. being practically all absorbed in the resistance of the armature, the auxiliary or "compensating field," and in the external resistance if any is used. So long as the voltage remains unchanged, the armature cur- rent will therefore remain unchanged, irrespective of the changes in the series field strength and the speed. The effect of a decrease in field strength is to decrease the torque, with a consequent decrease in speed until the retarding torque exerted by the permanent mag- nets on the disc is decreased to correspond to the turning effort of the armature. With an increase in field strength the reverse takes place, that is, the re- action between the armature current and the stronger field produces a stronger turning effort which increases the speed until the retarding effect of the permanent magnets increases to a corresponding degree. This condition of the operation of the meter holds true for an ordinary shunt motor until the counter e. m. f. is more than about 50 per cent of the line potential, 68 THE WATTHOUR METER below which point the speed of the shunt motor would increase with the field strength, and above which point the speed would decrease when the field strength was increased. Thus it will be seen that, in reality, there is no discrepancy between the motor of a meter and the ordinary shunt motor. Efficiency. The efficiency of a meter is based upon the actual watts lost in the resistance of the series field coils and the potential circuit (which includes the armature, the compensating field and the external resistance), the losses due to friction, and the losses in the disc due to eddy currents set up by the retarding magnets. During the early development of the watthour meter of the commutator type, the loss in the poten- tial circuit alone, in a 100 voltmeter was about 10 watts, and in the 200 voltmeters was about 20 watts ; such meters are now termed "low efficiency" meters ; the present type or "high efficiency" meter has a loss in the potential circuit of about 4 or 5 watts in the 100 volt meter, and a loss in the series field coils not exceeding i per cent of the total capacity of the meter in the smaller capacity meters and much less than this in the large capacity meters. The reduction in losses has been accomplished by increasing the resistance in the potential circuit and by almost doubling the number of conductors on the armature ; the number of arma- ture conductors being increased to produce a greater torque. General Construction. Fig. 45 illustrates interior views of three repre- sentative types of commutating meters, in which (a) is the Westinghouse, (b) the General Electric, and (c) the Duncan, all of American manufacture. The meter shown at (c) is designed for use on direct current circuits, although it may be said that the Duncan alternating current meter is very similar in construction, and ks operation essentially the same COMMUTATING TYPE. 69 70 THE WATTHOUR METER as the direct current meter. The use of the commu- tating type of meter on alternating current circuits will be dealt with later in this chapter. The Compensating or Shunt Field. The function of the compensating or shunt field is to compensate for friction, especially at light loads. When a meter is operating on a very small percentage of its rated load, the ratio of friction to torque is relatively great, therefore the lighter the load the greater will be the retarding effect of friction. In order to overcome this friction effect, the compensat- ing field is connected in series with the armature so that its flux will work in conjunction with the main or series field. In the General Electric and Westing- house meters, the strength of the compensating field is constant, and the "helping out" or compensating effect is altered by moving it closer to or further from the armature, so that more or less of its flux embraces the armature. In the Duncan meter, the amount of compensation is altered by means of the multi-point switch shown in the illustration. This switch is con- nected to various taps on the compensating winding and the variation is accomplished by cutting in or out a certain number of the coils, thereby altering the flux. When the compensating field was first used, it was permanently fastened to the inside of the series field coils. This method was soon superseded by mounting it on an adjustable rack, so that it could be moved toward or away from the armature and then clamped in the correct position. The compensating field should be so designed -that (in a new meter) it will allow a maximum boosting effect of about 10 per cent on light load, that is, it should have sufficient strength when adjusted for full compensation, to in- crease the speed of the meter by about 10 per cent when the meter is operating on 5 per cent of full load. This allows sufficient margin for adjustment as the friction increases. W r ith compensating fields designed to give a greater boosting effect, the meter-man is apt COMMUTATING TYPE. 71 to take advantage of the quick method of temporarily adjusting the meter and thereby compensate for ex- cessive friction which by all means should be located and removed. In the older types of meters (with especial refer- ence to the Thompson recording type), on account of the low armature resistance, it was necessary to place an external resistance in series with the compensating field, such resistance being mounted in card form on the back of the meter case. This method has been simplified by having the entire resistance of the poten- tial circuit (external to the armature) self-contained in the compensating field in all meters up to and including 250 volts. For the 500 and 6oo-volt types, it is still the practice to furnish a suitable external resistance for the potential circuit. Brushes. It is very important that the brushes be made of a material which will not vary in elasticity, and when once properly adjusted they should maintain their tension permanently. The control of the brush ten- sion is effected either by gravity or by a spring. The actual contact surface of the brush should be made of silver since it has been found from practice that this material gives better service under operating con- ditions. Each brush (i. e., each positive and each negative) is usually divided into two parts, so as to give a more even distribution of pressure at the point of contact, and to make the brush self-aligning. Brush friction has been considerably reduced by using a cylindrical rather than a flat type, The Commutator. During recent developments in the manufacture of meters, the diameter of the commutator has been materially reduced, and at the present time some makes employ a diameter of less than one-tenth of an inch in meters of no and 220 volts capacity. This reduction in the size of the commutator has greatly reduced the friction of that particular member. It is 72 THE WATTHOUR METER general practice to make the commutator bars of pure silver, since this metal suffers least from oxidization, and therefore it presents a smoother surface and more constant contact resistance, two features which are desirable. The commutator is usually built up directly on the shaft, the bars being insulated from it and from each other and are held intact by a metal ferrule on each end of the commutator, the ferrules, of course, being properly insulated from the bars. In some cases the commutator bars are insulated from each other by fibre bars or other solid insulating material, and in other cases simply by an air space. Each of these methods, under certain conditions, are liable to give trouble. When a hard insulating material is used, it is apt to wear down slower than the commutator bars, causing the brushes to "ride," and thereby opening the armature circuit, which will either cause the meter to stop or else cause severe sparking. In case a soft material is used it is apt to gum the commutator and give rise to the same trouble as too hard an insulation. The trouble due to air-insulated bars, is that under extreme conditions, the air space may become filled with dust and small particles of metal, thereby causing adjacent bars to become short-circuited. A meter should be inspected often enough, though, so that under average commercial conditions the commutator with air-space insulation will give good service and will very probably be superior to the commutator with solid insulation. The Armature. There are two general types of armature con- struction at the present, the spherical and the rect- angular. The tendency is to favor the spherical type, since this construction permits the field coils to be so designed as to allow a minimum leakage of magnetic flux, thus securing the highest possible torque for a given watt loss in the fields and armature windings, and thus approximating more nearly the condition of an ideal meter. In both the spherical and the rectangular wound COMMUTATING TYPE. 73 armatures, the winding is of the well-known ''Siemens drum type." The rectangular winding is usually sup- ported by two spiders made of small strips of hard wood and properly secured in position on the shaft. The supporting medium in the spherical wound arma- ture consists of two hemispherical pieces of fibre which are mounted directly on the shaft, the windings themselves being held in position by grooves which are stamped in the fibre shells. This construction is good mechanically and insures a very light weight of moving element. The full load speed of the commu- tating type of meter is usually about 40 r. p. m., which further permits of very light armature construction. Generally speaking the armatures of meters for use on no-volt circuits or thereabouts usually have 8 armature coils of about 1000 turns each, of number .003 copper wire, and those of 200 volts and above have 16 coils of 500 turns each, of the same size wire. This method of subdividing the coils on the higher voltages is followed so that there will not be such a great difference of potential between adjacent coils nor between the commutator bars. There is one com- mutator segment per coil, for instance, in a loo-volt meter there will be 8 commutator segments, and in a 2OO-volt meter there will be 16 segments. The total armature resistance in meters from 100 volts to 600 volts inclusive, of the ordinary house type, is usually between 1000 and 1200 ohms, the proper amount of resistance being placed in series to limit the current on the various voltages. The armature current is practically the same for all voltages from 100 to 600 inclusive, the total resistance of the poten- tial circuit being subdivided approximately as shown in the table below: DO CO O - ^ en fag Sg g g -2 s^2 TYPE OF METER ~| || -| Sg || ||| 2 D Cu "o i) "o o'o.i: i- 3 S 5 amp., 2 wire, 110 v 1,300 1,200 2,500 5 0.0447 5 amp., 2 wire, 220 v 3,800 1,200 5,000 10 0.0447 5 amp., 2 wire, 550 v 10,900 1,200 12,100 25 0.0454 74 THE WATTHOUR METER It is thus seen that the armature current is prac- tically constant. While on the subject of armatures, it should be borne in mind that the temperature coefficient of the armature and the disc should be the same, so that any decrease in torque of the armature, due to a rise in temperature, will be correspondingly offset by a rise in resistance of the disc, which in turn would decrease the effect of the retarding magnets. Other features of construction have been dealt with in Chapter I. COMMUTATING TYPE. 75 The Use of the Commutating Meter on Alternating Current Circuits. As previously explained, the commutating type meter is a simple shunt motor, and this being the case, the question may arise, "Why is it that such a meter can be operated with accuracy on alternating current circuits?" It should first be remembered that, owing to the iron in the magnetic circuit of an or- dinary shunt motor, there would be a great difference in phase relation between the current in the armature and the current in the fields if such a machine was supplied with alternating current, this being due to the much greater inductance of the field winding. The current in the fields would lag almost 90 degrees be- hind the current in the armature, therefore the torque produced would not be sufficient to cause rotation. The meter, being as it is, devoid of iron in its mag- netic circuit, will not suffer from such a phase differ- ence when supplied with alternating current. The commutating type of meter can be made to operate with accuracy on alternating current circuits by making an adjustment which is termed "lagging." If this type of meter is used on alternating current, precisely the same as on direct current that is, with- out any adjustments the current in the armature will lag a few degrees behind the impressed voltage, while the current in the field coils, for a load of unity power factor, will be in phase with the impressed voltage; the lag in the armature current being caused by the inductance of the armature and the compensating field. This, however, does not introduce a serious error at unity power factor. In order that the meter may regis- ter correctly on power factors other than unity, it is necessary that the current in the armature be in phase with the current in the series field when the meter is operating on a load of unity power factor. It is there- fore necessary that an adjustment be made that will bring the two currents in phase, thereby correcting the small phase difference above referred to. Such an adjustment is accomplished by shunting a part of the 76 THE WATTHOUR METER current in the series fields through a non-inductive resistance. By properly adjusting this resistance, the current in the series fields can be made to "lag" until it is in phase with the armature current. The principle or theory of this method of ad- justment may be explained as follows: The series coils have both resistance and inductance, and when shunted by a non-inductive resistance, the line current is divided into two components, one of which flows in the non-inductive resistance and the other in the field coils themselves. (The current in the non-induc- tive resistance is a small percentage of that flowing in the field coils.) The relative values of these two components of the line current are inversely propor- tional to the impedances of the two paths, and the phase angle between them will depend upon the ratio of the resistance to the reactance of the series coils. This is diagramatically shown at (a) in Fig. 47, where OV represents the impedance drop around the series -coils and the non-inductive resistance together ; i being the current in the resitance and i' the current in the series field coils. The voltage drop, Ri, in the non- inductive resistance, will be =OV, and i will be in phase with OV. The drop in the series coils, however, consist of two components, a resistance drop, R', i', which is in phase with i', and a reactive drop xi' at 90 degrees from i', the phase angle between i and i' being represented in the figure by $. The main line current is made up of these two components as shown at (b) in Fig. 47, and it can be seen from this that the angle p, which is the lag of the current i' in the series coils behind the line current I, can be adjusted by merely changing the value of the current i in the non-inductive resistance. After having properly lagged the meter on say 50% power factor, it is necessary to recalibrate it, since the torque exerted by the series coils will be less than before the adjustment was made, as the total line current is no longer flowing in the series coils. This recalibration is made by adjusting the re- COMMUTATING TYPE. 77 tarding magnets, after which the meter will be ac- curate for all power factors above 50%. An unlagged commutator meter will have a tendency to run fast on inductive loads. In any event, where it is neces- sary to lag commutating meters it is advisable to take the subject up with the manufacturer of the meter in question, and obtain their recommendations. In all V xz' Tt'i' Fig. 47. cases involving a great number of meters, it is advis- able to change the entire installation over to the in- duction type on account of its greater simplicity and superior operation on alternating currents. Three-Wire Meters. In the heart of cities, and in buildings where a large amount of current is used, the three-wire system of distribution is almost always to be recommended on account of the great saving in the amount of copper in the distributing wires, the most common system being the 220/1 lo-volt system, the current being fur- nished (in case of direct current) by either a three- 78 THE WATTHOUR METER wire generator, a two-wire generator with a balancer set, or by two generators operating on a three-wire connection. The question often arises as to what extent should the distributing company insist upon having the system balanced. In New York City the O'cx/rce Fig. 48. requirements are very rigid. For instance, all light- ing circuits taking more than five amperes must be equally divided between the two sides of the system, and all motors over 5 h.p. must be connected across the outside wires. On the other hand, some com- Fie. 49. panics pay no attention whatever to the "balance," and depend upon the average conditions to balance the load at the station bus-bars. At the point of distribu- tion, however, the conditions may not be so favorable as at the station, thereby resulting in poor service on COMMUTATING TYPE 7'.' one side or the other of the system. It is therefore recommended that some effort be made to keep the load fairly well divided between each of the two out- side wires and the neutral. To measure the power flowing- in a three-wire system, it is necessary to use two meters connected as shown in Fig. 48, or to use one three-wire meter whose internal connections are shown diagramatically in Fig. 49. The only way in which the three-wire 7=" Fig. 50. meter differs from the ordinary two-wire meter is that in the former the series field coils are divided into two equal sections, which are connected in the oppo- site sides of the system as shown above. There are devices on the market for automatically connecting the potential circuit to the opposite side of the line without reversing the direction of rotation of the meter in case one side of the three supply wires should be disconnected. 