ELEMENTS OF POWER SYSTEM ANALYSIS WILLIAM D. STEVENSON, JR. Professor of Elecirical Engineering North Carolina State College McGRAW-HILL PUBLISHING COMI ew yors. LONDON VERSION ADDS NO WA ena & < we rint-driNgMcGraw-Hill Ele Freve | and Electronic Engineering Series ELEMENTS OF POWER SYSTEM ANALYSIS VERSION ADDS NO WA ena & < we rint-driNgMcGraw-Hill Electrical and Electronic Engineering Series Frepenick Exacoxs Tarwax, Consulting Baitor Bary axp Gavur - Alternating-current Machinery BERANEK * Acoustics ace - Theory and Application of Industrial Electronies Cucera « Harmonics, Sidebands, and Transients in Communication Engineering Easrwan + Fundamentals of Vacuum Tubes Evans - Control-system Dynamies Frrzcnnato ap Kivosuey « Elects Grrent + Basie Electron Tubes Haren. ap Hosseimenrn » Engineering Electronies Hanwan » Fundamentals of Blectronie Motion Hesse axp Canny * Fundamentals of Blectrieal Engineering, Hie - Electronies in Engineering Jonsson + Transmission Lines and Networks Krave » Antennas Knaus + Electromagnetics LxPacr - Analysis of Alternating-current Cirenits LePage. asp Srety « General Network Analysis Miniwaw AND Seeny + Electronics Rogers + Introduction to Electrie Fields RopeNnens + Transient Performance of Electrie Power Systems Suey - Electron-tube Circuits Susxrxp + Direct-current Machinery SxILUING * Electric Transmission Lines Sua ‘Transient Bleetrie Curreits Spaxaexnurs + Vacuum Tubes Sruvenson - Elements of Powver System Analysis ‘Tentan + Radio Engineering,” ‘Tenwan asp Perri - Electronic Measuréments ‘Tuaren AND Brown « Servomechanism Analysis ¢ Machinery ERED “4 d REGISTERED 9) VERSION ADDS NOELEMENTS OF POWER SYSTEM ANALYSIS comyricatr 1955 British isue of frst edition, 1955, published in London, 1985, by McGraw-Hill Publishing Company Ltd, by arrangement with MeGraw-Hill Book Company Ine., New York Copyright in U.S.A. 1955, by the McGraw-Hill Book Company Ine., New York VERSION ADDS NO WA ena & , PoE! 'GISTERED ON VERSION ADDS NO WATERMARK & LLL ‘The Occurrence of Unsymmetrical Systems 12 The Beets of Unsymmetrieal Cievuits 143 General Equations Involving Unsymmetrical Ser LS AN ison of the cost of transporting energy icity, se R. E. Pierce and E. E, George, “Bzond “Transmission,” Trane. ATEE, vol. 67, pp. 1089-1004, 1] 1 * < Crint-drig2 ELEMENTS. OF POWER SYSTEM ANALYSIS. from water power available at remote points, ‘The water power is con- verted to electric power at the souree and transported by transmission lines to the point where it is converted to the desired form, such as light, heat, mechanieal energy, or chemical energy. ‘The transmission line cannot store energy, and all the energy furnished at the generating station is converted simultaneously at the load, except for the losses in the system, An electric power system consists of three principal eomponents: the generating stations, the transmission lines, and the distribution systems. ‘The transmission lines are the connecting links between all the generating stations and the distribution systems. A distribution system conneets all the individual loads in a given area to the transmission lines. A well- developed power system integrates a large number of generating stations so that their combined output is readily available throughout the region served. The locations of hydro stations are fixed by the presence of water power, but the choice of sites for steam stations is more flexible. Steam stations are usually spotted throughout the system so that. there is at least one generating plant near each large load center. ‘Thus, hydro stations often require the transmission of large amounts of power over long distances, but. steam plants usually require transmission over shorter distances. ‘The growth of loads may not be under the control of the power company, but often the availability of cheap power encour- ages the growth of loads in such favored areas. One job of the power engineer is to predict the future demand for power so that suitably located generating stations and well-coordinated, flexible, and reliable transmission gystems will be ready to supply the demand through enlarged distribution systems as required by the load. As the system grows, more energy sources must be exploited to satisfy the increasing demand, and more transmission lines must be built to link the new generating stations to each other, to an increasing number of distribution points, and to other power systems.* 1.2 The Growth of Electric Power Systems. ‘The development of a-c systems began in the United States in 1885 when George Westinghouse bought the American patents covering the a-e transmission system developed by L. Gaulard and J. D. Gibbs of Paris. William Stanley, an early associnte of Westinghouse, tested transformers in his Iggoray Great Barrington, Massachusetts, There, in the winter off Stanley installed the first experimental a-e distributi supplied 150 lamps in the town. ‘The first a-c t United States was put into operation in 1890 td For a description of the development of « largo Sporn, “"The Integrated Power System,” MeGraw-Hil York, 1950. VERSION ADDS NO WATERMARK & o 7 < Crint-drigGENERAL BACKGROUND a generated by water power a distance of 13 miles from Willamette Falls to Portland, Oregon.? ‘The first transmission lines were single-phase, and the energy was usually consumed for lighting only. Even the first motors were single- phase, but on May 16, 1888, Nikola Tesla presented a paper describing two-phase induction and synchronous motors. The advantages of poly- phase motors were apparent immediately, and a two-phase a-c distribu- tion system was demonstrated to the publie at the Columbian Exposi in Chicago in 1803, ‘Thereafter, the transmission of electrie energy by altemating current, especially three-phase alternating current, gradually replaced d-e systems, In January, 18M, there were five polyphase generating plants in the United States, of which one was two-phase and. the others three-phase.* One reason for the early acceptance of a-e systems was the transformer, which makes possible the transmission of electric energy at @ voltage higher than the voltage of generation or utilization. A higher voltage of transmission requires less line current for the transmission of a given amount of power and, therefore, results in lower 12 losses in the line. An ae generator is a simpler deviee than a d-c generator, and this is an additional advantage of a-e systems. Although most of the clectrie energy consumed in the United States is transmitted as alternating current, experiments have been carried on for ‘a number of years in this countzy on a system composed of a-e generators foeding a d-e transmission line through a transformer and an electron rectifier. Tn this system an electronic inverter changes the direct eurrent to alternating current at the end of the line so that the voltage can be reduced by a transformer. Direct-current transmission has been more popular in Europe, and most of the recent literature on d-e transmission has been published in Germany, England, and Russia. Direet-current transmission overcomes some of the disadvantages of a-e systems, as will become apparent: as the characteristies of a-c systems are studied. The disadvantage of elaborate inverting and rectifying equipment makes d-c © Much interesting material about the early development of electric equipment snd apparatus ean be found in the volumes of Transactions of the American Institute of Electrical Engineers for the period. For instance, 2 good deseription of thogla- rette-Portland line is given in C. F. Scott, “Long Distance Transmi and Power,” Trane. ALBE, vol. 9, pp. 425-442, 1802. For abou ‘liseoveries and developments which gave impetus to the elect PRR R GIQO wre AN ite Moths te Ra SREGISTERED Q) VERSION ADDS NO WATERMARK &, vol. 9, p. 27, 180 oO G * < Crint-drig4 ELEMENTS OF POWER SYSTEM ANALYSIS. transmission Jess economical than a-¢ systems for distances less than 450 or 500 miles, and there is some doubt. whether d-e transmission will ever be as reliable as alternating current. ‘The Federal Power Commission publishes monthly reports on various aspects of the generation and transmission of electric energy. ‘The first fof continuous series of annual reports giving data collected by the Federal Power Commission appeared in 1920, ‘Table 1.1 gives statisties on the total installed eapacity of generators and on the annual production ‘Tamu LA