80 THE WATTHOUR METER The armature circuit is usually connected be- tween the neutral and one of the outside wires in the three-wire meters, because such practice permits cheaper construction, due to the lower voltage im- pressed. (In Fig. 50, P represents the armature cir- cuit and FF the series field coils.) In either case- that is, with the armature circuit across the outside wires or across one outside wire and the neutral the three-wire meter is subject to error on unbalanced loads ; if connected to neutral it may register either slow or fast, depending upon whether the voltage be- tween C and B (Fig. 50) is less than or greater than one-half the voltage between A and B. If the poten- tial circuit is tapped from A and B, the meter will usually register high on unbalanced voltage, as the lower voltage will usually be on the heavier loaded side. It is very seldom that the unbalancing of the road on a three-wire system is such that it will cause any great degree of inaccuracy, but if extreme ac- curacy is a question of prime importance it is recom- mended that two two-wire meters be used rather than one three-wire meter on a poorly balanced system. High Capacity Meters for Switchboard Service. In order that the distributing company may have an exact comparison between the power actually de- livered from the station bus-bars and the delivered power from which a revenue is realized, it is of the utmost importance that switchboard meters be care- fully selected as to their accuracy and their capacity. The question of "over-metering" as brought out in Chapter I applies with even more force in the case of switchboard meters the detection of unwarrant- able losses depends primarily upon the switchboard meters. It will be found iti almost every case that it is more desirable, for several reasons, to use individual meters on the various generators or feeders than to use "total-output" meters. In the first place, if a single meter is used, its capacity will have to be greatly in excess of the average load, in order that it may take care of the "peak" load, consequently the COMMUTATING TYPE XI large meter will be running far below its maximum efficiency the greater part of the time. Secondly, if it is desired at any time to increase the capacity of the station, the individual method of metering will be found to be much more flexible than will the total output meter method. In the third place, it is much more convenient to test the smaller, individual meters, on account of their lighter connections and the ease Fig. 51. with which testing instruments may be inserted in the circuits. All switchboard meters should be so in- stalled that future testing may be done with the least possible trouble and inconvenience. High capacity meters for direct current switch- board service are, in almost every case, subjected to the influence of powerful stray fields produced by the bus-bars which are usually in close proximity to the meters ; short-circuits and overloads also give rise to 82 THE WATTHOUR METER disturbing influences. In order that switchboard meters be free from such disturbances, special con- struction is necessary. Fig. 51 is an example of a high capacity meter, the one illustrated being for 3000 amperes. The two armatures are "astatically" ar- ranged that is, they are so connected that should the influence of a stray field tend to weaken the torque of one armature, it will correspondingly strengthen the other, and vice versa. It will also be noticed that the retarding magnets are completely shielded by a rec- tangular metal box which is built up of soft steel punchings, which will effectually divert any stray lines of magnetic force which would otherwise affect the accuracy of the meter. Very often it is found necessary to place such a shield on a meter after it has been installed, after which it will also be necessary to recalibrate the meter, because the close proximity of the shield to the retarding magnets will cause a leakage of flux, thereby decreasing the retardation of the magnets. This effect is usually slight, but it is al- ways better to recalibrate the meter. The series field coils of the meter shown in Fig. 51 are of the "bus-bar" type, the magnetic field being produced by a straight copper bar which carries the current from one of the large studs past the armature to the other stud, the effect being that of a single turn. The standard sizes of this type range from 2000 to 10,000 amperes at potentials from 100 to 600 volts inclusive. Switchboard watthour meters ranging in current from 50 to 1500 amperes have the same astatic features as above noted, but instead of having the "bus-bar" field coil, they have several turns of heavy copper. Their damping system should also be encased in a protecting steel box when the meter is in the neigh- borhood of conductors carrying large currents. Another difference between the switchboard type of meter and the ordinary house type is that in the former, all resistance in series with the armature or the compensating field is usually external to the COM MUTATING TYPE 83 meter case, thus minimizing the heating effect from this source. In selecting a switchboard meter, the following points should be borne in mind : The meter should have a high torque, continued high accuracy, light weight of moving element, and should have its armatures and retarding magnets astatically arranged. A recent development has been made in the de- sign of switchboard meters which further protects it against the disturbing influence of stray fields. This is accomplished by making the "motor" a four-pole rather than a two-pole motor. By this arrangement it is possible to place two adjacent (positive and nega- tive) poles much closer together than in the case of the two-pole design. It will therefore be readily seen that a stray field coming from any direction will tend to more equally strengthen one pole and correspond- ingly weaken the other, than in the case of a two-pole meter. This type of meter can of course be used for or- dinary service as well as for switchboard service, pro- vided the conditions warrant the expense of the four- pole meter. Connections of Commutating Type Meters. The connections of the commutating type of watt- liour meters when used on direct current circuits are so simple that it is not deemed necessary to give but a few characteristic connections which are shown in the following figures : Fig. 52. 84 THE WATTHOUR METER Fig. 52 shows the connections of a small capacity (3 to 50 amps.) two-wire "T. R. W." meter, of which there are still quite a number in service. Fig. 53 is the same type meter for two-wire service in capacities of from 75 to 1200 amperes. Fig 54 is the "T. R. W." three-wire, 3^ to 150 ampere meter. From 73 Lo&ct Fig. 54. Fig. 55 shows the connections of the General Elec- tric type "C" watthour meter for 5 to 25 amperes, 500/600 volts, 2 wire (C-6 and C-7) 50 amperes, 100/250 volts, 2 wire (C-6). COM MUTATING TYPE 85 Fig 56 is the connection of a General Electric 75 to 600 amperes, 100/250 volts, 2 wire (C-6 and C-y). 50 to 600 amperes, 250/600 volts, 2 wire (C~7). It will be noticed that in Fig. 56 that only one line wire is carried through the meter on account of the large size of the conductors. Fig, 55. Fig. 56. Fig. 57. Fig. 58. Fig. 57 shows connections of a General Electric type "C": 5 to 50 amperes, 200/240 volts, 3-wire meter. Fig. 58 shows connections of a General Electric type "C": 75 to 300 amperes, 200/240 volts, 3-wire meter. 86 THE WATTHOUR METER La & & Fie. 60. 59 an d 60 show connections of the Westing- house two-wire and three-wire direct-current meters, respectively. CHAPTER V. THE MERCURY FLOTATION METER. WATTHOUR The Sangamo Electric Company of Springfield, 111., manufactures a type of watthour meter which is Fig. 61a. Sangamo Meter. radically different in operation from the induction and commtitating types of meters previously explained. The Sangamo meter is of the mercury flotation type and its principle of operation is based on an old dis- covery made by the scientist, Faraday, when he found 88 THE WATTHOUR METER that a pivoted metalliq disc carrying electric current would tend to rotate when under the influence of a magnetic field. The fundamental discovery of Faraday is very ingeniously utilized in the Sangamo meter. A copper disc is enclosed in a suitable chamber made of moulded insulating material which is divided horizontally into Fig. 61 b. SanRamo Meter. Cover Removed. two sections, the chamber being partially filled with mercury. In the lower part of the mercury chamber there are imbedded two copper terminals which serve to conduct the current to the copper disc through the intervening mercury. The mercury serves the double purpose of conducting the current to the disc and of buoying up the disc so as to make the weight on the lower bearing very slight. The copper terminals are MERCURY FLOTATION 89 arranged diametrically opposite as will be seen from Fig. 62. The exciting magnet which produces the flux which acts upon the disc is imbedded in one section of the moulded "mercury chamber." In the case of the direct current Sangamo meter the main line current, or a proportional part thereof, passes through the copper disc, the magnet being excited from the potential of the circuit upon which the meter is being used. In the case of the alternating current meter, the method of excitation is opposite 90 THE WATTHOUR METER from the direct current meter, that is, the magnet is excited by the line current, and the disc carries a current which is proportional to the potential of the circuit. The reaction of the current in the disc with the magnetic lines of force from the magnet will cause the disc to rotate at a speed which will be proportional to the product of the current and the impressed e.m.f., in other words, it will rotate at a speed which will be proportional to the power being expended in the circuit to which it is connected. In alternating current meters operating on a circuit whose power factor is other than unity, the speed of rotation will then be proportional to the current, the e.m.f., and the power lacior. Under ordinary conditions a variation in tempera- ture between 10 F., below zero, and no F., will not materially affect the operation of the mercurjf meter, but temperatures above or below these maximum and minim-urn values are liable to affect the accuracy. There is sufficient space in the mercury chamber and such a (comparatively) small percentage variation in the volume of the mercury with changes in tempera- ture that the expansion of the mercury will not cause it to leak out, as is sometimes supposed. The direct current Sangamo meter cannot be used on alternating current circuits because of the high self- inductance of the potential winding. Therefore, if alternating current be applied to a direct current meter very little current would pass through the potential coil, and even that would lag by so many degrees that it would produce a very small torque. Fig. 62 shows diagramatically the Sangamo direct current meter. The damping system in this type of meter is essentially the same as previously explained in connection with induction and commutating types of meters. In the above figure the damping magnets are shown at M, and the damping disc at D. The copper terminals which lead the current into the mercury and thence to the disc, A, are shown at EE. The external resistance, R, is in series with the poten- MERCURY FLOTATION 91 tial winding, SC. The light load adjustment is made by moving the slider, K, to the right or left, thereby causing part of the shunt current passing through the potential winding to flow through the armature. This current reacting w r ith the magnetic field from the potential winding produces a no-load torque which is sufficient to compensate for friction. The torque of the Sangamo meter is very low. the torque of a 5 ampere direct current meter being only about 20 gram-millimeters. It may be said, however, that in the case of the mercury flotation meter the pres- sure on the jewel bearing is relatively small, and it is therefore not necessary to have as high a torque as in other types. The reason for this low torque in the San- gamo meter is due to the fact that the armature is equiv- alent to only one turn, and as it is not practicable to carry more than 8 or TO amperes through the armature, the effective armature turns will necessarily be low. Current shunts are used in all direct current Sangamo meters having a capacity of more than 10 amperes, the shunt being external in large capacity meters and internal in the smaller sizes. The sliding connector, S, shown in Fig. 62 is used for adjusting the armature current with respect to the shunts. It is not feasible to build the mercury meter for three-wire direct current service, since it would neces- sitate the construction of two separately insulated mercury chambers and armatures, which would of course be too bulky, complicated and expensive to be warrantable. Two mercury meters have to be used where it is desired to measure the power in a three- wire direct current circuit, with this type of instru- ment. The alternating current meter of this type is shown diagramatically in Fig. 63. As will be noted, the armature circuit is in shunt with the source of supply in this case, as was above mentioned, rather i-han in series as is the case with the direct current tvpe The small potential transformer, N, has its primary, PT, connected across the line and induces '92 THE WATTHOUR METER through its secondary, MS, a current of high amperage -and very low e.m.f. This current flows directly through the armature as is shown. The actual value of this secondary current is between 12 and 20 amperes T F Fig. 63. -at an e.m.f. of approximately 0.05 volt. The trans- former also has an auxiliary secondary winding, AS, the terminals of which are connected through the variable resistance, RR, to the light load adjusting -coil, J, which is wound on the same core with the series MERCURY FLOTATION 93 field coils, FC. By moving the slider, K, along the resistance the compensating effect of the light load adjustment may be varied to allow for friction. The full load adjustment in the Sangamo meter is quite different from the usual practice, in that the damping effect of the retarding magnets is varied by "shunting" more or less of the magnetic lines through a soft iron disc, H (Figs. 62 and 63), which is placed directly above the magnet system, rather than by moving the magnets themselves. This soft iron disc is movable in a vertical direction as shown ; by bring- ing it in close proximity of the magnets, it weakens their effect, thereby causing the meter to run faster. The Ampere-Hour Meter. The Sangamo ampere-hour meter, is now being used quite frequently in connection with the charging and discharging of storage batteries. Storage batteries are rated on their ampere-hour capacity, and it will therefore be seen that an instrument which will indicate the amount of current that has been stored in, or the amount of current remaining in a battery is a valuable accessory to the electric automobile garage, or in fact to any one having the care of a storage battery. The construction of the Sangamo ampere-hour meter is essentially the same as that of the watthour meter, except that the exciting electro-magnet is re- placed with a powerful permanent magnet. The turn- ing effort of the armature will therefore be independent of the potential of the circuit on which the meter is used, being directly proportional to the current passing through it. This type of meter is usually furnished with a single pointer, which will show at a glance the condition of the battery with respect to the charge or discharge. The dial may be furnished with a movable contact, which by means of a relay can be made to open the main circuit through a "shunt-trip" type of circuit-breaker; this movable contact can be set so as- to open the circuit at any predetermined value. 94 THE WATTHOUR METER Connections. A few representative connections of the Sangamo meter are shown in the following figures : Fig. 64. Two-Wire D. C. Meters lie to 250 volts; 100 to 400 amperes, inclusive, box type "Current Shunt." New "Pocket Type" Shunt used in capacities 100 to 200 amperes, inclusive, except for street railway service. Fig. 67. Three-Wire Alternating Me- ters 110-220 volts single- phase; capacities 200 am- peres per side and over. With Current Transformer having two primary wind- ings. if J 1 OOOO 1 1 n^l -^ 1 N0d=) c IN SHUNT OuTg ^ Fig. 65. Fig. 66. Two-Wire Meters 110 and 220 volts, for A. C. Meters 5 to 100 amperes; D. C. meters 5 to 80 amperes, inclusive. Meters may be connected according to Fig. 65 or Fig. 66, but the former is preferable, as this method prevents tamper- ing with the meter connections. MERCURY FLOTATION 95 LINE LINE Fig. 68. Service type, ampere-hour me ter, 10 to 100 amperes inter- nally shunted. Fig. 69. 'Auto" type, ampere-hour me- ter, with contact device and auxiliary circuit for tripping a circuit-breaker. Capacity 10 to 100 amperes. CHAPTER VI. MISCELLANEOUS. The Pre-Payment Watthour Meter. The prepayment device as applied to the gas meter has demonstrated its usefulness after a number of years of service, it being especially valuable in Fig. 73. cities where many of the consumers are transient resi- dents, and also where those served find it a burden to make the usual monthly payments, preferring to pay as occasion demands. The slow 7 introduction of prepayment electric meters has been partly due to the limited demand, because usually the class of people that have been using electricity for light, power, heat- ing and cooking have not been of the kind that would MISCELLANEOUS 97 desire or that would necessitate the installation of pre- payment meters. Now electricity occupies such a broad field that it may be said to be used by all classes of people, therefore the distributing company is often confronted with the problem of "slow pay" customers. Especially is this true with small commercial estab- lishments, such as the poorer classes of restaurants, saloons, tailor shops, etc. The prepayment meter Fig. 74. when installed in such places will oftentimes obviate difficulties which may otherwise arise. The prepayment device can be furnished either as an integral part of the watthour meter, or as a sep- arate device. The construction and operation in either case is essentially the same, except that in the former case the connection to the meter is mechanical and in the latter case it is electrical. Fig. 73 illustrates, the prepayment device attached directly to the watthour meter, in which case a pinion in the registering mech- anism of the meter meshes directly with the debiting 98 THE WATTHOUR METER mechanism of the device. In case the prepayment device is used separately from the watthour meter, the debiting mechanism is controlled by an electro- magnet which is connected directly in the line, con- tact being made through suitable gears and commu- tating device in the meter register. Fig. 74 shows the prepayment device when used as a separate part of a watthour meter. Construction and Operation. The small knob shown protruding from the front of the case is provided with a slot for the reception of coins of the proper denomination. After the coin is placed in the slot the knob is given a half turn to the right, the coin engaging the shaft of the crediting mechanism and the main circuit being simultaneously closed. The coin is carried around with the turn of the knob and released into a chute which conveys it to a coin chamber in the base of the meter. The first coin placed in the slot will cause the indicating hand to move to the figure "i" on the crediting dial, the sec- ond coin will cause the hand to move to the figure "2," and so on, provision usually being made to accom- modate twelve quarter dollar coins at once. Thus $3.00 worth of current may be paid for in advance, and as each quarter's worth is used, the debiting mechan- ism will cause the indicating hand to recede to the next figure. The dial, therefore, only indicates the number of coins to the credit of the consumer and does not take into account the coins for which energy has al- ready been delivered. The total number of coins placed in the device can always be readily translated from the watthour register by multiplying the reading in kilowatt hours by the rate per kilowatt and divid- ing the result by the denomination of the coin. When all the energy which has been paid for has been de- livered, the crediting hand moves back to the zero position and opens an internal switch, which cannot be again closed until another coin is deposited. The switch contacts are made of laminated copper strips which insure good electrical contact. MISCELLANEOUS 99 The force which actuates the debiting device con- sists of a large spiral spring. This spring exerts prac- tically a constant force, since it is so designed that it is always operating under a low percentage of its maximum tension. The gearing mechanism of the spring is "differential" in operation, so that the escape- ment is independent of the knob. This permits the consumer to place more coins in the device before all energy paid for has been delivered, without in any way disturbing the crediting and debiting mechanism. The prepayment device is usually made for rates rang- ing from 5 to 20 cents per kilowatt-hour in steps of one-half cent. Each prepayment device is marked with the rate per kilowatt-hour with which it should be used ; if, however, it is desired to change this rate of charge it is only necessary to change the gear ratio of the "rate device," the construction being such that this is easily accomplished. The coin receptacle is in the back of the meter and so located and protected that the meter cover may be removed (for testing and inspection), without giving access to the coin recep- tacle. It consists of a drawer which can be slipped in or out from the bottom of the meter case, so that it is not necessary for the collector to remove the meter cover when taking out the coin. Lugs (as will be seen from the illustration) are furnished so that the coin receptacle can be locked by means of a suitable padlock. The manufacturing companies have perfected the prepayment device to such a degree that it is as trust- worthy as the gas meter device, and they have de- signed it so that "beating" is practically impossible. When a coin is once placed in the slot, the knob can- not be turned back until a half 'turn has been com- pleted and the coin has dropped into the chute. A coin of smaller dimensions than that for which the slot is designed will not allow the knob to be turned. The credit knob is provided internally with a sharp edge which will shear off any thread or fine wire that may have been attached to the coin with fraudulent intentions. 100 THE WATTHOUR METER The prepayment attachment for the electric meter has the advantage over the gas device in that it may be placed at any point remote from the meter which it controls. The electric meter may be in the attic, the basement or on the back porch, and the prepay- ment device in the kitchen or other convenient place. The diagrams, Fig. 75, show the connections of the prepayment device above described \vhen used in connection with alternating and direct current watt- hour meters as manufactured by the General Electric Company. The Wright Demand Indicator. The need of an instrument which will indicate the maximum demand made upon the current supply of a distributing company has lead to the development of the device shown in Fig. 76, which is known as the "Wright Demand Indicator." As will be noted from Chapter VIII, relative to rates, the cost of serving a customer whose average load for 24 hours, compared with his maximum, is high, will be lower than the cost of serving a customer the ratio of whose average to maximum load is low. In other words, the customer with a good "load factor" is more profitable to the distributing company. For example, suppose that a given customer has a connected load of lamps amount- ing to 10 kw. which are being supplied with current for only two hours in the evening, and suppose that another customer has a connected load of motors amounting to 2 kw. that takes current for ten hours. In either case, the kilowatt-hours per day are the same, but the lighting load comes when the demand upon the station is at the highest point ; whereas, the motor customer is being served when the generating machinery would otherwise be running partially loaded. It is therefore evident that the customer whose demand is practically constant, such as the motor customer, is the most profitable, and is entitled to a better rate. The Wright Demand Indicator can be, and is used to great advantage in determining the maximum MISCELLANEOUS 101 O D [0 x. J n> J D V - 1 1 ^ p_ ? V rF > I ts D > I I ' , II 1 c 102 THE WATTHOUR METER load on transformers. Its systematic use in determin- ing the actual maximnm loads on the transformers of a distributing system will insure the transformers against excessive overloads. As is the case with meters, it is not infrequent that transformers of too large a capacity are used for supplying a given con- nected load ; the maximum demand indicator is valu- able in determining the proper capacity of distributing transformers. The indicator may be mounted on the pole and connected in the circuit by replacing one of the primary fuse-plugs with the plug having loose connections which are attached to the terminals of the indicator. When the indicator is used in this way it should be mounted on a suitable board, which will facilitate handling and also prevent breakage. The indicator can also be applied to motors driving ma- chine tools, etc., to ascertain whether or not they are being operated in excess of their guarantees. Around the upper, or left hand bulb, shown in Fig. 76, and in close thermal contact with it, is a band of resistance wire through which passes the main line current or a proportionate part thereof; shunts being provided with high capacity direct current indicators, and current transformers with the larger sizes for use on alternating current circuits. The current in pass- ing through the resistance band heats the air in the glass bulb, which in turn causes the air to expand, thereby forcing the liquid up into the right-hand tube, the liquid then falling into the central or index tube, which is set in front of the scale. The heat generated in the resistance band is proportional to the square of the current passing through it (watts dissipated in resistance = resistance X the square of the current flowing). The difference in temperature of the air in the two bulbs causes the liquid to flow as it does, and since any external temperature affects the air in the two bulbs similarly, no error will result due to changes in temperature of the surrounding air. The tube is reset by simply tilting it and allowing the liquid to flow out of the index tube back in to the MISCELLANEOUS 103 Fig. 77. Fig. 76, Fig. 78. 104 THE WATTHOUR METER U tube. In resetting the indicator, air bubbles from one arm of the U may be carried over into the other arm, thereby unequalizing the pressure and causing the calibration to be disturbed. To prevent this trouble, the little traps shown in the illustration are located in the bottom of the U, and when the tube is inverted (or partially so), they remain covered with the liquid, clue to the action of the capillaries in the channels of the U tube, and thereby prevent the passage of air from one side of the U to the other side. The Induction Type Watt Demand Indicator. The Wright Demand Indicator which has just been described deals with the current only, and does not take into consideration the power factor of the circuit on which it is used nor fluctuations in line volt- age. It is often necessary to know the maximum watt demand, especially in the case of motor installations. To fulfill the requirement of such a device the General Electric Company has designed the instrument shown in Fig. 78, which is known as the "Polyphase Maxi- mum Watt Demand Indicator." This instrument will indicate and register the maximum watt demand on single, two or three phase systems having a balanced or an unbalanced load, and irrespective of the power factor and voltage fluctuations. This type of indicator is simply a modification of the polyphase watthour meter, the ordinary retarding magnets being replaced by a greater number of very powerful permanent magnets arranged as shown, and with both electrical elements acting on the upper disc. The register as ordinarily furnished on watthour meters is replaced by a circular scale having two con- centric pointers, one of which is connected to the disc shaft through a suitable train of gears ; the other pointer being driven by the first pointer. As the load on the indicator causes the first pointer to deflect, the second pointer is carried to the maximum position reached by the first pointer, in which position it is held by a ratchet. The upper, or "motor" disc is opposed and con- MISCELLANEOUS 1.05 trolled by phosphor-bronze springs which confine the rotation of the disc to a definite number of revolu- tions. The torque acting on the disc is proportional to the power passing through the indicator, therefore by using a control spring of many convolutions, the graduation of the scale can be made uniformly. A curve showing the relation of the percentage of deflection to the time during which it takes the pointer to reach such deflection is shown in Fig. 79. It will be noticed that the curve rises very rapidly until the 90% position is reached, and that the time from 90% Fig. 79. to 1 00%. is relatively great; for this reason the indi- cators are usually rated by defining the time lag as the interval of time taken to record 90% of any change in load. The time lag depends upon the torque of the electrical element and upon the retarding effort of the permanent magnets ; by altering the effect of these two factors, a time lag from one to thirty minutes can be secured. By changing the position of the retarding magnets in an indicator having a definite rating, the time lag may be varied from io r ^ to 15%. The polyphase maximum watt demand indi- cator can be used on single phase circuits by connect- ing the potential coils in multiple and the current coils in series and "dividing the scale deflection by two. CHAPTER VII. MAINTENANCE AND TESTING. The care and maintenance of recording watthoui meters should receive the most careful attention from distributing companies, and only competent men should be placed in charge of the meter department because, as pointed out in Chapter I, negligence in the proper care of the meter system may result in a serious financial loss. It is therefore essential that the dis tributing company be equipped to test and make minoi adjustments of its meters. In order to secure the best results it is necessary that some systematic method of inspecting and testing be adopted. Almost all of the larger companies, realizing the importance of metei accuracy, have separate and well-organized meter de partments which are equipped for testing, repairing and re-calibrating service meters, this department being held responsible for their proper operation. With small companies it is often impractical to have a separate meter department, but it will be found, even by the smallest distributing companies that it is economy in the end to have some systematic method of testing and caring for meters. In small stations, where the size of the system so warrants, it is advisable to employ the entire time of at least one man to see that the meters are kept in proper con dition ; where this is not warrantable, it can usually be arranged to have the same man do all of the metei work, rather than having two or three men, each doing a part of it, because where there is one man, the responsibility is then definitely placed, and further- more, he becomes more efficient and he will usually take more interest and pride in seeing that his meters are always in the best of condition. MAINTENANCE AND TESTING 107 Reading the Meter and Keeping of Records. The interval between the readings of each indi- vidual meter should be as nearly uniform as possible, because if the interval is greater for one month than it is for the next, the customer's bill will as a rule be Kind... Style.. Cycle .Set -..Motors K.W.. ..Tested Inc Are Amp Volt Conit. Folio .Route No Rate Date Reed DEC. Fig. 80. 108 THE WATTHOUR METER correspondingly affected, which will in a great many cases lead to dissatisfaction on the part of the con- sumer, with the resulting annoyance and explanations necessary on the part of the distributing company. The best way in which to obviate such troubles, and to insure a uniformity of meter reading is to begin each month at a fixed date and always have the meter- reader go over the route in the same order. In reading meters it is usual practice to have a special form of "loose-leaf" book which has on each page twelve (or six) facsimile prints of the meter dial, upon which can be marked the corresponding position of the pointers. Such a card is reproduced in Fig. 80. the reverse side of the card being used for any notes \vhich may be necessary. The meter reader should first note the actual reading of the meter and put the figures down in the column set aside for this purpose : he should then copy the exact positions of all of the pointers. By taking both readings thus, one acts as a check on the other, and with a little practice a man becomes efficient and accurate. Some distributing companies use the form shown in Fig. 81, which does not provide for the check afforded by having both the direct reading and the positions of the pointers copied. There is a great difference of opinion as to the most advantageous method of tran- scribing the meter readings to the record book. The objections advanced against the method of copying the positions of the meter pointers is the fact that it takes considerably more time than it does to simply transcribe the numerical value direct ; it also involves more work on the part of the book keeping orga- nization, and there is also liability of error when the book-keeper transcribes the reading from the meter reader's book to the record book. The method of simply transcribing the reading numerically is undoubtedly the most rapid and the most satisfactory if a well experienced and careful man can be employed for this work, but the method of transcribing numerically and also copying the po- MAINTENANCE AND TESTING 109 sitions of the pointers is usually to be preferred on account of the check which it affords. Figure 82 represents a very convenient form of file-card which may be used for the office records and Our No C Mfgrs. No. Location of Meter... DATE READ READING READ BY JUNE JULY FORWABDBD KW KW AUG KW SEPT. KW OCT. KW NOV. KW DEC. KW JAN. KW FEB.. KW MAR. KW APR. KW MAY KW JUNE KW Fig. 81. to which the figures from the reader's book are trans- ferred ; the reverse side of the card is similar and the record may be continued thereon. Under no condi- tions should the record cards be taken from the files. THE WATTHOUR METER SONIQV3H 1N31IS MAINTENANCE AND TESTING METER TEST REPORT. Request Date,. 