Ixenaraxo Enncrnic precrmcas, Caracrry asp ANNUAL PRODUCHION OF cenoy 1W Tue Usiren Srares* Year | lnstelled capacity, | Annual eneray production, kw whe 120 | 12,718,608 39, 404,639,000 1930 | 82,885,365 91,111,548,000 1 | 39,026,881 | 141, 837,010,000, 1950 | 68,019,010 329, 141,343,000, * Bouree: Fede Power Commision of electric energy in the United States at ten-year intervals since 1920. Although these statisties record the growth of power systems in the first half of the twentieth century, statisties alone do not show the impact of the two world wars on the electrical industry. World War I revealed the neod for interconnection of power systems operating on a standard frequency in order to furnish larger blocks of power than were available from individual systems. Both wars dramatized the role of electricity in building military power, and both were followed by a greater demand for electric energy. Prior to World War IL, the greatest net increas in one year in the installed capacity of genersting stations was 3,791,000 kw, in 1925. This figure was not surpassed until 1948, when the net increase in one year was 4,287,831 kye. Since then, statisties on yearly growth indicate a doubling of installed capacity every ten years. Annual energy production is also expected to double every ten years and reach 1 talllion kwhr in 1968. Tn the early days of a-c power transmission in the United States, the operating voltage increased rapidly. Tn 1800 the Willametig Porgmsd line was operated at 3,300 volts, In 1907 a line was operatin Voltage rose to 150 ky in 1913, 220 ky in 1923, 244 ky ky on the line from Hoover Dam to Los Angeles, which 1936. Tn 1952, construction was completed on 4 }-kv system of t and Blectrie Cy * See P. Sporn, EL, Peterson, T. W. Gross, and H. P Extre-high-voltege Transmission System of the Amery pany,’ Trans, ATBE, vol. 7, pp. 81-72, 1961 Mion of a ° ON Sen S REGISTERED 9, q VERSION ADDS NO WATERMARK &, & e Ameri * < Crint-ariGENERAL BACKGROUND 5 of line voltage is principally a matter of balancing the initial investment in lino construction and apparatus with the cost of operation. Up toa certain point, inereasing the voltage results in lower losses for a given size of conductor or in a smaller conductor for a given power loss. Much of the saving achieved in conductor cost by designing for higher voltage is lost becatise of the increased loss in the surrounding air, which’ is ionized by the high voltage gradient at the wire, and because of the increased cost of insulators, transformers, switehes, and circuit breakers. “The cost of the latter items increases so rapidly at the higher voltages that some maximum voltage exists above which it is not economical to design transmission lines at present. Radio inffuence is also a factor affecting the selection of voltage. The final determination of a system voltage of 330 ky for the above-mentioned line resulted from studies of tests on a 500-ky experimental line near the Tid station of the Ohio Power Company. ‘The Ameriean Gas and Electric Service Corporation, in cooperation with eight manufacturers of high-voltage equipment, obtained data on insulators, Hine conductors, switchgear, transformers, lightning arresters, instruments, radio influence, and the effects of atmos: pherie ionigation at voltages up to 500 kv. ‘Through such tests and through experience in operating a 600-mile, 400-Ky line in Sweden, the economic and technical limitations of high-voltage transmission are being studied." Until 1917, electric systems were usally operated as individual units because they started as isolated eystems and spread out only gradually to cover the whole country. ‘The demand for large blocks of power and increased reliability suggested the interconnection of neighboring systems. Interconneetion is advantageous economically because fewer machines are required a8 a rescrve for operation at peak loads (reserve capacity) and fewer machines running without load are required to take eare of sudden, unexpected jumps in load (spinning reserve). ‘The reduction in machines is possible because one company can usually call on neighboring companies for additional power. Interconnection also allows a company to take advantage of the most economical sources of power, and a com pany may find it cheaper to buy power than to generate it in an obsolete plant. Interconnection has inereased to the point where pow exchanged between the systems of different companies as routine. Figure 1.1 is the map of a small transmission 3 shows eight points of interconnection with other sys tinued service of systems depending on water p Ko part “oN of their generation is possible in times of unusug ‘treme water 'S-REGISTERED ON Power Transmission nnd the Novel Features of the Syey vol. 72, Dt, pp 1089-2095, 1053 VERSION ADDS NO WATERMARK &, & * < Crint-drig‘ NENTS OF POWER SYSTEM ANALYSS shortage only because of the power obtained from other systems through intereonnections. Interconnection of systems brought many nev problems, most of whieh have been solved satisfactorily. Iaterconnection increases the umount of current whieh flows when short circuit occurs on a system and requires the installation of breakers able to interrupt-a larger current. Lr [="REGISTERED 9) VERSION ADDS NO WATERMARK &, & * < Crint-drigGENERAL BACKGROUND ry transient conditions. ‘There is a definite limit. to the amount of power an ae generator is capable of delivering and to the load which a synchro- nous motor ean cary. Instability results from attempting to increase the mechanical input to a generator or the mechanical lod on a motor beyond this definite amount of power ealled the stability limit. A limit ing value of power is reached even if the change is made gradually, Dis turbances on a system, caused by suddenly applied loads, by the oceur- rence of faults, by the loss of excitation in the field of a generator, and. by switching, may cause loss of synchronism, even if the change in the system caused by the disturbance would not exceed the stability limit if the change were made gradually. The limiting value of power is called the transient stability limit or the steady-state stability limit accord ing to whether the point of instability is reached by a sudden or a gradual change in conditions of the system. Fortunately, engineers have found methods of improving stability and of predicting the limits of stable operation under both steady-state and transient conditions. Stability studies of a two-machine system are less complex than studies of multimachine systems, but many of the methods of improving stability ean be seen by the analysis of a two-machine sys- tem. The a-c calculating board is a great help in predicting the stability limits of a complex system and in comparing various methods of increas ing stability, but the same calculations must be made for each machine represented on the board as are made for the machines in a simple: which is more suited to analytic calculations, 1.6 The Power System Engineer. ‘This chapter has attempted to sketch some of the history of the basic developments of electric power systems and to describe some of the analytic studies which are important in planning the operation, improvement, and expansion of a modem power system. The power system engineer should know the methods ‘of making load studies, fault analyses, and stability studies, for such studies affect the design and operation of the system and the selection of apparatus for its control. Before we can consider these problems in more dotail, we must study some fundamental concepts relating to power systems in order to appreciate how these fundamental concepts affect the larger problems. system REGISTERED ©, q VERSION ADDS NO WATERMARK &, & * < Crint-drigCHAPTER 2 INDUCTANCE OF TRANSMISSION LINES. 2.1 Introduction, An electric transmission line has four parameters which affect its ability to fulfil its function as part of a power system. These parameters are resistance, inductance, capacitance, and conduct ance. In this chapter we shall study inductance, and we shall consider the other parameters in the two following chapters When current flows in an eleetrie circuit we explain come of the proper- ties of the circuit by the magnetic and electric fields which are present. Figure 2.1 shows an open two-wire Tine and its associated magnetic and clectrie fields. The lines of mage netie flux form elosed loops linking the circuit, and the lines of eleetrie ‘tux originate on the positive charges fon one conductor and terminate on the negative charges on the other conduetor. Variation of the eur rent in the conductors causes a change in the number of lines of Fic. 2:1 Magnetic and electric fields magnetic Mux linking the cireuit focinted with a tvo-wir ie Any change in the flux linking a circuit induces voltage in the circuit, and the induced voltage is proportional to the rate of change of flux. Inductance is the property of the circuit that relates the voltage induced by changing flux to the tate of change of eurrent 2.2. Deftsition of Inductance. ‘Two fundamental equatiog explain and define inductance. The fist equation relatogj to the rate of change of flux linking a ¢ _w& eu where ¢ is the induced voltage in volts and ¥ is # of the circuit in weber-tums, The number of we] “4 a = REGISTERED J) VERSION ADDS NO WATERMARKINDUCTANCE OF TRANSMISSION UNES 1s of each weber of flux and the number of tums of the cireuit linked. For the two-wite line of Fig. 2.1 each line of flux links the circuit only once, and 1 volt is induced if the rate of change of flux is 1 weber/see. If we had been considering a coil instead of the circuit of Fig. 2.1, most of the lines of flux produced would have linked more than one turn of the coil If the flux linking 100 turns of a coil changed at the rate of 1 weber/sce, the induced voltage in each turn would be I volt, but the induced voltage in the coil would be 100 volts since the turns are in series, ‘Therefore, the induced voltage is proportional to the rate of change of flux linkages. A coil having five turns is shown in Fig. 2.2. The closed loops represent some of the magnetie flux linking the turns of the coil, Two of these loops are seen to Tink only one turn of the coil. They contribute a total of two flux linkages. Two other loops link three turns and therefore contribute six flux linkages. Four loops link all five turns to give twenty flux linkages. "Thus, for the loops shown, there are 2+ 6 -+ 20 = 28 flux linkages. If each loop or Tine of flux represents 1 weber, the unit, of flux linkages is a weber-turn, and the coil has 28 weber-turns. Decreasing this flux to zero at a uniform rate in 1 see would induce 28 volts in the coil When the current in a cireuit is changing, its associated magnetic field (which is described by the flux linkages) must be changing. If constant permeability is assumed for the medium in which the magnetic field is set up, the number of flux linkages is directly proportional to the current, ‘and therefore the induced voltage is proportional to the rate of change of current, ‘Thus our second fundamental equation is Fic. 2.2. Flus linking coil volts 2.2) a vo Co whore Z, the eonstant of proportionality, is the inductance of the ciggait in henry’, eis the induced voltage in volts, and di/at is the ra of current in amperes per second. Equation (2.2) m [= REGISTERED VERSION ADDS NO WATERMARK &, & honrys 7 < Crint-drig6 ELEMENTS OF POWER SYSTEM ANALYSIS If the flux linkages of the cireuit vary linearly with current, whieh means the magnetic circuit has a constant permeability, i= hens ay from which arises the definition of the self-induetance of an electric circuit as the flux linkages of the circuit per unit of current. The inductance ‘of one conductor of a cireuit is equal to the flux linkages of the conductor per unit current in the conductor. In a two-wire line the number of flux linkages of the circuit is the sum of the flux linkages of each wire. In the rationalized mks system of units, Lin henrys is equal to weber-turns per ampere. In terms of inductance the flux linkages are Y= Li weber-turns 25) In Bq, (2.5), if Fis instantaneous current, ¥ represents instantaneous flux linkages. When the current is alternating, the flux linkages are alternat- ing, and the mms value of the flux linkages is the product of the inductance and the rms current. Thus You = LE weber-turnst 26) ‘The rms voltage drop due to the flux linknges is, V = job! volts (27) V = jeahm —volts @8) ‘Mutual inductance between two cireuits is defined as the flux linkages of one eireuit due to the current in the second circuit per ampere of cur- rent in the second cireuit. If the current 2 produces Ys flux linkages with circuit 1, the mutual inductance is ve i ‘The voltage drop in eireuit 1 eaused by the flux linkages of eireuit 2 is Vi = joMsels Mis henrys ivr volts ‘Mutual inductance is important in considering the influence of power lines on telephone lines and in considering the coupling between parallel power lines. 2.3 Partial Flux Linkages. Only flux lines external to the eondy have been shown in Fig. 2.1, Some of the magnetic field ‘the conductors although the amount of internal flux thu ican bo wsetel thigh equa R QO y capital ters fSREGISTERED 9) VERSION ADDS NO WATERMARK &, & 1 We shall designate instantaneous eurrents and voltages ‘rns values by eapital letters. Masimum values will by with the subserint mor max. ‘Thus the instantancous i the maximum flux linkages are Lf, and the rms value * < Crint-drigINDUCTANCE OF TRANSMISSION UNES v ‘The changing lines of flux inside the conductors also contribute to the induced voltage of the circuit and, therefore, to the inductance. ‘The correct: value of inductance due to internal flux may be computed as the ratio of flux linkages to current by taking into account the fact that each line of internal ffux links only a fraction of the total current. ‘The flux linkages of the internal flux in @ tubular element are the product of flux in the clement and the ratio of the current encireled by the tubular ele- ment to the total current in the conductor. ‘Thus a line of flux which encircles only half the current in & conductor contributes only half a flux linkage. Partial flax linkages are those linkages produeed by flux which links only part of the current. ‘The total number of flux linkages d internal fx is the summation ofall the partial linkages, ‘The summation of all the partial flux linkages in weber-turns divided by the current in the cireuit in amperes is the inductance in henrys due to internal flux. ‘The principle outlined above for computing inductance is applicable to inductance resulting from external as well as internal flux. By this principle inductanee is defined as flux linkages per ampere, and the value of flux linkages is the summation of fiux times the fraction of the total current linked. ‘The fraction is less than one for lines of flux inside the conductor and greater than one for flux surrounding several tums of a coil. For Fig. 2.2 the fractions of current linked by the lines of flux shown are 1, 8, and 5, ‘The method of computing flux linkages by multi- plying each line of flux by the fraction of current enclosed should become increasingly clear as the topic of inductance is developed further. ‘We shall show later that the method of partial flux linkages is valid for computing the internal inductance of a cylindrical wire by deriving internal induetance in another manner and comparing the results of the twvo methods. 2.4 Inductance of a Conductor Due to Internal Flux. In order to obtain an accurate value for the inductance of a transmission line, it is necessary to consider the flux inside each conductor as well as the external flux. Let us consider the long, eylindrieal conductor whose cross section isshown in Fig. 2.3, We will assume that the return path for the eurrent in this conductor is so far away that it does not appreciably affect the magnetie field of the conduetor shown. ‘Then the lines of flux are con- centric with the conductor. ‘The magnetomotive foree (mmf) in ampere-turns around vg ie AN aint eae ee seca or 02 Ceased ey fold intenety around the path" Ths ers La 2 VERSION ADDS NO WATERMARK &, 4th ed, pp. 428-432, John Wiley & Sone, Ine., New Yop oO 7 * < Crint-drig® SENENTS OF POWER SYSTEM ANALYSIS where H is the magnetic field intensity in ampere-turns per meter, s is the distance along the path in moters, I is the current. in amperes ‘enclosed, and the dot between Hand de indicates that the value of H is the component of the Geld intensity tangent to de. Let the field intensity at a d x moters from the center of the con ductor be designated H,. Since the field is symmetrical, M7, is constant at all. points equidistant from the center of the conductor. If the inte» gration indicated in Eq. (2.9) is per- formed around a cixeular