111 Size. . Name Address Location of Meter. . Company No. Mfgr. No.: . Make Type Testing Constant Reading Constant. I . .Amps. Wires .. Volts % Slow % Fast % Slow .... % Fas t \ Load f Load |Load Full Load 2 Cp. | 4 Cp. j 8 Cp. I 16 Cp. | 32 Cp. j Arcs j Fai Misc. Remarks: Date Meter Tested Meter Tester. N. B.If these Requests fail to follow in Numerical Order promptly advise Accounting Department Fitr. 83a. 112 THE WATTHOUR METER It is not infrequent that check readings have to be made, sometimes at the request of the consumer and sometimes because of apparent discrepancies in the monthly reading. When such readings are taken it is advisable to use a "re-read" card of a form similar to that shown at (b) in Fig. 83. Fig. 83 (a) shows a convenient form of record blank for use in testing meters. Foiio Line RE-READ METER L'M-D&Co 8-08 Date issued . Reading Address Watt Hrs _ K. \V.. .. Meter Number Date Read _ Constant. - By (Tse back of this slip for remarks ) Fig. 83b. During the regular, or periodic, testing of service meters, there will invariably be found a number of slow meters and it is good policy to use a card sim- ilar to Fig. 84, so that the consumer may expect the DEAR SIR : Your meter located at... _.. being our No upon being tested has been found c / c slow. Fearing that any additional billing on your previous consumption would be inaccurate, and possibly unjust to you, we are not rendering any additional bill, but are standing whatever loss has been incurred. We trust all this is satisfactory. Yours truly, Fig. 84. MAINTENANCE AND TESTING 113 next bill to vary from the last one. Such precautions involve very little time and expense, and in many cases they save heated arguments and preserve the good will of the consumer. Date Tested .....Testing Const %l %J Dafe Rec'd Testing Const %l %l POUND LBFT POUND LEFT t/ % % J /2 Full Full Date Tested Testing Const %l % Date Rec'd Testing Const __ , %l %l FOUND LEFT POUND LEFT V 4 % _^_ Full Vz Full Date Tested Testing Const - %l - %S Date Rec'd Testing Const %l %\ FOUND LEFT POUND LEFT '/4 y 4 JL Full '/* Full Fig. 85a. ELECTRIC 'METER RECORD. OUR NO. MANUFACTI-RT .It's No.j MAKE TYP E DATE SET READING NAME ADDRESS DATE OUT READING ' Fig. 85b Too much stress cannot be laid upon the syste- matic keeping of the data regarding the individual performance of the meters, and for this purpose it is advisable to keep record cards similar to that shown in Fig. 85, (a) being the front side and (b) being 114 THE WATTHOUR METER the reverse side. Such data will give an accurate and ready insight into the continued performance of the individual meters, and will act as a guide in the future selection of the best meter for the service of the existing conditions. Installation of Meters. With few exceptions, new meters will be found to be well within the limit of good accuracy, inas- much as they are carefully tested and adjusted at the factory before they are shipped. Before a meter is installed, however, its accuracy should be checked as a matter of record and in order to make any minor adjustments that may be found to be necessary. (In transporting a meter from place to place it is not at all unlikely that the finer adjustments may be af- fected.) When installing a meter care should be taken, as far as possible, to select an easily accessible place which is free from vibration, jar and moisture, and a place where it will be protected from the weather; it should be installed in such a position that it can be easily read a point which is too often over- looked. The meter should not be roughly handled during installation or in carrying it from place to place, as it must be remembered that it is a delicate device and should be handled accordingly. W^hen putting a meter into service care should be taken to see that the moving element does not rest on the jewel bearing until it is installed and ready for operation. The different makes of meters have different methods of accomplishing this result, but in all of them provision is made for protecting the jewel bearing during transportation. Before a service meter is put into operation it is necessary to see that it is level, since friction is liable to result if this precaution is not taken. (Some manufacturers furnish a small pocket spirit level for this purpose). Meters should never be located beneath water pipes nor near steam pipes ; they should be placed MAINTENANCE AND TESTING 115 within 8 feet of the floor line so that the periodic testing may be accomplished with the greatest ease. Meters should not be placed closer together than 15 inches between centers; if placed closer than this they may "interfere" with each other through the effects of stray fields. Meters should not be installed close to conductors carrying heavy currents nor in the vicin- ity of iron girders or posts. The subject of "over-metering" has already been mentioned in several places; as a general rule it may here be stated that for residence lighting the meter should have a capacity of approximately 50 per cent of the connected load ; for small store lighting, win- dow lighting, out-door multiple arc lamps, etc., the meter should have a capacity of about 90 per cent to 100 per cent of the connected load, while for medium and large sized stores this percentage will be approximately 75 to 80 per cent. For metering a motor load, the meter should usually have a capacity of 100 per cent, except where a number of motors are installed, some of which may be running idle or lightly loaded most of the time, in which case a smaller meter could be used. Occasionally it is necessary to install a meter having a greater capacity than the connected motor load, as for instance in the case of hoists and high speed elevators. A very convenient and reliable method of level- ing meters without the use of a spirit-level consists of placing a coin, such as a quarter of a dollar, on the front of the disc and as near to the edge as possible ; if the meter is out of "plumb," the disc will move so as to bring the coin toward the side which is the lowest. The meter can then be leveled so that the disc will remain stationary with the coin resting on the front edge ; the coin should then be placed on the edge of the disc in a position ninety degrees from its former position, and the meter then leveled from front to back, without changing the pre- vious adjustment. When the disc is perfectly level, 116 THE WATTHOUR METER the coin can be placed at any position around the edge, and the disc will remain stationary. Precaution should be taken to see that the meter is connected properly into the circuit, so that it will rotate in the right direction, especially is this true with polyphase meters, and with single, phase meters when used to measure polyphase power. It is not infrequent that two single phase meters are used to measure the power being supplied to polyphase in- duction motors ; the power factor of an induction motor when running lightly loaded is often below /* I fie Meter Properly Leveled fastened to Wall Is the Wall f Stone Wood or Partition \ Brick Cement f Dampnett ribration Does Location of Meter J Chemical Fumes Damage \Dust External Magnetic Fields W IKING: Old or New Are House and Service Wires in Proper Meter Terminals _ Permit Illegal Use of Current (Evidence of S. C. on Meter Cover.) Start* on folarity Creeping Sate Sev if in See. Meter Left Inspector's Report 50 per cent, in which case one of the two single phase meters will run backward when properly con- nected ; care should therefore be taken to see that the meters are so connected that both of them will run forward, when the motor is operating at or near full load. The card shown in Fig. 86, illustrates the form of "Inspector's Report" as used by the Pacific Gas & Electric Co., of San Francisco, Cal., the practice of having a report made out like this for all new installations is to be highly recommended. MAINTENANCE AND TESTING 117 Testing and Adjusting. To insure the continued accuracy of any watt hour meter it is necessary that it be tested and ad- justed from time to time. In the smaller meters it is not necessary to make these tests oftener than about once a year, except in cases of complaint, and when meters are operating under adverse condi- tions ; in the larger sizes, where a small variation in accuracy represents a considerable amount of money, the tests should of course be made oftener, especially where a meter of large capacity is called upon to register a small percentage of its rated load for a great part of the time. As a rule, the conditions under which the meter operates will dictate in a large meas- ure the frequency of the tests. The experience of the large number of distribut- ing companies that have adopted some method of systematically testing and adjusting their meters has proven that the increased revenue resulting from more accurate meters has much more than offset the additional expense to which they have been put, be- sides reducing trouble and complaints due to occa- sional fast meters. In testing meters which are in service it is much better to test and make any necessary adjustments at the point of installation, rather than to bring the meter into the testing room, since with the proper instruments the same accuracy can be obtained and a great saving in time can be effected. Such practice also avoids those injuries to the meter which are liable to occur in transportation to and from the customer's premises. Of course, where it is necessary to make repairs, it is best to bring the meter to the shop. That particular part of the meter which usually gives the most trouble and which requires the most fre- quent renewal is the jewel bearing. Jewels and pivots can be renewed at the point of installation without disturbing the meter. Friction in the jewel bearing will result in the meter running slow, and a new jewel should be substituted whenever a defective one 118 THE WATTHOUR METER is located. When installing a new jewel, a new pivot or ball should also be installed, as the old one is more than apt to have minute particles of the defective jewel imbedded in it, which will constitute an effective cutting tool and will soon ruin a new jewel. In the case of meters employing the ball and jewel bearing it is best to handle the ball with a pair of tweezers rather than with the fingers, as the moisture from the hands may cause the ball to rust. There are several accurate methods used for test- ing watt hour meters, and the choice of the method or methods to be used is usually determined by the rela- tive convenience of that method: the most common are (i) the voltmeter and ammeter method, (2) the indicating wattmeter method and (3) the portable ro- tating standard method. Testing with Indicating Instruments. The voltmeter and ammeter method can only be used with direct current watt hour meters, or with alternating current watt hour meters when the power factor is unity or its exact value known ; the connec- tions for this method are shown in Fig. 87, where V Fig. 87. is the voltmeter, A is the ammeter and W the single phase watthour meter being tested. If E is the voltage impressed upon the circuit, and I the current in amperes, then the power (unity power factor) is P=E x I. The indicating wattmeter method of testing watt MAINTENANCE AND TESTING 119 hour meters is applicable to alternating currents re- gardless of what the power factor may be, so with this method the power, P, is read direct. Fig. 88 shows the connections for this method of testing. Each revolution of the meter disc represents a certain number of watt hours of electrical energy pass- ing through the meter, which is given in some types of meters directly in the form of the meter "con- stant," and in such meters, if R is the number of revo- lutions of the disc in t seconds (as measured with a stop watch), and with constant power passing through Fig. the meter, then the watt hours would be =R x K y where K is the meter "constant." But the power P=watt seconds per second=watt hours x 3,600 di- vided by the time, t, therefore we have p = R K 3,600 The constant, K, for the General Electric meters will be found marked on the meter disc, and are also re- produced in the accompanying tables. 120 THE WATTHOUR METER DIRECT CURRENT METERS. 100-120 Volt Type CH> 1 200-240 Volt. 2 and 3 wire. Type C-6 s 500-600 Volt Type C-7 a E O, IH . be'z: o ec to a: C Qu |E IH ll ll E Is l~* . O 5J "S w S3 .2 ' I! 5 2 500 none 12 .4 250 none 24 1 100 none 60 10 .4 250 24 .75 133.33 " 45 2 50 120 15 .6 166.66 36 1.25 80 75 3 33.33 180 25 1.0 100 60 2.00 50 120 5 20 300 50 2.0 50 120 4.00 25 240 10 10 " 600 75 3.0 33.33 180 6.00 16.66 " 360 15 66.66 10 900 100 4.0 25 240 7.50 13.33 ' 450 20 50 10 1200 150 6,0 16.66 360 12.50 80 10 750 30 33.33 10 1800 300 12.5 80 10 750 25.00 40 10 1500 60 16.66 10 3600 600 25.0 40 10 1500 *50.00 *20 10 300 125 80 100 7500 *Applies to 600 amperes, two wire meters only: 600 ampere, three wire meters are not manufactured. General Electric, Type "I," Standard 60 Cycle, Single Phase Watthour Meters. 100-130 Volt 2 Wire 200-260 Volt 2 and 3 W T ire 500-600 Volt Amps Meter "K" Watts per r.p.m. Meter Watts "K" per r.p.m. Meter "K" Watts per r.p.m. 3 2 12 . 4 24 1 60 5 .3 18 6 36 1.5 75 10 .6 36 1. 25 75 3 180 15 1 60 2 120 5 300 25 1 .5 90 3 180 7.5 450 50 3 180 6 360 15 900 75 5 300 10 600 25 1500 100 6 360 12. 5 750 30 1800 150 10 600 20 1200 50 3000 200 12 .5 750 25 1500 60 3600 300 20 1200 40 2400 100 6000 Polyphase, 60 Type "D-3." 3 .4 24 75 45 2 ' 120 5 .6 36 1. 25 75 3 180 10 1. ,25 75 2. 5 150 6 360 15 2 120 4 240 10 600 25 3 180 6 360 15 900 50 6 360 12. 5 750 30 1800 75 7 .5 450 15 900 40 2400 100 12, ,5 750 25 1500 60 3600 150 15 900 30 1800 75 4500 For General Electric meters used with current and potential transformers, but calibrated without them, the constant to be used is that marked on the meter disc, divided by the product of the ratios of the potential and current transformers. The worm reduc- tion in all General Electric meters is 100 and will be MAINTENANCE AND TESTING 121 found stamped on the back of the register. The regis- ter ratio multiplied by ioo=number of revolutions of disc for one revolution of the right hand pointer. In all cases, the meter, K, is the actual number of watt hours per revolution of the disc. *Rating in (volt-amperes) Meter Constant, K= Full load r.p.m. of disc x 60 -p, . Watt hours of right hand dial Register Ratio= - - Worm reduction x K (*For polyphase meters, the rating should be multiplied by 2. In case a meter has a double rating, such as 110/220 volts, the latter voltage should be applied in the formula. The approximate full load speed of all G. E. type "I" and "D-3," 60 cycle meters is 30 r.p.m.) Westinghouse Meter Constants. For different makes of meters, the testing for- mula given takes a different form since the constant K is made to embrace different factors. For the Westinghouse meter, the formula be- comes P _RXK t where K represents the watt-seconds for one revolution of the meter disc. The values of the constant, K, for the Westing- house types B and C and for the direct current meters are as follows : 2-wire D. C. and self-contained single phase, K = volts x amps x 2.4; 2-wire single phase used with current transformers only (but checked without), K = volts x 5 x 2\4; 2-wire, single phase used with current and potential transformers (but checked without), K = 5x100x2.4; 2-wire, single phase, used with transformers of either or both forms (and checked with), K volts x amps x 2.4; 3-wire, single phase, self-contained, K = volts x amps, x 4.8. 3-wire, single phase, used with current transformers (but checked wtihout), Revolts (as marked on meter) x 12; Type "C" polyphase, self-contained, K = volts x amps x 4.8 ; Type "C" polyphase, used with current transformers only (but checked without), K = 5 x volts x4.8; 122 THE WATTHOUR METER Type "C" polyphase, used with current and potential transformers (but checked without), K = 2400; Type "C" polyphase used with transformers of either or both forms (and checked with), K^ volts x amps x 4.8. In all cases, the volt and ampere values referred to are those as marked on the name plate of the meter. The full load speed of the types B and C is 25 r.p.m. For the Westinghouse type A meter, the full load speed is 50 r.p.m., and the constant, K, for this type is exactly one-half the value of a similarly rated type C meter. Fort Wayne Type K Meter. The calibrating equation of the Fort Wayne type K meter is as follows : p _RXK X 100 t where t is the time in seconds during which the meter makes R revolutions, and where K is the constant, which will be found in the following tables : Fort Wayne, Type "K," Single Phase, 60 Cycle Watthour Meters whose Serial Number is 344,999 or less. Values of the Constant, K. a o 2 > 2 . O >. 2 > E ^ > o 'io '^Q '58 '58 < eg 5? rg^H fi? E 'io o 'io Q "S 'i8 < CN-^H CNf^; &?.o4 3/.9J J/JB 3tZ5 30.-)S30.t /y/s /f.9f /.xt / /.xx XfT* Jtf-ftt Xf./3 X7.F4- X7.X K.o /.9X /90 /.et /.ft 33.4* SJ.tf Zt ZS.3SZS.6t Z3.& X3.Z4>Z*& XX.73 ZX-& ZXAt X*. ?