path eon- contrie with the conductor at x met from the center, H, is constant over the path and tangent to it. Equa- tion (2.9) becomes tance $iids = 1, (2.10) and 2eeH, = Te @n) where Z, is the current enclosed density, hen, assuming uniform current ety (2.12) a=5 1 in the total cure in the conductor. ‘Then substituting Ba, 13) ing. 211, we obtain Dest = BT ea) and Hy = 5237 amp-turns/meter a4 ‘The flux density x meters from the center of the conductor is axl zi ate SEL, webers/meter* (215) where x is the permeability of the conductor. B, = ull, = If the curvont is alternating, the maximum value of His ahve of the current is used in Iq. (2.8). Similarly, i 7 is rms Seld intensity, and flux computed from rms 27 is the rms ¥ is applicable to dirset current or instantancons, maximum, § In the rationalized mks eystem of units the pormeat 4e X 107 henry met and the relative permeability 8 m/s FS-REGIS TERED © q VERSION ADDS NO WATERMARKINDUCTANCE OF TRANSMISSION UNES 1 In the tubular element of thickness da, the flux d@ is B, times the cross- sectional area of the element normal to the flux lines, the area being dz times the axial length, ‘The ffux per meter of length is webers/meter of length (2.16) ‘The flux linkages dy per meter of length, which are caused by the flux in the tubular element, are the product of the flux per meter of length and the fraction of the current linked. ‘Thus ula a a a z SIE de weber-turns/meter 7 Integrating from the center of the conductor to its outside edge to find Yous the total flux linkages inside the conductor, we obtain (2.18) ot a relative permeability of 1, « = de X 10-7 henry/meter, and I Vin = 5X 10-7 weber-turns/meter (2.19) Lia = 14 X 10-7 henry/meter (2.20) 2.5 Flux Linkages between Two Points Exter- nal to an Isolated Conductor. As a stop in computing inductance due to flux external to a conductor, lot us derive an expression for the flu Iinkages of an isolated conduetor due only to that, portion of the external fhix which lies between two points distant D, and Dz meters from the center of the conduetor. In Fig. 2.4, Py and Ps are two points of distances Di and Ds from the conductor which carries a current of J amp. Since the flux paths are concentric circles around the conductor, all the flux, between P and Ps lies within the concentric cylindrical surfaces which pass throug) Py and P:, At the tubular element which is 2 meters from the center of the conductor field intensity is H,. ‘The mmf around the RED A Pion ci >, [ss: REGISTERED 9) VERSION ADDS NO WATERMARK QerH, = I2» NENTS OF POWER SYSTEM ANALYSIS The field intensity is H, = 5 amp-turns/meter 222) and the flux density in the clement is webers/moter? (2.23) ‘The flux d¢ in the tubular element of thickness de is, al dg = fh de webers/meter of length (224) ‘The flux linkages dy per meter are numerically equal to the flux dé, since flux external to the conductor links all the current in the conductor ‘once and only once. The total flux linkages between P: and Ps are obtained by integrating dy from 2 = Di tor = Ds. We obtain gon [pede = Bin Bt ebmrturm/mter—¢ or, for a relative permeability of 1 bu= 2x i072 webertums/neter — 226 Tes inductance de only to the Suid tetveon Ps and ai m= 2x10 PE heneyymeer In Eqs. (2.25) to (2.27), note that “In” denotes the natural logarithm (base ©. Converting henrys per meter to millihenrys per mile and using the logarithm to the base 10, we obtain 7at og Iu = nillibenrys/mile (2.28) 2.6 Inductance of a Single-phase Two-wire Line. Before proceeding to the more general case of multiconduetor lines and three-phase lingg, let us consider a simple two-wire line composed of solid, round, Figure 2.5 shows a circuit having two conductors of rad conductor is the return circuit for the other, First flux linkages of the cireuit caused by the current j of flux set up by curtent in conductor 1 at a distal Mine VU ‘%5 fo or greater REGISTERED ©) VERSION ADDS NO WATERMARK "Throughout this book “In"” denotes the natural i enotes the common logarithm (base 10).INDUCTANCE OF TRANSMISSION LINES a than D + rz from the center of conductor 1 does not link the cireuit and cannot induce a voltage in the circuit. Stated in another manner, such a line of flux links a net eurrent of zero, since the current in conductor 2 is equal in value and opposite in direction to the eurrent in conductor 1 The fraction of the total current linked by a line of flux external to conductor 1 at a distance equal to or less than D ~ r2is one. Between D ~ reand D + rs (that is, over the surface of conductor 2), the fraction of the total current in the cirenit linked by a line of flux set up by current in conductor I varies from one to zero. ‘Therefore, it is logical to simplify Fro, 2.5 Conductors of diferent radii and the magnetic eld due to current in oon: Aluotor | only. tho problem, when D is much greater than r; and y; and the flux density through the conductor is nearly uniform, by assuming that all the external flux set up by current in conductor | extending to the center of conductor 2 links all the current 7 and that flux beyond the center of conductor links none of the current. In fact, it ean be shown that calculations made on this assumption are correct even when D is small.* ‘The induetanee of the cireuit due to current in conductor 1 is d mined by Eq. (2.27) with the distance D between conductogs 1 aj substituted for Ds and the radius 71 of conductor 1 substitu| For external flux only Ine = 21072 henge See E, W. Kimbark, “Electrical Ty 7, John Wiley & Sons, Tne., Ni or 180 f-&-REGISTERED VERSION ADDS NOn ELEMENTS. OF POWER SYSTEM, ANALYSIS. For internal flux only Liga = 36 10 henty/meter 230) ‘The total inductance of the ircut due to the current in conductor 1 only (@ 42h 2) X 107 — henrys/meter 31) ‘The expression for inductance may be put in a more concise form by factoring Eq, (2.31) and by noting that In & = 14, whenee b= 2x109({ +m?) (232) t= 2x10 (ne +n?) (238) Upon combining terms, we obtain = 2x 107 In, (231) If we substitute r for ne¥ 2x107In2 fees (235) x 1y=2% 107 InP henrys/mete 2.35 Ly = 0.7411 log?) millihenrys/mite 236) ‘The radius r{ is that of a fictitious conductor assumed to have no internal flux but with the same inductance as the actual conductor of radius r ‘The quantity &! is equal to 0. Equation (2.35) gives the same value for inductance as Eq. (2.31). The difference is that Eq. (2.35) omits the term to account for internal flux but compensates for it by using an adjusted value for the radius of the conductor, We shoilld note carefully that Eq. (2.31) was derived for a solid, round conductor and that Eq. (2.35 found by algebraic manipulation of Eq. (2:31). ‘Therefore, the multiplying factor of 0.7788 to adjust the radius in order to account for internal flux applies only to solid, round conductors. We shall consider other conduetors later. Since the current in conductor 2 flows in the direction opposi that in conductor 1 (or is 180° out of phase with it), the fl produced by current in conductor 2 considered alone direction through the circuit as those produced by eurre] ‘The resulting flux for the two conductors is deter the mmfs of both conductors. For constant permg flux linkages (and likewise the inductances) of sidered separately may be added. to fe of owever, the Os SS S REGISTERED ©, q VERSION ADDS NO WATERMARKINOUCTANCE OF TRANSMISSION UNES n By comparison with Eq. (2.35) the inductance due to current in con- ductor 2 is = 2x 107 In? henrys/meter (37) and for the complete circuit, L=ltl:=4xX 107 InP henrys/meter (2.38) vit If, = = rf, the total inductance reduces to 1x 10-7In? — henrys/meter (2.39) r= 1482 ig? matihengnile a) Equation (2.40) is the inductance of the two-wire line taking into account the flux linkages caused by current in both conductors, one of which is P Fin. 266 Cross-sectional view of a group of n conductors carrying currents whose sum is rero, Point P is remote from the conductors, the return path for eurrent in the other. This value of inductance is sometimes called the inductance per loop meter or per loop mile to distinguish it from the inductance of the cireuit due to the current in one conductor only. The latter, as given by Eq. (2.36), is one half the total inductance of a single-phase line and is called the inductance per conductor 27. Flux Linkages of One Conductor in a Group. A more general problem than that of the two-wire line is presented by one condyetor, group of conductors where the sum of the currents in all the is zero. Such a group of conductors is shown in Fig. 2 1,2,3,... a carry the currents Ls, I, I, of these conductors from a remote point P are indicl Di», Dep, Daoy Dar. Let us determine ¥y conductor 1 due to fy including internal fhux li the flux beyond the point: P, By Eqs. (2.19) an = REGISTERED © q VERSION ADDS NO WATERMARK“ ELEWENTS OF POWER SYSTEM ANAS (+ 2nin 2) 10 eat) Dir Vier = vars = 2X 10-77; In weber-tums/meter (2.42) The Aux linkages Yare with conductor 1 due to Ty, but excluding flux beyond point P, is equal to the flux produced by Iz between the point P and eonduetor 1 (that is, within the limiting distances Dep and D from conduetor 2). So Dp Dis Yara = 2X 10-72 In 2.43) ‘The flux linkages yar with conduetor 1 due to all the conductors in the group, but excluding flux beyond point P, is pe ewan (m Beg PE 4 net + ink) eas which becomes, by expanding the logarithmic terms and regrouping, 1 ‘+n ° 1 1 1 dar = 2X 10 (n ind + hn + ln p+ Di $Jyln Dip + Lain Dap + Tain Dap + > + Tal a) (248) Since the sum of all the currents in the group is zero, hththts+h=0 and, solving for I,, we obtain Ine (ht et Its + In) (2.46) Substituting Bq. 2.46) in the second term containing 7, in Eq. (2.45) and recombining some logarithmic terms, we have 2x 10 + hin vr (mind J+ Tain De Ds Dyp Bot hingt +: Dar Now letting the point P move infinitely far % S\REGISTERED © terms containing logarithms of ratios of distagg pS, REGIST Zl SION ADDS NO WATERMARK + hin BE + DalINDUCTANCE OF TRANSMISSION LINES 2s tesimal, since the ratios of the distances approach one, we obtain Varin H + isin u By letting point P move infinitely far away we have included all the flux linkages of conductor 1 in our derivation. Therefore, Bq. (2.48) expresses all the flux linkages of conductor 1 in a group of conductors, provided the sum of all the currents is zero, If the currents are alternat- ing, they must be expressed as instantancous currents to obtain instan- tancous flux linkages or as complex rms values to obtain the rms value of flus linkages as a complex number 2.8 Inductance of Composite-conductor Lines. ‘The commonest eon ductors for overhend power transmission lines are composed of strands of wire with alternnte layers spiraled in opposite directions. Spiraling alternate layers in opposite directions prevents unwinding and makes the outer xdius of one Inyer coincide with the inner radius of the next. Stranding provides flexibility with large cross-sectional area, ‘The num- ber of strands depends on the number of layers and on whether all the strands are the same diameter. The total number of strands in con- centrically stranded cables, where the total annular space is filled with strands of uniform diameter, is 7, 19, 37, 61, 91, or more. A general formula for the total number of strands in such cables is Number of strands = 327 — 32+ 1 where 2 is the number of layers, including the single conter strand, A 500,000-cireular-mil conductor may be composed of 37 strands having individual diameters of 0.1162 in. or of 19 strands having individual diameters of 0.1622 in, Table A.1 in the Appendix lists the character- isties of concentrically stranded conductors of hard-drawn copper and of conductors having 12, 8, and single strands. The strands of copper conductors are usually uniform in diameter and composed of copper only. Figure 2.7 shows a typical steel-reinforeed aluminum cable (ACSR). he conductor shown has 19 steel strands forming a central 0 which are two layers of aluminum strands. There arg 30. strands in the two outer layers. ‘The conductor stranf as 30AI/198t, or simply 30/19. Various tensil capacities, and eonductor sizes are obtained by usinl tions of steel and aluminum, Table A.2 in theadag the characteristics of ACSR, indicates the sizes and steel strands, the number of strands of ex the number of VERSION ADDS NO WATERMARK &, & * < Crint-drig1» FLEMENTS OF POWER SYSTEM. ANALYSIS layers for the usual types of strnading. A type of conductor known as ‘expanded”” ACSR has a filler such as paper separating the inner steel strands from the outer aluminum strands ‘Steel wites coated with a thick layer of copper are used to obtain high tensile strength combined with good current-carrying capacity. Some- times cables are composed of copper strands in the outer layers and copper-coated steel wires in the inner layers, Hollow copper conductors are sometimes used on high-voltage lines. One type of hollow copper conductor consists of interlocked sections of copper forming a spiral 4 aluminwm conductor, 19 stocl strands, 30 aluswinum strands. (Aluminum Company of America) a along the axial length of the conductor. Such a conductor is self- supporting and has some degree of flexibility. Another type of hollow conductor consists of copper strands with the inner layer twisted in the direction opposite to that of a twisted copper I beam around which the strands are spiraled Stranded conductors come under the general classification conductors, which means eonductors composed of two or strands electrically in parallel. We are now induetance of a transmission line composed of but we shall limit ourselves to the ease where all thd and share the current equally. ‘Phe method cal all types of conductors containing strands of RE Ot sds VERSION ADDS NOINDUCTANCE OF TANSMISION UNS 7 duetivities,? but this will not be done here since values of internal induc- tance of spevfie conductors are generally available from the various mnatinfacturers and ean be found in handbooks, The method to be devel- Gped indicates the approach to the more complicated problems of non- Homogeneous conductors and unequal division af current between strands ‘The method is applicable to the determination of inductance of Hines consisting of eireuits electrically in parallel since two conductors in parallel ean be treated as strands of a single composite conductor. Figure 2.8 shows a single-phase line composed of two conductors. Tn onder to be more general, exch conductor forming one side of the line ment of an indefinite number of eon- 6 ductors. ‘The only restrictions are ° vee that the parallel filaments are cylin: ° rival and share the current equally. °O Conductor X is composed of n iden- teal, parallel laments, each of which Sao ae carries the current 1/n. Conductor pyg. 9" Singl-phase line contting of Y, which isthe return citeuit for the fe composite current in conductor X, is composed of m identical, parallel filaments, each of which carries the current — J/m. Distances between the elements will be designated by the letter D with appropriate subscripts, Applying Eq, (2.48) to filament a of conductor Br we obtain for Hux linkages of lament a 2 aF(4 Vomiy... 1 vcwetlin Lam tind L) @ max wel (ing ting bingy ts +ngc) eam from whieh = 2x 10-71 In VP Don wD “DasDac > Dan weber-turns/meter Dividing Eq. (2.50) by the current I/n, we find that the inductance of filament a is ¥ Tn 1x 1° y Deez De Ne DasDac* * * Dan + Seo for instanoe L. F. Woodruff, “Bleetrie Power Wiley & Sons, Inc., New York, 1988. VERSION ADDS NO WATERMARK &, & 7 < Crint-drigELEMENTS OF POWER SYSTEM ANALYSIS. Similarly, the inductance of filament b is Tin 2n x 107 nS henrys/meter (2.52) ~ Dow ‘The average inductance of the filaments of conductor X is Lat bn + bat t+ In La = (2.53) Conductor X is composed of n filaments electrically in parallel. If all the filaments had the same inductance, the inductance of the conductor would be 1/n times the inductance of one filament. Here all the fila ments have different inductances, but the inductance of all of them in parallel is 1/n times the average inductance, ‘Thus the induetan conductor X is Ix = Tay _ Lg ¢ Ly tbe tt tly 254) Substituting the logarithmic expression for inductance of each filament in Eq. (2.54) and combining terms, we obtain V0uDwDu }(DraDusDre (DasDasDar => Dan} howrys/moter (2.55) where rf, andr, have been replaced by Dus, Du, and Dyn, respectively, to make the expression appear more symmetrical. Note that tho numerator of the argument of the logarithm in Eq. (2.55) is the mnth root of mn terms, which are the produets of the dis- tances from all the n filaments of conductor X to all the m filaments of conductor Y. For eaeh filament in conductor X there are m distances to filaments in conductor Y, and there are n filaments in conductor X. ‘The product of m distances for each of n