/ ft Z./43 /-ft JSo X..O *0 /.ff /.f6 /.jy /.S3 /.fx / f< 100. In calibrating watthour meters it will be found to be more convenient to use the term "percentage of accuracy" or "correction factor'' rather than the term "percentage error," since the former method involves less work in making the computations. For example : p' Percentage of accuracy = : X 100, p Correction factor == 7,7 If the percentage of accuracy is less than 100 the meter is slow ; if it is greater the meter is fast. If the cor- rection factor is less than i the meter is fast, and if it is more than I the meter is slow. If the watthour meter under test is found to be inaccurate at or near full load, the retarding magnets should be adjusted until the meter registers correctly. If it is slow, the magnets should be moved in toward the center of the disc, which operation will increase the speed, while if the meter is fast the retarding magnets should be moved out toward the periphery of the disc. If the meter under test is found to be slow on light loads, undue friction should be looked for and eliminated, and the light load adjustment (as pre- viously explained for various kinds of meters) reset. If the meter is fast on very light loads, or if it "creeps," the light load adjusting device is very prob- ably exerting too much torque, and its effect should be decreased until the meter is within 2 per cent ac- curacy on a 5 per cent load. 140 THE WATTHOUR METER Shop Methods of Testing. A very convenient and flexible laboratory test- ing board is shown diagrammatically in Fig. 96, which can be used for testing single phase watthour meters of voltages from 100 to 500 inclusive, and of any am- pere capacity, the load being regulated by switching more or less of the lamps in circuit by means of the .single pole switches, L. The indicating wattmeter may be of 5 or 10 ampere capacity, and can be conveni- ently mounted in a horizontal position on a swinging bracket; the current transformer being of a 5 or 10 to i ratio, or if desired, several current transformers of Fig. 96. different ratios may be used. For testing no-volt watt- hour meters of capacities not greater than the capacity of the indicating wattmeter, the d.p.d.t. switch S" is thrown in the dowhward position, thus putting poten- tial on the loo-volt tap of the indicating wattmeter and on the potential winding of the watt hour meter through the variable resistance ; at the same time the t.p.d.t. switch S' is thrown down, thus connecting the indicating wattmeter directly into the circuit. By throwing S' up, the current transformer is connected into circuit when it is desired to test higher capacity meters. The correct load, as near as possible, is ob- MAINTENANCE AXD TESTING 141 tained by closing the switches L, and a finer adjust- ment is accomplished by means of the variable resist- ance shown in the diagram. This variable resistance is most conveniently made up by wrapping a bare resistance wire on a suitable mandrel and having a sliding contact which will not interrupt the circuit in passing from one turn of the wire to the next. By means of this variable resistance, the tester can hold the load on the wattmeter constant while the test is being made. For testing 2oo-volt meters, the switch S" is thrown in the upward position ; and S' is thrown down or up according to whether the meter in test is below or above the capacity of the indicating wattmeter. For testing 5oo-volt meters the switch S is closed,, which puts the potential winding of the watt hour meter directly across the 5oo-volt tap of the compen- sator, at the same time putting the potential coil of the indicating wattmeter across the 5 = the magnetic flux in the transformer core, OV = the voltage of primary winding, OV ] = the voltage of the secondary winding, IN = the magnetizing current, NM= the energy component which supplies the losses of the transformer and the load. One source of error is due to the fact that OI is not exactly 180 degrees displaced from OI 1 , and there- fore the current which flows through the meter (from the secondary side), will not have the proper phase relation with respect to the current in the potential coils of the meter. As already stated, however, for all practical purposes the error thus introduced is neg- ligible except in the case of low power factors, and in transformers of poor design. It can be seen by refer- ring to the above diagram that if the secondary circuit has the proper amount of inductance to cause the secondary current OI 1 to lag by the same angle that the exciting current, IM, lags, that the secondary cur- rent will be exactly 180 degrees out of phase with the primary current, which would result in there being no error from this cause. The second error referred to, which is caused by the varying ratio of the transformer, is due to the exciting current, IM, not being effective in inducing current in the secondary winding; the secondary cur- rent being induced by the component, OM, of the primary current. If IM varied directly as the primary current, this error could be corrected by adjusting the ratio between the primary and secondary turns; such is not the case, however, and an error is introduced. MAINTENANCE AND TESTING 147 Fig. 100 is a curve showing the accuracy of a well- designed current transformer. In calibrating watthour meters for use in connection with current transformers, the meter should be calibrated to register correctly on the flat part of the curve. There will then be a slight error at either end of the curve, that is, there will be an error on very light loads on or overloads, but if the meter is carefully calibrated in accordance with the Fi e . 100. Fig. 101. curve of the transformer it will be accurate over the greater part of the range, and the error at either extreme will be small. Fig. 100 shows a typical calibration curve of a good current transformer, and Fig. 101 shows the cali- bration curve of a standard induction watthour meter. These two curves are combined as shown in Fig. 102, the resultant curve, B, being the resultant calibration curve of the meter when used in connection with the transformer. It will be seen that if the meter is ad- 148 THE WATTHOUR METER justed so that it will be a little slow on full load (about 0.5 per cent), and if the light load adjustment is set so that the meter will be slightly fast (about 0.5 per cent) on 10 per cent load, the resultant curve will be more nearly correct, and when a high degree of accuracy is required, this is recommended, the amount of such adjustment being determined by referring to the calibration curve of the transformer with which the meter is to be used. It should be remembered that a current trans- former must always have a load on its secondary side ; if the meter or instrument with which it is being used should be disconnected while current is still on the Fie. 102. primary, the transformer should either be disconnected from the line, or else have its secondary short-circuited. If the secondary is left open-circuited there will be no counter magneto-motive force from the secondary, consequently the magnetic flux will increase to such a degree that it will cause the iron core to become overheated to an extent that may injure the trans- former. The load carried by potential transformers is constant, and if the load is light, the error in the transformer ratio will be very small. The secondary e.m.f. of the potential transformer leads the primary e.m.f. by a small angle, 6, Fig. 103 ; the angular dis- placement referred to in connection with current transformers is also leading; it therefore follows that the angular displacement in a potential transformer MAINTENANCE AND TESTING 149 compensates, in a large degree, for the angular dis- placement in the current transformer ; if this angular displacement is the same for both the current and potential transformers the error from this source will be entirely eliminated from the meter with which they are used. The angular displacement referred to depends upon the magnitude and character of the load imposed upon the transformer, as well as upon its design. Fte. 103. Fig. 103 shows a vector diagram of the regulation of a potential transformer, in which OE=primary e.m.f., . OE'=secondary e.m.f., OI=primary current, OI'=secondary current, RI=total resistance drop, XI=total reactance drop, $=angular displacement between primary and secondary e.m.f.'s. In using 5-ampere no-volt meters with current and potential transformers and leaving the regular register on the meter it is necessary to use a multiply- ing constant, which, multiplied by the register reading, gives the kilowatt hours consumed. This multiplying 150 THE WATTHOUR METER constant is obtained by multiplying the ratio of the current transformer by the ratio of the potential trans- former. In order to obtain a multiplying constant of 10, loo or 1,000 it is often necessary to use a special regis- ter and to change the disc constant of the meter. The new disc constant is obtained from the following for- mula: 100 x register ratio x transformer ratio 10,000 x C in which C is the multiplier of 10, 100 or 1,000. Applying the above formula to an example we will take the case of a 5-ampere meter having a disc constant of K=.3, and used with a 20 : i ratio poten- tial transformer and a 24 : i ratio current transformer, from which the "transformer ratio" in the formula will be (20x24)^480:1. Suppose a register is chosen having a ratio of 662-3, and a multiplier (C) of 100 is used. Substituting these values in the above for- mula we derive a value of K=.3i2, which should be used instead of K=.3, which would of course result in a slightly different operating speed of the disc. Meter Troubles Some, of the most common troubles encountered in connection with watthour meters may be sum- marized as follows : Excessive Vibration If a meter is placed in such a position as to be subject to excessive vibration, creeping often results ; vibration is also severe on the lower bearing. This trouble can in many cases be remedied by placing rubber washers between the meter and the wall upon which it is supported ; in severe cases a spring suspended board is recom- mended. Wherever possible, meters should be in- stalled in places that are entirely free from vibration or jarring effects. It is not unusual to find meters installed near doors which are often closed and opened, and especially is this bad practice where the \vall or partition is of light construction. MAINTENANCE AND TESTING 151 Humming and Rattling of Induction Meters This trouble is usually due to loose laminations in the magnetic circuit, and can be remedied by tightening them up. Rattling may also be due to a vibiation of the disc and shaft in very loosely aligned meters, which trouble can usually ^e removed by carefully examining, locating and tightening the exact parts that may be loose. Excessive humming is sometimes caused by the potential winding being loose on the potential pole, which trouble can be easily remedied by driving small, flat wooden wedges between the insulating sleeve (upon which the coil is mounted) and the core. Humming is not an inherent phenomenon of the induction meter, and trouble from this source can always be traced to some simple mechanical defect. It sometimes happens that the wall upon which the meter is mounted acts as a ''sounding Voard," thus magnifying the humming of the meter. This can be corrected by the use of rubber wr.shers as above men- tioned for excessive vibration. It is best, however, to remove the meter to a place that will not be subject to such trouble. Weakening of the Retarding Magnets Magnets, after having been in service for som.e time, may be- come weak, due to the "aging" of the steel; this should not occur, however, if they have been thor- oughly and properly treated before leaving the factory. Weakening due to aging is a very serious defect and shows the lack of proper methods or care in their manufacture. Such trouble should be guarded against in the selection of meters. The retarding magnets may also be weakened by the effects of powerful stray fields, or by heavy short circuits on the "load" side of the meter. The retarding magnets should never be moved closer than one-quarter of an inch to the periphery of the disc, because if they are there will be a leakage of magnetic flux around the disc, from pole to pole of the magnets, thereby decreasing the number of 152 . THE WATTHOUR METER lines of force that actually cut the disc. This of course will have the same effect as the weakening of the magnets, and will cause the meter to run fast. Magnets that will not produce the necessary retarding effect when moved within a quarter of an inch of the edge of the disc are too weak and they should be discarded. Sometimes it is found necessary to place iron shields around the damping system of meters which are not already provided with such a protection against stray fields, and when this is done the meter should be re-calibrated, since the proximity of the iron shield may allow a leakage of flux which would result in the same trouble as placing the magnets too near the periphery of the disc. Bent Shafts and Buckled Discs These troubles may be due to one of three causes; by abuse, by the effects of short circuits of a severe nature, or to faulty manufacture ; the only remedy is to install new parts. Creeping Creeping may be due to "over-com- pensation" of the light load adjustment, vibration, high voltage, or a combination of any or all of these effects, or in commutating meters by the external re- sistance being short-circuited. Some types of induc- tion meters have two small holes punched in their discs, the holes being diametrically opposite. When the part of the disc which has the hole in it comes under the influence of the electrical element, the torque is thereby sufficiently decreased to allow the disc to stand in that position when there is no current flowing in the series coils. This method very effec- tually prevents "creeping," but it does not affect the accuracy of the meter. Creeping in the commutating type of meter can be very effectively eliminated by clamping over the edge of the disc a small piece of U-shaped soft iron wire; when the piece of wire comes under the in- fluence of the retarding magnet the attraction of the magnet tends to hold the disc in that particular posi- MAINTENANCE AND TESTING 153 tion, therefore preventing creeping on no load. The size of the clip can be so selected that it will pre- vent creeping, but which will not prevent the meter from starting on light loads nor affect the light load accuracy. A modification of the last named method con- sists in attaching a piece of iron wire to the shaft of the meter in such a manner that its free end extends out radially and comes under the influence of the retarding magnets ; the effect of the wire can be varied by bending it so that its free end comes closer to or further from the magnets. Defective Jewels The simplest and probably the easiest way of detecting roughness or defectiveness in the jewel bearing is to take the point of a sharp needle and gently "feel" the entire surface of the jewel. A fracture or any roughness can thus be detected. In this connection it might be stated that one of the best materials for cleaning the jewels and pivots is the pith from a cornstalk. After the jewel and pivot have been thoroughly cleaned, the pivot should be wiped with a clean rag which has been moistened with a high grade of watch oil, but under no conditions should it be flooded with oil. Changing Position of Commutator on Shaft If the commutator is shifted from its correct position on the shaft it will cause the meter to run slow ; if it is shifted 90 degrees the meter will stop, and if shifted more than 90 degrees the meter will run backward. Backward Rotation of Commutating Meters on Light Load It may sometimes be found that the commutator meter will run in a reverse direction on light loads, while on heavier loads it will run in the proper direction. Such a trouble will be found due to a reversed connection of the compensating field, so that instead of helping the main field out, its action is differential. Care should therefore be taken to see that the compensating field is properly connected. Open-Circuited Armatures The current in enter- ing the commutator divides at each brush and flows 154 THE WATTHOUR METER through the armature in two multiple paths of equal resistance; if one of these paths is opened, it will therefore be seen that the equivalent resistance of the armature will be doubled, which will cause the meter to run at about half speed. The same result may be accomplished by using an ordinary 16 candle- power lamp and moving the connection from one bar to the other; the lamp will not light until after the defective coil has been passed. The defective coil can very easily be located with a voltmeter. Apply the normal voltage to the two brushes and with one of the voltmeter leads per- manently attached to one brush, move the other lead over the commutator from bar to bar, during which operation no deflection will be indicated on the volt- meter until the defective coil is reached, unless the movable voltmeter terminal happens to be passed over that half of the armature in which there 'are no open-circuits. Friction in Upper Bearing It sometimes happens that the upper bearing is pressed down too tightly against the upper end of the shaft, thereby causing excessive friction which will result in the meter run- ning slow on light loads ; this trouble is easily reme- died by loosening the binding screw and raising the bearing slightly. Friction in the Registering Mechanism If the registering mechanism is allowed to accumulate dirt and grease it will develop undue friction ; care should therefore be taken to see that the registers are kept in good, clean condition. Dirt on Meter Disc Small pieces of trash or dirt on the meter disc will, if they come in contact with the retarding magnets, or the stationary element, act as a brake on the meter. CHAPTER VIII. RATES. The fixing of a scale of rates for the sale of elec- trical energy which will be fair to both the consumer and the distributing company is a difficult problem, which will here be briefly outlined. We will confine ourselves to showing why it is that electrical energy cannot be sold to all classes of consumers at the same rate, and further to reproduce schedules of rates as adopted by some of the leading distributing companies throughout the United States. Electrical energy cannot be stored in large quan- tities except at a great expense, but must ordinarily be "manufactured" as the demand necessitates. For this reason, it is necessary for the distributing com- pany to provide generating and distributing equip- ment to handle the