filaments results in mn terms. ‘Lhe mnth root of the product of the mn distances is called the geometric mean distance between conductor X and conductor Y Ds or GMD and is also called the mutual GMD betwg ductors. Geometric mean distance is a mathematical will discuss later in more general terms, ‘The denominator of the argument of the logat the n® root of n® terms, There are n filamenj there are n terms consisting of +’ for that fila from that filament to every other filament in itor X. ‘Thus we, VERSION ADDS NO WATERMARK & o Ori ‘int-dSsINDUCTANCE OF TRANSINISION LINES, ” account for n? terms, Sometimes rf is called the distance from filament a to itself, especially when it is designated as D,,. With this in mind the terms under the radical in the denominator may be described as the product of the distances from every filament in the conductor to itself and to every other filament. ‘The n® root of these terms is called the self GMD of conductor X, and the r’ of a separate filament is called the self GMD of the filament, for reasons which we shall see later when we discuss GMD as a mathematical concept. Sometimes self GMD is called geometric mean radius or GMR. Self GMD may be abbreviated Lx 0.7411 log 22 nillihenrys/mile (2.57) If we compare Eq, (2.57) with Eq. (2.36), the similarity. between them is apparent. ‘The equation for the inductance of one conductor of 8 composite-conductor line is obtained by substituting in Ra. (236) the GMD between conductors of the composite-eonductor line for the distance between the solid conductors of the single-conductor line and by substituting the self GMD of the composite conductor for the self GMD. (©) of the single conductor. Equation (2.57) gives the inductance of one conductor of a single-phase line. The conductor is composed of all the strands which are electrically in parallel. ‘The inductance is the total number of flux linkages of the composite conductor per unit of line current, Equation (2.36) gives the inductance of one conductor of a siugle-phase line for the special case where the couductor isa solid, round ‘The inductance of conduetor ¥ is determined in a similar manner, and the inductance of the line is L=Lr+Ly 29 Geometric Mean Distance. In the preceding section we derived an expression for the induetanee of a composite-conductor line. We found in the expression for inductance due to the current in one fa term which is the geometric mean of the distances betyreey of the same conductor only. Geometric mean distan concept which is helpful in calculating inductancys By definition the GMD from one point to a (A the geometric mean of the distances from the on to each of the VERSION ADDS NO WATERMARK &, & fe < Crint-drig» ELEMENTS OF FOWER SYSTEM ANALYSS other points. For instance, the GMD from an extemal point to four points on a ciele is the geometric mean of the four distances shown in Fig. 2.9. Here the geometric mean of the distances is IMD = WD:DaDaDi If the number of points on the circle is increased without limit, the geo- metrie mean of the distances from the external point to the points on the circle approaches the GMD from the point to the circle, It is equal to the distance from the point to the center of the cirele.® The GMD from any point on a cirele to all other points ‘on a circle is equal to the radius of the circle. Piitiofouryenisunaemie "| The concept of the GMD from a point to an area is important and can be visualized by dividing the area into large mumber of equal elements and taking the geometric mean of the distances from the point to the clements of area. If there are n elements, the geometric mean of the distances is the nth root of the product of the n distances. ‘The GMD from the point to the area is the limit approached by the GMD from the point to the elements of the area as the number of elements increases without limit. ‘To find the GMD between two areas, each area is divided into a num- ber of equal elements, say m equal elements for one area and n equal elements for the other. ‘The GMD between the areas is the limit of the ‘mnth root of the mn produets of the distances between the m elements, of one area and the n elements of the other area as m and n increase without limit. Figure 2.10 shows the six distances between two of the ‘m equal elements into which one area is divided and three of the n equal elements into which the other area is divided. ‘To find the GMD between the areas all distances between elements must he considered, and the number of elements in each area must be infinite. The GMD between two circular areas can be shown to be equal to the distance between their centers. ‘The self GMD of an area is the limit of the geometric 1 distances between all the pairs of elements in that au *'Sce B. B. Ross and F, W. Grover, “Formulas and Tal Mutual and Self Inductance,” Seientitie Paper 169, Bull. Bo ppp. 1-287, 1912. Other formulas for GMD have heen See alao J. C. Maxwell, “A Treatise on Electricity 208-301, Clarendon Press, Oxford, 1881. FS"REGISTERED ©) VERSION ADDS NO WATERMARK &, & * < Crint-drigINDUCTANCE OF TRANSMISSION UNES 3 of elements inereases without limit. The self GMD of a circular area ean be shown to be equal to the radius of the circle times ¥, Since 1” in our formulas for the inductance of @ round wire is the radius of the wire times &, we ean sce the reason for calling r! the self GMD of the Since the cross-sectional areas of the filaments of the composite con- ductors considered in deriving Eq. (2.55) could be clements of areas such as those in Fig. 2.10, the induetance of a line composed of conductors of irregular area ean be found by calculating GMD values. The self GMD of each area and the mutual GMD between the two areas must be found. ‘The inductance due to current. in each conduetor is found by Fra, 2:10 ‘The ax distances From twa equal elements of one area to three equal ele monte of another ares Eq. (2.56) or (2.57), and the two inductances are added to find the inductance of the line. Uniform current, density throughout is assumed. ‘Table 2.1 gives some formulas for self and mutual GMD. The GMD method does not apply strietly to nonhomogeneous con- duetors such as ACSR or to eases where the current density is not uniform throughout the conductor, An approsimate value for inductance of ACSR is obtained by neglecting entirely the current carried by the steel conductors. ‘The eurrent in the steel conductors is relatively small, and the inductance depends on the amount of current in the conduetor, since the permeability is not constant and the flux linkages are not a linear function of current. If inductance is determined experimentally for ACSR or other conductors not having uniform current density, an equivalent self GMD may be found? Let D. be the eq GMD which when substituted in the inductance formpla y value of the experimentally determined induetance +p Dm h L=2x 107 Ine henrys * S0e I. F. Woodruf, “Bleetrio Power Transmission,” Sons, Tuc, New York, 1908, VERSION ADDS NO WATERMARK &, & * < Crint-ari2 ELEMENTS OF FOWER SYSTEM ANALYSIS ia 21 Grose Maas Distasces Deeeription | —_ilustration Self GMD of a eireular | GMD from one eireular | area to another GMD from cireular line | to enclosed area GMD from external | point to circular area | 1D between n equally | spaced points on circle GMD froin one annular ‘area to another Saf GMD of « reetang- ular are F=e1 | 0.2987, depending on ratio a/b Self GMD of an annular | ‘the equivalent self GMD is D, = Dyce (2.59) where L is in henrys per meter. ‘The value of Du is not affected by nonuniform distribution of the current inside the conductor as long the, external magnetic field is not changed. Table A.2 in the A the self GMD (GMR) of various sizes of ACSR. 2.10 Examples of Inductance Compute ples of the method of calculating inductanc Example 2.1 One cireuit of a single-phase transmission solid wires, each 0.1 in, in radius, ‘The returgag 's composed of ta RSION ADDS NOINOUCTANCE OF TRANSMISSION UNES 2 wires, each 0.2 in, in radius, ‘The arrange- ment of conductors is shown in Fig. 2.11 Find the inductance due to the current in each side of the Tine and the inductanee of the complete line in millihenrys per mile Solution 8 a 3 Find the GMD between sides X and Y. . WDaDacDosDueDeaDee Dat = Dre = 30 ft Due = Dus = Dar = VIO FBO Das = AO FA = 50 ft Sax Sav Du = SOF X 50% 1,800" Fic, 2.11 Arrangement of 30% X 50! X 1,300" = ‘Phen find the self GMD for