maximum demand or "peak load," and since the peak load usually lasts for only a few hours, the system is being operated for the greater part of the day at a production much below its full capacity. The operating expenses, however (except fuel, etc.), remain practically the same, as do the fixed charges, the maintenance, depreciation and interest on the investment. The charges which are proportional to the quantity of energy being generated, such as water, fuel, etc., constitute the smaller portion of the total cost, therefore it is evident that the distributing company can sell electrical energy to consumers using it for a good many hours per day cheaper than it can to consumers using it for only a few hours per day, since the revenue from the "long hours" customer will be greater even at a lower rate, while the manufac- turing expense will not be much greater. To illustrate the above statements take as an example two consumers, each taking the same amount of power, but one of which takes this power for ten hours per day while the other takes it for two hours 156 THE WATTHOUR METER per day. The equipment necessary to supply each cus- tomer is practically the same, as are also the fixed charges. A profit of one cent per kilowatt-hour above operating expenses from the customer taking power for ten hours per day would be more profitable to the distributing company than would a profit of two or three cents per kilowatt-hour from the customer tak- ing power for only two hours. In the first case the .gross profit above operating expenses would be ten cents per kilowatt per day, while in the second case it would be four or six cents. Suppose that the fixed charges amounted to 4 cents per kilowatt per day, the two hour customer would yield a profit of only two cents per kilowatt demand per day, while the ten hour customer, at a rate of 2 cents lower, would yield a profit of 6 cents per kilowatt demand per day, or three times as much. In the case of very small con- sumers, the cost of bookkeeping, meter reading, test- ing, etc., is disproportionately high ; the losses in the -distributing system are also out of proportion, which further increases the cost of supplying energy to the small consumer. The distributing company can afford to sell energy during the "off-peak" period cheaper than it can dur- ing the hours of the peak load, since during the period of maximum demand, the equipment is usually taxed to its utmost. An increase in the peak load means an increase in the equipment or else a greater strain and depreciation on the present installation, while an in- crease in the "off-peak" load can be readily handled with the resulting increase in revenue. Generating and distributing equipment has to be provided of sufficient capacity to take care of the peak load. During off-peak hours, a large part of this equipment is idle. The charges (maintenance, depreciation and interest on the investment) due to this excess equipment provided to handle the peak load are properly chargeable to the cost of manufacture during peak load hours, which makes the cost of producing a kilowatt-hour during this time high. RATES 157 Since the cost of manufacture is higher during the peak load, it is only fair and just that the consumers demanding current at this time should pay a corre- spondingly higher rate. Another point which should be borne in mind when determining the rates made to different custom- ers is the nature of the load with regard to the power factor. The capacity of the generating and distribut- ing equipment is limited by the amount of current to be handled, from which it is evident that the cost of supplying energy to a load of low power factor will be higher than the cost of supplying a similar load (in kilowatts) of a higher power factor. Especially is this true during the period of maximum demand. REPRESENTATIVE SCHEDULES OF RATES. The Commonwealth Edison Co., of Chicago, 111. Schedule A. Regular Lighting Rate. The following is the regular rate for electricity for light- ing purposes, or upon an interior distributing circuit carrying electricity for lighting and also for heating or power through the same meter, as measured by a meter or meters owned and installed by the company: Thirteen cents (13c) per kilowatt hour for all electricity consumed in each month up to and including an amount that would be equal to thirty hours' use of the consumer's maximum demand in such month, and seven cents (7c) per kilowatt hour for all electricity consumed in such month in excess of that amount. Maximum recording meters will be installed by the com- pany for the purpose of ascertaining the maximum demand,, except where the capacity of the consumer's installation is less than one kilowatt, in which case the maximum demand will be estimated. A discount of one cent per kilowatt hour on the con- sumer's total monthly consumption will be allowed on monthly bills paid on or before ten days after their respective dates. The rate stated in this schedule A covers and includes, for incandescent lighting, the free installation and use of the proper supply of incandescent lamps of the company's pres- ent standard carbon filament types, and of the same voltage,. 158 THE WATTHOUR METER efficiency and candlepower as the incandescent lamps now furnished by the company. An abatement or reduction of one-half cent ( l / 2 c.) per kilowatt-hour from the aforesaid rate shall be allowed to a consumer furnishing, maintaining and renewing all the lamps or other forms of electric illuminants used by him. Schedule B. Regular Power Rate. The following is the regular rate for electricity used for power purposes exclusively, as measured by a meter or meters owned and installed by the company: Eleven cents (lie) per kilowatt hour for all electricity consumed in each month up to and including an amount that would be equal to thirty hours' use of the consumer's maximum demand in such month; and six cents (6c) per kilowatt hour for all electricity consumed in such month in excess of that amount. When the electricity is taken from the company's direct current system, the greatest number of kilowatts used at one time (the peak of the load) in any month shall be deemed the maximum demand for such month; and maximum recording meters will be furnished by the company for the purpose of ascertaining the maximum demand, except where the capacity of the consumer's installation is less than one kilowatt, in which case the maximum demand will be estimated. When the electricity is taken from the company's alter- nating current system, the maximum demand for any month shall be the number of kilowatts equal to a percentage of the total kilowatt capacity represented by all motors connected, which percentage shall be in accordance with the following table of percentages: Where installations are under 10 horsepower, and only one motor is used 85% Where installations are under 10 horsepower, and more than one motor is used '. 75% Where installations are from 10 horsepower to 50 horse- power, both inclusive (irrespective of number of motors) 65% Where installations are over 50 horsepower (irrespective of number of motors) 55% The horsepower capacity of any alternating current motor or motors shall be assumed to be that which is in- dicated by the manufacturer's standard nominal rating or ratings; and each horsepower shall be deemed to be equal RATES 159 to seven hundred and forty-six watts. The company shall, however, have the right ,from time to time, to test any such motor or motors, and if it be found on any such test that the actual horsepower used by such motor or motors exceeds its or their rated capacity, the kilowatt equivalent of the maxi- mum horsepower actually used shall constitute the consumer's maximum demand. A discount of one cent (Ic.) per kilowatt-hour on the con- sumer's total monthly consumption will be allowed on monthly bills paid on or before ten days after their respective dates. The consumer shall pay to the company each month not less than fifty cents (50c.) per horsepower, or fraction thereof, in rated capacity of motor or motors connected. Schedule C. Wholesale Rates for Electricity. Any consumer entering into a written contract to use the company's electricity for either lighting or power, or both, for a period of not less than five years in any single premises occupied by him, will, at his option, be given a wholesale rate for such premises, in lieu of the rates stated in Schedules A and D, which wholesale rate shall consist of both a primary and a secondary charge in accordance with the following specification of charges: Direct Current. Contract Without Guaranty. Primary Charges. For Each Month: $3.20 per kilowatt of the consumer's maximum demand in such month up to and including 20 kilowatts. $2.50 per kilowatt of the excess of the consumer's maximum demand in such month over 20 and up to and including 50 kilowatts. $2.20 per kilowatt of the excess of the consumer's maximum demand in such month over 50 kilowatts. Secondary Charges. For Each Month: 6c per kilowatt-hour for the consumption in such month up to and including 2000 kilowatt hours. 3c per kilowatt-hour for the excess consumption in such month over 2000 and up to and including 5000 kilowatt hours. 1.4c per kilowatt-hour for the excess consumption in such month over 5000 kilowatt-hours. 160 THE WATTHOUR METER Contract with Guaranty. If the consumer will guarantee that his maximum de- mand in each year of the contract term shall be not less than 200 kilowatts, the following primary and secondary charges will be made: Primary Charges. $28.00 per kilowatt per year reckoned upon 200 kilowatts, the guaranteed maximum demand; and $25.00 per kilowatt per year for the excess, if any, over 200 kilowatts of the consumer's actual maximum demand recorded in the year. Such primary charges for each year to be paid by the consumer in installments as follows: At the end of each month he shall pay $2.33 1-3 per kilo- watt reckoned upon 200 kilowatts, and $2.08 1-3 per kilowatt for the excess, if any, over 200 kilowatts of the maximum demand recorded in the year previously to that time. At the end of the year he shall pay the difference, if any, between the sum of the prescribed monthly installments for the year, and the amount constituting the full primary charge for the year. Secondary Charges. For Each Month: 6c per kilowatt-hour for consumption in such month up to and including 2000 kilowatt-hours. 3c per kilowatt-hour for the excess consumption in such month over 2000 and up to and including 5000 kilowatt- hours. 1.4c per kilowatt-hour for the excess consumption in such month over 5000 kilowatt-hours. Alternating Current Transformed. Contract Without Guaranty. Primary Charges. For Each Month: $3.20 per kilowatt of the consumer's maximum demand in such month up to and including 20 kilowatts. $2.20 per kilowatt of the excess of the consumer's maximum demand in such month over 20 and up to and including 50 kilowatts. $2.00 per kilowatt of the excess of the consumer's maximum demand in such month over 50 kilowatts. RATES 161 Secondary Charges. For Each Month: 6c per kilowatt-hour for the consumption in such month up to and including 2000 kilowatt-hours. 3c per kilowatt-hour for the excess consumption in such month over 2000 and up to and including 5000 kilowatt- hours. l.lc per kilowatt-hour for the excess consumption in such month over 5000 and up to and including 30,000 kilowatt- hours. .9c per kilowatt-hour for the excess consumption in such month over 30,000 kilowatt-hours. Contract With Guaranty. If the consumer will guarantee that his maximum demand in each year of the contract term shall be not less than 200 kilowatts, the following primary and secondary charges will be made: Primary Charges. $26.00 per kilowatt per year reckoned upon 200 kilowatts, the guaranteed maximum demand; and $21.50 per kilowatt per year for the excess, if any, over 200 kilowatts of the consumer's actual maximum demand recorded in the year. Such primary charges for each year to be paid by the consumer in installments as follows: At the end of each month he shall pay $2.16 2-3 per kilowatt reckoned upon 200 kilowatts, and $1.75 1-6 per kilowatt for the excess, if any, over 200 kilowatts of the maximum demand recorded in the year previously to that time. At the end of the year he shall pay the difference, if any, between the sum of the prescribed monthly installments for the year, and the amount constituting the full primary charge for the year. Secondary Charges. For Each Month: 6c per kilowatt for the consumption in such month up to and including 2000 kilowatt-hours. 3c per kilowatt-hour for the excess consumption in such month over 2000 and up to and including 5000 kilowatt- hours. 162 THE WATTHOUR METER l.lc per kilowatt-hour for the excess consumption in such month over 5000 and up to and including 30,000 kilowatt hours. .9c per kilowatt-hour for the excess consumption in such month over 30,000 kilowatt hours. Alternating Current Untransformed. Contract With Guaranty. If the consumer will guarantee that his maximum demand in each year of the contract term shall be not less than 200 kilowatts, the following primary and secondary charges will be made: Primary Charges. $25.00 per kilowatt per year reckoned upon 200 kilowatts, the guaranteed maximum demand; and $20.50 per kilowatt per year for the excess, if any, over 20G kilowatts of the consumer's actual maximum demand recorded in the year. Such primary charges for each year to be paid by the consumer in installments as follows: At the end of each month he shall pay $2.08 1-3 pel kilowatt reckoned upon 200 kilowatts, and $1.70 5-6 per kilo- watt for the excess, if any, over 200 kilowatts of the maximum demand recorded in the year previously to that time. At the end of the year he shall pay the difference, if any. between the sum of the prescribed monthly installments for the year, and the amount constituting the 'full primary charge for the year. Secondary Charges. For Each Month: 6c per kilowatt-hour for the consumption in such month up to and including 2000 kilowatt-hours. 2.7c per kilowatt-hour for the excess consumption in such month over 2000 and up to and including 5000 kilowatt- hours. Ic per kilowatt-hour for the excess consumption in sucb month over 5000 and up to and including 30,000 kilowatt hours. .8c per kilowatt-hour for the excess consumption in such month over 30,000 kilowatt-hours. Bills for both primary and secondary charges will be ren dered monthly and a discount of ten per cent. (10%) upon the. RATES 163 secondary charges will be allowed on all bills paid on or before ten days after their respective dates. Schedule U. Automobile Charging in Private Garages. The rate for electricity for charging automobiles in pri- vate garages is either the regular power rate specified in Schedule B, or the power rate under contract for one year or longer specified in Schedule D as the consumer may prefer, subject, however, to the following additional provisions: The net minimum charge to be paid by the consumer each month shall be not less than sixty-six and two-thirds cents (66 2-3c) for each kilowatt of the consumer's maximum demand in such month, and no monthly bill shall be less than one dollar and fifty cents ($1.50). Where alternating current charging boards are used no monthly bill shall be less than one dollar and fifty cents ($1.50) for each charging board. Schedule V. Automobile Charging in Public Garages. The rate for electricity for charging automobiles in public garages is either the regular power rate specified in Schedule B, or the power rate under contract for one year or longer, specified in Schedule D, as the consumer may prefer, subject, however, to the following additional privisions: If the consumer agrees not to make use of the company's service for this purpose during the two hours of the day be- tween four and six o'clock P. M., his net rate for electricity furnished for charging automobiles shall not exceed five cents (5c.) per kilowatt-hour. Where alternating current charging boards are used no monthly bill shall be less than one dollar and fifty cents ($1.50) for each charging board. Schedule W. Rates for "Throw-Over" Switch Service. Where a consumer s premises are supplied with electricity either for light or power, or both, from some plant in the building in which the premises are situated (whether such plant belongs to the consumer or not), and such consumer desires to be in a position to use, or in fact uses, the com- pany's electrical service, not regularly but only occasionally and during the temporary break-down or cessation of such plant; or where a consumer's premises are supplied with power of any kind from any plant in the building in which the premises are situated (whether such plant be an electric plant or not, or be owned by the consumer or not), and such con- sumer desires to be in a position to use, or in fact uses, the 164 THE WATTHOUR METER company's electrical power service, not regularly but only occasionally and during the temporary bread-down or cessa- tion of such plant, the consumer will be charged and must pay to the company for such emergency service the rate herein- after in this schedule provided, to-wit: Such rate will be that specified in Schedule A. D. or E, ac- cording to the purpose for which the service is used, with the additional requirement that the consumer shall pay, irre- spective of the amount of his consumption, a minimum monthly charge depending upon the number and capacity of lamps, motors and other apparatus arranged for connection with the company's service, which charge shall be in accord- ance with the following table of minimum charges: For each incandescent lamp so connected, ten cents (lOc) per month where the lamp has a capacity of fifty (50) watts or less, at rated voltage, and at the rate of ten cents (lOc) per month for fifty (50) watts of capacity where the lamp has a capacity exceeding fifty (50) watts. For each arc lamp so connected one dollar ($1.00) per month where the lamp has a capacity of five hundred (500) watts or less, at rated voltage, and at the rate of one dollar <$1.00) per month for five hundred (500) watts of capacity where the lamp has a capacity exceeding five hundred (500) watts. For each motor so connected, other than a motor used for operating elevators, hoists or similar machinery, one dollar and fifty cents ($1.50) per month per rated horsepower of such motor. For each motor so connected used for operating eleva- tors, hoists or similar machinery, five dollars ($5.00) per month per rated horsepower of such motor. The company will furnish emergency service under this schedule only when the premises are situated on its existing lines having the requisite capacity, and only when the con- sumer signs a contract for the service, running for one year or longer and specifying the number and capacity of lamps, motors or other electrical apparatus in his premises that are to be supplied with the company's electricity during such occasional periods and providing that the consumer shall so arrange his wiring that no lamps, motors or apparatus other than those specified in the contract can be thrown on the com- pany's service by means of switches, or otherwise. For the purpose of this service the company will enter its service main into the building in which the consumer's premises are situated (providing the consumer, in case he shall not own RATES 165 the building, shall obtain the necessary consent from the owner), and the consumer must, at his own expense, install switches and such other equipment as may be necessary for connecting his premises with such service main at the point of entry into the building. For service under this schedule the company will not furnish lamps or renewals for the same. The Edison Electric Illuminating Co., of Boston, Mass. Lighting Rates Commercial. Electricity for any use will be sold, under the following schedule, to any customer who has signed an agreement for electric service, embodying the terms and conditions of the company. A price of 12 cents per kilowatt-hour will be charged for all electricity furnished under this schedule, and the mini- mum charge will be $1.00 per month per meter. Power Rates Commercial. Electricity for power use will be sold, under the following scheduTe, to any consumer who has signed an agreement for electric service, embodying the terms and conditions of the company. "Power" is defined as general motor service, cook- ing, heating, electroplating, charging storage batteries, and similar service, but does not include the running of dynamos for electric lighting purposes. A price of 12 cents per kilowatt-hour will be charged for all electricity furnished under this schedule, with the follow- ing deductions, and the minimum charge will be $1.00 per month per meter: A price of 9 cents per kilowatt-hour will be charged for all electricity furnished in excess of 23 and not exceeding 103 hours' use of the *demand for each month. *The demand is the greatest amount of electricity used by the customer at any one time. Until such time as the company installs one or more indicators, automatically to determine the demand, either in whole or in part, it may estimate the demand. The demand on any circuit, when an indicator is installed, will be the average of the regular monthly readings of the indicator, between October 1st and the following February 1st in each year. The demand so determined, beginning February 1st of each year, shall be the demand for the next twelve months, except that the demand in no case shall be less than 1/3 of the highest reading during the previous twelve months and in no case shall be less than one kilowatt; and provided that if any direct-connected elevator (as defined by the company) be in- stalled the demand shall not be taken at less than 10 kilowatts. The customer has the privilege of having the indicator cut out one night in each month, provided a 48-hour written notice is given to the company. 166 THE WATTHOUR METER A price of 6 cents per kilowatt-hour will be charged for all electricity furnished in excess of 103 hours' use of the demand for each month. Whenever that portion of a customer's bill which is cal- culated at the 9-cent and 6-cent rate, or both, exceeds $10.00 per month, a discount of 70 per cent will be allowed on such excess over $10.00. Whenever a customer's bill, after the foregoing deduc- tions have been made, exceeds $100.00 per month, a discount of 30 per cent will be allowed on all in excess of $100.00. Elevator Rates Commercial. Electricity for direct connected elevator use will be sold, under the following schedule, to any customer who has signed an agreement for electric service, embodying the terms and conditions of the company. A "direct-connected ' elevator is denned as being an elevator running in guides, and in which the car starts at the same time as the motor. A price of 12 cents per kilowatt-hour will be charged for all electricity furnished under this schedule, with the follow- ing deductions, and the minimum charge will be $1.00 per month per meter: A price of 5 cents per kilowatt-hour will be charged for all electricity furnished in excess of 300 kilowatt-hours and not exceeding 600 kilowatt-hours per month. A price of 3 cents per kilowatt-hour will be charged for all electricity furnished in excess of 600 kilowatt-hours and not exceeding 4000 kilowatt-hours per month. A price of 2,y 2 cents per kilowatt-hour will be charged for all electricity furnished in excess of 4000 kilowatt-hours per month. Yearly Lighting Rates Commercial. Electricity for any use will be sold, under the following schedule, to any customer who has signed an agreement for yearly electric service, embodying the terms and conditions of the company. A price of $60.00 per year, payable in equal monthly install- ments will be charged per kilowatt of the *demand up to and including 15 kilowatts. "The demand is the greatest amount of electricity used by the customer at any one time. Until such time as the company installs one or more indicators, automatically to determine the demand, either in whole or in part, it may estimate the demand, but in no case shall it be taken at less than 2/10 of a kilowatt. The demand on any circuit, when an indicator is installed, will be the greatest reading of the indicator between November 1st RATES 167 and the following February 1st of each year, and the demand so determined, beginning- February 1st of each year, shall be the demand called for by the agreemnt for the next twelve months, except that the demand in no case shall be less than 1/3 of the highest reading during the previous twelve months. The customer has the privilege of having the indicator cut out one night in each month, provided a 48-hour written notice is given to the company. A price of $36.00 per year, payable in equal monthly in- stallments, will be charged per kilowatt of the demand for all kilowatts exceeding 15 and up to and including 55. A price of $30.00 per year, payable in equal monthly in- stallments, will be charged per kilowatt of the demand for all kilowatts exceeding 55. These prices do not include the supply of electricity. A price of 5 cents per kilowatt hour will be charged for all electricity furnished under this agreement up to and including 1500 kilowatt-hours per month. A price of 3 cents per kilowatt-hour will be charged for all electricity furnished under this agreement exceeding 1500 kilowatt-hours and up to and including 5500 kilowatt-hours per month. A price of 2^ cents per kilowatt-hour will be charged for all electricity furnished under this agreement exceeding 5500 kilowatt-hours per month. Permanent Electric Rates. Electricity for any use in specified premises will be sold, under the following schedule, to any customer who has signed an agreement for at least 50 kilowatts of permanent electric service, embodying the terms and conditions of the company. A price of $60.00 per year, payable in equal monthly in- stallments, will be charged per kilowatt of service up to and including 15 kilowatts. A price of $36.00 per year, payable in equal monthly in- stallments, will be charged per kilowatt of service for all kilowatts exceeding 15 and up to and including 55. A price of $30.00 per year, payable in equal monthly in- stallments, will be charged per kilowatt of service for all kilowatts exceeding 55. These prices do not include the supply of electricity. A price of 5 cents per kilowatt-hour will be charged for all electricity furnished under this agreement up to and includ- ing 1500 kilowatt-hours per month. A price of 3 cents per kilowatt-hour will be charged for all electricity furnished under this agreement exceeding 1500 168 THE WATTHOUR METER kilowatt-hours and up to and including 5500 kilowatt-hours per month. A price of iy 2 cents per kilowatt-hour will be charged for all electricity furnished under this agreement exceeding 5500 kilowatt-hours and up to and including 105,500 kilowatt-hours per month. A price of 1*4 cents per kilowatt-hour will be charged for all electricity furnished under this agreement exceeding 105,- 500 kilowatt-hours per month. The company will deliver its electricity at the customer's premises, and, in consideration of not supplying lamps and care, will deduct from the net amount of the bill, as otherwise rendered, y 2 cent per kilowatt-hour. The company will provide capacity for intermittent over- loads up to 40 per cent in excess of the kilowatts applied for by the customer. An excess price of 20 cents per kilowatt hour will be charged for all electricity furnished at any time in excess of the kilowatts applied for by the customer. Terms and Conditions. For the purpose of determining the amount of electricity used, a meter shall be installed by the company upon the cus- tomer's premises at a point most convenient for the com- pany's service, upon the reading of which meter all bills shall be calculated. If more than one meter is installed, unless for the company's convenience, each meter shall be consid- ered by itself in calculating the amount of the bill. When more than one meter or discount indicator is installed under this agreement, for the company's convenience, the sums of the consumptions and demands shall, in all cases, be taken as the total consumption and demand. All bills shall be due and payable upon presentation and shall be rendered monthly, unless either the customer or the company desires bills rendered weekly, in which case it may be done by adjusting to a weekly basis all the monthly figures referred to in the schedule of rates. A minimum charge will be made of $1.00 per month per meter, unless otherwise provided. The customer will be responsible for all charges for elec- tricity furnished under this agreement until the end of the term thereof and for such further time as he may continue to take the service; except that where the customer has the right to terminate the agreement by notice, which shall be in RATES 169 writing, he shall remain liable for all charges for ten days thereafter. The customer will be responsible for all damage to, or loss of, the company's property located upon his premises unless occasioned by the company's negligence. The company shall not be responsible for any failure to supply electricity, or for interruption or reversal of the sup- ply, if such failure, interruption or reversal is without de- fault or neglect on its part. The company reserves the right to install a circuit breaker, so arranged as to disconnect the service in the premises, if the company's capacity at that point is exceeded. If a customer, who is not paying a rate calling for an annual fixed cost, desires to use the electric service as auxiliary to another source of power (excluding, however, small sources of power not exceeding two horsepower) he may do so only by paying a minimum charge of $3.00 per month per kilowatt for as many kilowatts as it is possible for him to use on the service at any one time; this number to be determined by a circuit breaker, so arranged as to discon- nect the service if the number of kilowatts is exceeded. If lamps and care are supplied under this agreement, they will be supplied only for such installation as uses the company's electricity exclusively. It is agreed that all lamps, plugs, meters and such other appliances as are furnished by the company shall remain its property. And it is further agreed that all wiring upon the premises of the customer, to which the company's service is to be connected, shall be so installed that the company may carry out this contract, and shall be kept in proper condition by the customer. Permission is given the Company to enter the custom- er's premises, at all times, for the purpose of inspecting and keeping in repair or removing any or all of its apparatus used in connection with the supply of electricity, and for said purpose the customer hereby authorizes and requests his landlord, if any, to permit said company to enter said premises. The benefits and obligations of this contract shall inure to and be binding upon the successors and assigns, survivors and executors or administrators (as the case may be) of the original parties hereto, respectively, for the full term of this contract. 170 THE WATTHOUR METER Birmingham (Alabama) Railway, Light & Power Co. The consumer hereby agrees to pay the company, monthly, within ten days after presentation of bills, for said incan- descent light service at the base rate of twelve cents per kilowatt hour, as measured by meter or meters to be furnished and installed by the company, subject to the following dis- counts : LIGHTING RATES. On monthly bills under " " over 25 k.w.h. 10 25 150 250 400 500 1,000 1,500 2,000 2,500 3,000 3,500 % 40 % 45 % 47V 2 % 50 % The consumer agrees to pay the company a net minimum monthly bill of ( $ ) Dollars as a readiness to serve charge. POWER RATES. Discount for Monthly Usage. 500 1000 k.w.h. per month 10% discount 1000 2000 " " " 12% 2000 3000 " " 14% 3000 4000 " " " 16% 4000 5000 " " " 18% Over 5000 " " 20% In addition 10 to 15 k.w.h. used per h.p. installed 4% discount 15 20 " " " " " 6% 20 25 " " 8% 25 30 " " " " " 10% 30 35 " " " " " 12% 35 40 ' " 14% 40 45 " " 16% 45 50 " " " " " 18% 50 55 " " 20% 55 60 " " " " " 22% 60 65 " " " " " 24% 5 70 " " " " " 26% 70 75 " " " " " 28% 75 gO " " " " " 30% 80 85 " " " " " 32% 85 90 " " " " " 34% 90 95 " " " " . " 36% 35 100 " " " " " 38% Over 100 " " " " 40% 10% additional for off peak service. RATES 171 The San Francisco Gas & Electric Co., of San Fran- cisco, Cal. Lighting Rates. All current delivered to the consumer shall be registered by a meter or meters installed and owned by the Company and the rate and minimum charge shall be as follows per meter: per 6% 6 i* 4% 4% k.w.h. cents if the monthly consumption be less than from k.w.h. 79 go 129 130 189 190 244 245 309 310 374 375 449 550 674 675 849 8501099 1100 or over The consumer agrees to pay the company a net monthly minimum bill of $1.00 for each meter installed, provided the bill for current consumed at the above premises, does not equal or exceed the said sum of $1.00 per month. The consumer further agrees to pay a minimum of $1.50 per month, for each horsepower in motor or motors installed, provided the bill for current consumed does not equal or ex- ceed the said sum of $1.50 per horsepower per month. Power Rates. All current delivered to the Consumer shall be registered by a meter or meters installed and owned by the Company and the rate and minimum charge shall be as follows per meter: Per 1000 9 cents 8.55 8.1 7.65 7.2 6.75 6.3 5.85 5.4 Per k.w.h. 5.25 cents 5 4.75 4.50 4.25 4 If the monthly consumption I. above the minimum and \ ' shall be from 2,000 to 3,000 w.h. per 16 c.p. lamp. " ' " 3,000 to 4,000 4,000 to 5,000 5,000 to 6,000 6,000 to 7,000 " ' " 7,000 to 8,000 8, 000 to 9, 000 " ' " 9,000 or over If the monthly consumption shall be from 375 to 649 k.w.h. 650 to 924 " ' " 925 to 1,199 1,200 to 1,474 " 1,475 to 1,749 " " 1,750 or over 172 THE WATTHOUR METER The consumer agrees to pay to the company a net monthly minimum bill of $1.00 for each meter installed, pro- vided the bill for current consumed at the above premises does not equal or exceed the said sum of $1.00 per month. The consumer further agrees to pay a minimum of $1.50 per month, for each horsepower or fraction thereof in motor or motors installed, provided the bill for current consumed does not equal or exceed the said sum of $1.50 per horse- power or fraction thereof per month. From the data given above, a very good general idea can be gained of the fixing and application of lighting and power rates throughout the United States, though of course local conditions may cause a wide variation from the figures given. APPENDIX. Definitions. In many respects electric circuits closely resemble a water system, in which the pressure is analogous to the voltage of the electric circuit and the quantity (in cubic feet per second) to the current flowing in the wires. This comparison will often aid in the solution of various electrical problems. Definitions. Ampere=the unit of electrical current, and is that current which will deposit silver at the rate of o.ooiiiS grams per second when flowing through an electrolytic solution of silver nitrate. Ohm=the unit of resistance, and is equivalent to the resistance of a column of pure mercury, at o cen- tigrade, 103.6 centimeters high, of uniform cross sec- tion, and weighing 14.4521 grams. Volt=the unit of electrical pressure (electro- motive-force), and is that pressure which will maintain the flow of one ampere of current against the resist- ance of one ohm. Let R=the resistance of a given circuit, E the voltage impressed and I, the current in amperes, then for direct current, E = R X I (Ohm's Law), which is the fundamental equation of direct current circuits. Watt=the unit of electrical power; the watts equal the product of the volts and the amperes in direct current circuits, and to the product of the volts, the amperes and the power factor in alternating current circuits. (See Chap. II.) Kilowatt=one thousand watts. 