side X. 5.8 ft conductors for Example 2.1. D, = /DesDisDuPrxDDiDsD Du = (CLES) a0 0! [oi x0, V2 X 208 x (2 x 20) 5 358 a Lx = 07411 log 7a = 1.00 millihenry/mile 35.8 0.7411 log 35; = 1.38 millihenrys/mile L Lx + Ly = 2.38 millihenrys/mile, Example 2.2 A conductor is composed of seven identical copper g a radius r as shown in Fig. 2.12. Find the factor multiplied to find the self GMD of the conductor, by which the square root of the area of the col should be multiplied to obtain the self GMD of t * REGISTERED ON VERSION ADDS NO WATERMARKPy ELEMENTS OF POWER SYSTEM ANALYSIS. Solution First we find the distances Dis, Dass, and Dus, as follows: Dy 2% Diy = 4r Dy = VD = Dee VP = rf = 2 V3 en-strand ‘The self GMD of the se conductor ia the 49th root of 49 distances. So D, = YEV Dae De DDO < NS self GMD of one strand and the seli GMD values of every other strand. ‘The term Du:tDustDuD a ah A is the product of the distances from Se N one outside strand to every other Std Ie ie raised to the sth Miranda "The tenn ("accounts Guise stand, ‘Thon here are for D, we ebuain D= VE XV xe PKB K ‘Po find D, in terms of total conductor area in circular mils, let A = total conductor area in circular mils diameter of each strand in mils radius of each strand in mils A= Td? = 284 Qa og 7 2am VA = Oats VA ls Ifa single-phase line consists of two stranded enbles computed in Example 2.2, it is seldom ne‘ between strands of the two sides, for the C to the distance between centers of the eables GMD is important only where the various! electrically in parallel are separated from eae Tt Dy, See equal AN ‘REGISTERED ©) * (REGIST “VERSION ADDS NO WATERMARKINDUCTANCE OF TRANSAISSION UNES as nearly approaching the distance between the two sides of the circuit For iustance, in Example 2.1 the conductors in parallel on one side of the line are separated by 20 ft, and the distance between the two sides of the line is 30 ft. Here the ealeulation of mutual GMD is important. For stranded conductors such as that of Example 2.2, the distance Detween sides of the line is usually so great that the mutual GMD can be taken as equal to the center-to-center distance with negligible error. 2.11 The Use of Tables. Tho self GMD of conduetors of any number of strands ean be computed as in Example 2.2. ‘The engineer seldom has to make such computations, however, since tables listing values of self GMD are generally available for standard conductors. All manu- factuters furnish data, including values of self GMD, for their conductors, and tables provide the most practical method of obtaining the desired values, especially for nonhomogeneous conductors such as ACSR, In order to use the tables intelligently the engineer must understand thor- oughly the meaning of the tabulated data. Inductive reactance rather than induc ice is usually desired. ‘The ‘a single-phase two-conduetor inductive reactance of one conductor of line is Qafl. = Bef X OTALL X 10-4 Log PE = 1057 1077 log PE mame (250 where Dy is the distance between the two conductors. The self, GMD D, may be found in the tables and substituted in the equation. Dy and D, must be in the same units, Most tables list values of D, for 60 ep: 25 eps, and direct current. The value of equivalent D, varies with frequeney beeause current density does not remain uniform throughout the conductor as frequency inereases, as we shall see in Chap. 4. As current distribution becomes nonuniform the inductance due to internal flux decreases, and different degrees of nonuniformity are accounted for by different values of D, at different frequencies. ‘The nonuniform distribution of eurrent due to the frequeney of the current is called skin efieet. Tn the equations and formulas already discussed in this chapter we neglected skin effect by assuming uniform current density. to be applied to the value of internal inductance comput tion of uniform distribution of current Some tables give values of inductive reactance in One method is to expand the logarithmic term of ¥ k as follows: AN [Sn REGISTERED 9 VERSION ADDS NO WATERMARK &, & Xi = 4.657 x 10-flog t+ 4.657 X 10-4flog D, D. * < Crint-driga6 ELEMENTS OF FOWER SYSTE® ANALYSIS If both D, and Dp are in feet, the first term in Bq, (2.61) is the inductive reactance of one conductor of a twe-conduetor line having a distance of 1 ft between conductors, as may be seen by comparing Eq. (2.61) with Eq, (2.57). Therefore, the first term of Eq. (2.61) is called the inductive reactance at L-jt spacing. Tt: depends upon the self GMD of the eonductor and the frequeney. ‘The second term of Eq. (2.61) is called the inductive reactance spacing factor. This second term is independent of the type of conductor and depends on frequency and spacing only. ‘The spacing factor is equal to zero when Dw is 1ft. If Dy is less than 1 ft, the spacing, factor is negative. ‘The procedure for computing inductive reactance is to look up the inductive reactance at 1-ft spacing for the conductor under consideration and to add to this value the inductive reactance spacing factor, both at the desired line frequency.” In the Appendix, ‘Tables A.1 and A.2 include values of inductive reactanee at 1-ft spacing, and Table A.3 lists values of the inductive reactance spacing factor. Exomple 2.3 Find the inductive reactance per mile of a two-conductor single-phase line operating at 60 eps. The conductors are each No. 1/0 seven-strand hard-drawn copper wire spaced 18 ft between centers. Solution ‘The area of the stranded conductor is ‘Table A.1). From Example 2.2 D, = 04114 WA in, VANE VIDED 5 9.0 ~ oo1113 which is the value listed in Table A.1 for D, at 60 eps. Agreement of calculated and tabulated values indicates that skin effect is negligiblo for this case. For one conductor Xp = 4.657 X 10-7 X 60 log 5 Gt yg = 0.807 ohm/mile dour If only D, is given in the tables, the above method is used, tive method follows: ERED on Lewis “es eae oY “REGISTERED 9) VERSION ADDS NO WATERMARK &, & w-This method of computing induetive rexetance was ‘and appeared in C.F. Wagner and R. D. Byans, ‘Sy MoGeaw-Hill Book Company, Inc., New York, 193% “Standard Handbook for Eleetsieal Engineers,” pp. Company, Ine., Now York, 1041 Op; $ Crint-drigINDUCTANCE OF TRANSHISSION LUNES v i eaboracy > Indutivereactanoe al Lftapacing = 4.057 X 10-* X 60g gy! ag = 0.546 ohm/mile Inductive reactance spacing Factor = 4657 X10-* ¥ 00 log 18 351 chm mile Induotive reactance of one conductor = 0.516 + 0381 897 ohm/mile ‘The latter method is preferred if tables are available giving inductive reactance at I-ft spacing and the inductive reactance spacing factor, for thon it is necessary only to add these two values found in the tables. Since the conductors composing the two sides of the line are identical, the inductive reactance of the line is Xi = 2X 0.897 = 1,794 ohms/mile 2.12 Inductance of Three-phase i Lines with Equilateral Spacing. ‘o far in our discussion we have con- sidered only single-phase lines. ‘The equations we have developed are quite easily adapted, however, to the J nN calculation of the inductance of \ three-phase lines, Figure2.13shows the conductors of a three-phase line spaced at the corners of an equilateral < triangle. If we assume that there 2 is no neutral wire, or if we assume Fro, 213 Crosracctional view of the balanced three-phase currents, Saullsteraly, spaced conductors of » eee ee eee acon gag) meehase ine determines the flux linkages of conductor a, So tea 2xtoe(nnd + nnd + tind) weber-turns/meter (2.62) Since [= (Un + 1.), Hq. (2.62) becomes 210-2 tenes) y ror? sithaf REGISTERED 9} VERSION ADDS NO WATERMARK or8 ELEMENTS OF POWER SYSTEM ANALYSIS is tho samo in form as Eq. (2.36) for a single-phase line. conductors, D, replaces r‘ in the equation. Because of symmetry, the inductances of conductors 6 and ¢ are the same as the inductance of conductor a. Since each phase consists of only one con- ductor, Eqs. (2.64) and (2.65) give the inductance per phase of the thre phase line. 2.13 Inductance of Three-phase Lines with Unsymmetrical Spacing. When the conductors of a three-phase line are not spaced equitaterall the problem of finding the inductance becomes more difficult, ‘Then the flux linkages and inductance of each phase are not the same. A different inductanee in cach phase results in an unbalanced circuit and in induced voltages in adjacent communication lines even when the phase currents are balanced. These undesirable characteristies