174 THE WATTHOUR METER Watthour=the unit of electrical energy, and is equivalent to the flow of one watt for one hour. Kilowatt-hour=one thousand watt hours. Inductance : The inductance of an electrical circuit is the property of that circuit whereby it can convert electric energy into magnetic energy, and vice versa. Inductance bears a close resemblance to "inertia" in mechanics, and has been called "electrical inertia." Inertia is that property of a moving body whereby it resists any change in its velocity ; if a rapidly moving body is suddenly stopped, as, for example, a hammer striking a nail, a great force is exerted by the body against the obstacle which brings it to rest ; the mag- nitude 'of the force depending upon the suddenness with which the moving body is stopped and upon the inertia of the body. In an electric circuit containing inductance, if the current is suddenly interrupted, a high e.m.f. is produced which tends to cause the cur- rent to continue. The magnitude of this "induced" e.m.f. depends upon the suddenness with which the current is interrupted and upon the inductance of the circuit. The inertia of any given body is proportional to its mass ; the energy stored in a moving body is W = T/2 MV 2 , where M is the mass (= weight divided by the gravitational constant), and where V is the velocity. The magnetic energy stored in an electric circuit due to its current and inductance is W = ^2 LI 2 , where L is the inductance and I is the current. Henry=unit of inductance ; a circuit having an in- ductance of one henry will have an e.m.f. of one volt induced in it by a current changing at the rate of one ampere per second. Milli-henry=o.ooi henry. Power Factor=the ratio of true watts to apparent watts (see Chap. II). Cycle one complete wave or alternation of cur- rent or e.m.f. (See Fig. , Chap. II.) Frequency=number of cycles per second. APPENDIX 175 Impedance=the vector sum of the resistance and the reactance of an electric circuit and is expressed by the following" equation : Z = V R 2 + X 2 , in which R is the resistance in ohms, and X is the reactance (= 2 TT i X L, where f is the frequency in cycles per second and L is the inductance in henrys). The voltage drop in an alternating current circuit con- taining both reactance and resistance is E = ZX I, where Z is the impedance as above expressed and I is the current in amperes. Determination of Temperature Rise by Resistance Method. In testing electrical machinery such as generators, motors and transformers, it is impossible to obtain the internal temperature of the windings by use of ther- mometers ; the following formula will therefore be use- ful in determining the average temperature of such windings : rFl \ - 1 IdegreesC, in which 238 is a constant ; R is the initial resistance of the winding at a room temperature, t, and F is the final resistance. If the room temperature differs from 25 C, the calculated temperature should be corrected by l /2% for each degree C. Thus with a room temperature of 15, the rise in temperature should be increased by 5%, or if the room temperature is 35, the rise in tem- perature as calculated should be decreased by 5%, etc. Adjusting Meters for Use With Current and Potential Transformers. The usual method of testing watthour meters used with current and potential transformers is to test and adjust the meters without the transformers, taking into consideration, of course, the ratio of the transformers. Where a great degree of accuracy is not required, this 176 THE WATTHOUR METER procedure will answer very well ; but where, as with large consumers, a small percentage error represents a considerable sum of money, the errors introduced by the current and potential transformers should be taken into account and, as far as may be, compensated. The errors introduced by the current and potential transformers are (i) errors in ratio of the transformers and (2) errors due to improper phase relations between the primary and secondary currents and e.m.f's. The errors due to ratio can be easily compensated by adjusting the meter in accordance with the ratio curves of the transformers at unity power fatcor. The ratio of the potential transformer will remain constant as its load is constant. The ratio of the current trans- former will not remain constant, but will vary with the load. It tends to make the meter fast at full load and slow at light load. This can be compensated by ad- justing the meter to be a little slow at full load and fast at light load. The errors due to improper phase relations be- tween the primary and secondary currents and e.m.f.'s are of more serious nature and more difficult to elim- inate than errors due to ratio. The errors from this source are negligible at unity power factor, but may be considerable at low power factors, depending on the design of the transformers. The diagram, Fig. I, shown below, is a vector dia- .gram of a current transformer. OI is the primary or Fig. 1. APPENDIX 177 line current, OI' is the secondary current and lags be- hind OE', the secondary e.m.f. of the current trans- former, by an angle depending on the power factor of the secondary load (meter coils and leads). II" is the exciting current, the magnetizing component of which is at right angles to and the energy component in phase with the primary e.m.f., OE, of the current trans- former. OI" is that component of the primary current inducing a current in the secondary, or is the second- ary referred back to the primary. It will be seen that OI' leads OI by and angle a which will tend to make the meter run fast on inductive loads. Fig. 2 is the vector diagram of a potential trans- former. OE is the primary e.m.f. OE' is the secondary e.m.f. OI is the primary current, OF the secondary Fig. 2. current and IF the exciting current. E"N is the e.m.f. consumed by resistance or is the RI drop, EN the e.m.f. consumed by the reactance or is the XI drop, and EE' is the impedance e.m.f. of the transformer. In the fig- ure it will be seen that the secondary e.m.f. (OE" re- ferred to the primary) leads the primary e.m.f by an angle a'. This will tend to make the meter run slow on inductive loads. Obviously, if a of the current transformer equals a of the potential transformer there will be no error from this source, and it is attempted in the design of meter transformers to approximate this condition. For this reason considerable resistance is introduced into the 178 THE WATTHOUR METER potential transformer to increase this angle. This high resistance results in a transformer of poor regulation, but the regulation is not important, as the load is con- stant and the range recommended narrow. In practice, the angle a of the current transformer is greater than the angle a of the potential transformer. This difference can be compensated in the lag of the meter. Suppose, for example, the meter is not lagged properly by the angle = O there will be no error from this source. In other words, the meter is lagged so that it will run slow on inductive loads. The combined effect of angular displacement in the potential transformer and of the meter not being lagged quite ninety degrees just compensates for the angular displacement of the current transformer. These angles may be determined and corrections made as follows : The results obtained by this method :are sufficiently accurate for adjusting service meters since finer corrections could not be made on the meters themselves. With an indicating wattmeter the power flowing in a circuit of 50 per cent power factor (or other known power factor) is read. A one to one ratio current transformer of the type used with the watthour meters is then inserted in the circuit and the current for the watt-meter is taken from the secondary of the current transformer. The power is again read. Assuming that a power factor of 50 per cent is used, from the first reading we get W = El cos 60 and from the second reading W = El cos (60 a). The angle a is the angular displacement due to the current transformer. From the above W cos 60 COS (60 a) = W From which we can readily obtain a. Before substitut- ing in the formula W should be corrected for error in transformer ratio. The current coils of the watt-meter are again connected directly in circuit and the potential supplied by potential transformers, two transformers APPENDIX 179 being used, one to step up from the line voltage and the other to step down again to the watt-meter. We can now obtain the angular displacement due to the potential transformers by applying the same formula as given above for current transformers. The angle thus obtained for the potential transformers will be twice the angle of one transformer, and as it is the angle of one transformer with which we are concerned the result obtained should be divided by two. This will give us the angle a'. We will then lag the meter, not for 90, but for 90- - (a a'). Another and quicker way of applying this method is to determine the error introduced by two potential transformers at unity and at 50 per cent power factor. Each transformer is responsible for ^ of the error. Now connect the watthour meter in circuit with the current transformer and with two potential trans- formers, one potential transformer, stepping up from the line voltage (testing circuit) and the other stepping down to the watthour meter. The indicating watt-meter should be connected di- rectly in the circuit without current or potential transformers. The watthour-meter is than adjusted at unity power factor and at 50 per cent power factor to disagree with the indicating watt-meter by the amount of the error due to one potential transformer. There will be errors due to three transformers, one current and two potential. The w r atthour-meter is to operate with but two transformers and should be adjusted to compeiir'vte for the errors of only two transformers. By not compensating for the error of one of the poten- tial transformers, as outlined above, the desired results will be accomplished. It is not strictly correct to take l / 2 the error of the two potential transformers, as outlined above, as the error of one transformer; it is, however, very close, closer than adjustments can be made on the watthour- meter. INDEX Accuracy, equation and per cent, 127, 128, 139. factors affecting, 3, 5. induction meter, curve of, 43. initial, 3. Adjustments, 139, 30, 175. commutating meters, 70, 75. for friction, 6. polyphase meter elements, 6i Ampere, definition, 173. Ampere-hour meters (see mer- cury flotation). Armatures, commutating meters, 72. astatic arrangement, 82. open-circuited, 153. Backward rotation, induction me- ters, 116. commutating meters, 153. Balance, three-wire systems, 78. Balance loop, induction meters, 61. Bearings, upper, 9, 154. construction, lower, 4, 10. Brushes, 71. Calibration, curves of, 43, 147, 148. Capacity, selection of, 7, 115. overload, 7. Commutating meters, alternating- currents, 75. adjustments, 70, 75. armature construction, 72. armature, open circuits in, 153. astatic arrangement of, 82. backward rotation, 153. brushes, 71. commutator, 71, 153. comparison to shunt motor, 66. Commutating meter, compensating field, 70. constants (see testing). construction, general, 68. efficiency, 68. four pole type, 83. heating of frames, 8. lagging, 75. switchboard type, 80. temperature coefficient, 74. three wire type, 77. parts, 74. principle of operation, 65. Compensating field, 70. Connection diagrams, commutat- ing type, 83 to 86. indicating instruments, 18 to 22. 2 single phase meters on 3 wire single phase circuit, 44. 2 single phase meters on 4 wire 2 phase circuit, 46. 2 single phase meters on 3 wire 3 phase circuit, 46, 50, 53. 3 single phase meters on 4 wire 3 phase circuit, 54. 3 single phase meters on 6 hase circuit, 57. single phase 3 wire meter, 44. single phase meter with 3 wire transformer, 45. 1 polyphase meter, 3 and 4 wire circuits, 60. 2 polyphase meter on 6 phase circuits, 62. 3 phase, 4 wire meter, 58. manufacturer's diagrams, spe- cial note, 64. measurement of power, 15 to 23. mercury meter, 94. metering high potential circuits. 63. Constants (see testing). Construction, general, 7. Covers, 7. Creeping, 6, 38, 152. Current transformers (see trans- formers). 182 INDEX Definitions (see Appendix also), 1. Demand indicators, 100. Diagrams (see connections). Discs, troubles, 152, 154. construction and material, 10. Distributing system, losses in, 2. Duncan meter, 68, 124. E Efficiency, high, meters, 68. low, meters, 68, 14. Elements, adjustments of, 61. interference of, 62. Formulae (see testing). Fort Wajme meters, adjustments, 30, 36. constants, 122, 123, 127, 128. Frames, 7. Frequency, effect of variations of, 39 to 43. Frequency, definition of, 174. "Friction torque," 6. Gear, worm, 9. ratio, 120. General Electric meters, adjust- ments, 30, 33. constants, 119, 120, 127, 128. Glass covers, reasons for not using, 8. Grounding secondaries of trans- formers, 63. H Heating of frames, commutating meters, 8. Henry, definition of, 174. Humming of induction meters, 151. Impedance, definition of, 175. Inductance, definition of, 174. Induction meters, adjustments, 30, 33, 36, 37, 39, 61, 175. accuracy of (calibration curve), 43. backward rotation of, 116. Induction meters, used as balanced load indicator, 52. "balance loop," 61. calibration, curve of, 43, 147, 148. connections of (see connection, diagrams). constants (see testing). creeping, causes of, 38. double lagging, 33, 39. elements, interference of, 62. Fort Wayne, adjustments of, 30, 36. Fort Wayne, constants of, .'122, 123, 127, 128. frequency, effect of variation of, 39 to 43. General Electric, adjustments of, 30, 33. General Electric, constants of, 119, 120, 127, 128. humming of, 151. interference of elements, 62. lagging, 28, 33, 39. light load adjustment device, 33, 36, 37. Vector diagram of, 29. parts of (illustration), 35. polyphase type of, 58. 4 wire type, 58. power factor, influence of, 28. adjustment devices, 30. determination, by use of, 55. principle of operation, 24. reasons for extensive use induc- tion type, 24. single phase meters, advantages of, 53. single phase meters on poly- phase circuits, 45 to 58. single phase meters on unbal- anced 3 phase circuits, 49. speed of, 121. standard meters on old volt- ages, 54. . 3 wire type, 44. Westinghouse, adjustments of, 30, 37. \Vestinghouse, constants of, 121, 127, 128. Indicators, demand, 100, 104. Installation, 114, 116. Inspection records, 116. Interference of elements, poly- phase meters, 62. INDEX 183 Jewel, lower bearing, 4. Jewels, installation of, 117. selection of, and life, 11. testing for defective, 153. K Knopp method of testing, 132. Knopp "milli-hour" stop watch, 133. Lagging (see commutating type). (See induction type). Leveling, conventional method of, 115. Light load adjustment (see com- mutating induction meter). Lightning, protection of current transformers from, 64. Locations of meters, 114. Losses in distributing system, 2. M Magnets, retarding, 5. process of manufacture, 13. weakening of, 151. Maximum demand indicators, 104. Measurement of power, 15 to 23. Mercury flotation meter, alternat- ing current type, 92. ampere-hour type, 93. direct current type, 89. connections of, 94. constants and testing (see test- ing). principle of operation, 87. Ohm, definition of, 173. Overload capacity, 7. Overmetering, 6, 80, 115. Percentage of error, equation of, 127, 128, 139. Pivots, 4. Phantom loads, 130. Power factor, testing for leading or lagging, 132, 141. adjustments for, 28. 29, 30, 75. determination of, 55. graphically represented, 16. Power, measurement of (see con- nection diagrams). definition of, 22. equations of, 17. Prepayment meters, 96. R Rates, 155. Ratio, soecial potential trans- former, 54. Reading, systems of, 107. Recording watt-meter, definition of, 1. Recording mechanism, 5, 13, 154. Records, systems of, 107, 116. Register ratio, 121, 149. Revenue, relation to meter sys- tem, 1. Rotating standard, method of test- ing (see testing). Sangamo meters (see mercury flotation). Selection of meters, 3, 7, 83. Shunt field (see commutating meter). Shafts, 9, 152. Switchboard type, commutating meters, 80. Temperature rise, determination of, 175. Testing adjustments (see adjust- ments), constants and formulae, Duncan, 124. Fort Wayne, 122, 123, 127, 128. Gen. Elec., 119, 120, 127, 128. Sangamo, 125. Westinghouse, 121, 127, 128. equations of accuracy, 127, 128, 139. methods of indicating instru- ment, 118. 184 INDEX Testing for shop work, 140. Knopp, 132. phantom load transformer, 130. rotating standard, 125, 128. special portable set, 137. polyphase meters, 144. records, 113. Three wire meters, commutating type, 77. induction type, 44. Three wire system, balance of, 78. Torque, value of high, 4. Total output meters, 80. Transformers, current, errors of, 145, 175. calibration curve of, 147. double primary, 45. grounding of secondaries, 63. loads imposed upon, 64. protection from lightning, 64. Transformers, phantom load, 130. Transformers, potential, errors >f, 148, 175. grounding of secondaries, 63. loads imposed upon, 64. special connection of, 54. Troubles, general, 150. Vibration, methods of preventing. 150. (See installation). Volt, definition of, 173. W Watches, Knopp's "milli-hour/ 133. Watt, definition of, 173. Watthour, definition of, 174. Watthour meter, definition of, 1. Worm gear, 9. 7 DAY USE RETURN TO DESK FROM WHICH BORROWED fl stamped below. ITY OF CALIFORNIA LIBRARY OF THE UNIVERSITY OF CALIFORNIA 5ITY OF CALIFORNIA LIBRARY OF THE UNIVERSITY OF CALIFORNIA SITY OF C1LIFORNU LIBRARY OF THE UNIVERSITY OF CALIFORNIA