ean be overcome by exchanging the positions of the conductors at regular intervals along the ond. Cond. cond. Pos. Fic. 2.14 Transposition eyale line so that each conductor occupies the original position of every other eonduetor over an equal distance, Such an exchange of conductor posi tions is called transposition. A complete transposition cycle is shown in Fig. 2.14. The phase conductors are designated a, 6, and c, and the positions occupied are numbered 1, 2, and 3. ‘Transposition results in each conductor having the same average inductance over the whole cycle. If an untransposed telephone line paraliels an untransposed power line, the flux produced by the power line induces a voltage of power-line frequency in the telephone line. ‘Transposition of the power line without transposition of the telephone line eliminates interference of the power line with the telephone line except for unbalanced cases where power currents flow in the earth or in overhead ground wires. For balanced three-phase currents in a transposed power line, the magnetic field linking an adjacent telephone line is shifted 120° in time phay rotation of the conductor positions in the transposition gycld length of one transposition cycle of the power line, induced in the telephone line is zero, because it is th voltages equal in magnitude and displaced 120° fi hot necessary to transpose a power line to pry telephone line, for the same result is accompl the telephone line. ference with & [©:REGISTERED 9) VERSION ADDS NO WATERMARK &, & * < Crint-drigINDUCTANCE OF TRANSMISSION LUNES ” Modern power lines are usually not transposed at regular intervals, although an interchange in the positions of the conductors may be made at switching stations in order to balance more closely the inductance of the phases." Fortunately, the dissymmetry between the phases of an untransposed line is small’ and may be neglected in the solution of many problems. If the dissymmetry is neglected, the inductance of the untransposed line is calculated as though the line were correctly trans- posed. ‘The inductive reactance of each phase of the untransposed line is taken as equal to the average value of the inductive reactance of one phase of the same line correctly transposed. ‘The derivations to follow are for transposed lines, ‘The error is small, and the ealeulations are less laborious if the inductance of an untransposed line is ealeulated by the same equations, To find the average inductance of one conductor of a transposed line, the flux linkages of a conductor are found for each position it oveupies in the transposition cycle, and the average flux linkages are determined Let us apply Eq, (2.48) to conductor @ of Fig. 2.14 to find the flux linkages of a in position 1, when b is in position 2 and c is in position 3, as follows It 5 weber-turns/meter (2.66) With a in position 2, b in position 3, and ¢ in position 1, weber-tums/meter and, with a in position 8, D in position 1, and e in position 2, weber-turns/meter (2.68) ‘The average value of the flux linkages of a is ; RED al “us Mano High ot SB “REGISTERED 9, q VERSION ADDS NO WATERMARK rad Lines and Elimination of Electrostatic ATER, vol, 70, pp. 1837-1841, 1951” ELEMENTS OF FOWER SYSTEM ANALYSIS With the restriction that J, = — (Ie + I), 20" (510 tangy " Dad’ Diy X 10°72, In weber-turns/meter (2.70) and the average inductance per phase is x 10+ In Bet henrys/meter Le TALL log 23? millihenrys/mile 71) where Du = WDaDaDar (2.72) ‘Faquation (2.71) may be written Dag = 0.7111 log 4" miliihenrys/mile 3) where D, is the self GMD of the conductor. D,,, the geometric mean of the three distances of the unsymmetrical line, is the equivalent equilateral spacing, as may be seen by the comparison of Eq, (2.71) with Bq, (2.65) We should note the similarity between all the equations for inductance of a conductor, If the inductance is in millihenrys per mile, the factor 0.7411 appears in all the equations, and the denominator of the logarith- mic term is always the self GMD of the conductor. ‘The numorator is the distance between wires of a two-wire line, the mutual GMD between sides of a composite-conductor single-phase line, the distance between conductors of an equilaterally spaced line, or AN, emilet ogatend pang of an unsymmetrical line, os“ ample 24 Fra, 2.15 Arrangement of eee eee a A single-citeuit three-phase line operated at 60 eps is arranged as shown in Fig. 2.15. Bach conductor is No. 2 single-strand hard-drawn copper wire. Find the inductance and inductive reactance per phase per mile, Solution ‘The diameter of No, 2 wiro is 0.258 in. 0.258 X 0.7788 Deo 0.00836 ft Day = VIB KIBX 545 = 078 li Siig REGISTERED X1, = 2x60 X 2.083 X 10-* = 0.787 olf Dase mall VERSION ADDS NO WATERMARKor, from Tables A.t and A.2, Inductive reactance at Lft spacing = 0.581 Inductive reactance spacing factor for s = 9.2058 Inductive reactance per phase = 0.7868 ohm/phase/mile 2.14 Porallel-cirevit Three-phase Lines. ‘Two three-phase circuits that are identical in construction and electrically in parallel have the same inductive reactance. ‘The inductive reactance of the single equiv- lent circuit, however, is half that of each of the individual circuits considered alone only if they are so widely separated that there is negli- gible mutual inductance between them. If the two cirouits are on the same towers, the method of GMD may be used to find the inductance per pe ee ee ee, eee IND NIN (a) Prose ain (8) Phase aia (e) Phase ain poston 1 poston 2 pesiton 3 ye, 2.16 Arrangement of the eonduetors of a double-eireuit three-phase line in the three parte of the transposition eyele 4 phase by considering all the conductors of any particular phase to be strands of one composite conductor. Lat us consider the two three-phase circuits with flat, vertical spacing shown in Fig. 2.162. One circuit is composed of conductors a, b, and c. ‘The other is composed of conductors a’, b’, and e’. Conductors a and a’ tre in parallel and compose phase a. Similarly, conduetors b and b are in parallel composing phase 6, and ¢ and e’ in parallel compose phase «. In the other parts of the transposition eyele, conductors a and a’ take first the positions originally oceupied by b and b’ and then the positions occupied by ¢ and c’, as shown in Figs. 2.160 and 2.16¢. The induetgpee is lowered if the individual conductors of a phase are separate, as possible and if the distances between phases are ky results in a low Da and a high D,. Tt is accomplished it line by having the conductors of two of the phases Fig. 2.16, rather than horizontally adjacent By the method of GMD, the equivalent equi Deg = WDasDicDon ADDS NO WATERMARK &, & * < Crint-drig2 ELEMENTS OF POWER SYSTEM ANALYSIS. where Da. mutual GMD between phases a and b in position 1 = Wadglg = Vd Dy. = mutual GMD between phases & and cin position 1 = Da = Vag Da = mutual GMD between phases cand ain position | = -V/2dh Thus Der If the self GMD of each individual conduetor of phase a is r’, the self GMD in position 1 for the entire phase consisting of conductors a and : vit Habaghth¥ ¢ Da = VFI In position 2 the self GMD of phase ais a Da = vk sand in position 3 the self GMD of phase a is Daa VRE = Veh ‘The average value of the flux linkages of the phase for the whole trax position eyele determines the averageinductance. We sav fin Eqs. to (2.55) and again in Eqs. (2.69) and (2.70)] that the average of logatith- mie terms is equal to the logarithm of the geometrie mean of the argu- ments of the logarithms. Therefore, the equivalent self GMD of one phase for the transposition eyele isthe geometric mean of the three values of self GMD of the phase in the three parts of the transposition eyee. Thus D, = VWDaDaDa D, = (@pspon 77) Equations (2.76) and (2.77) are the same for all three phases if ris the same for all three phases, since phases b and ¢ occupy the same positio ‘as phase a for equal distances, Tho induetanee per phase is AO] Deg De L = 0.7411 log. 0.7411 to [ heise ge Equation (2.78), derived by the GMD method, is place sine the to pus conductor cea pale ERED pe tas strands of one conduetor. ‘The inductance of fonduclor Sy QR A. [=.REGISTERED 9) VERSION ADDS NO WATERMARK orate ()"] athensINDUCTANCE OF TRANSMISSION LUNES o As the two circuits of Fig, 2.16 are moved farther apart, the ratio g/f approaches one, If the two circuits are very far apart, the mutual inductance between them is negligible, and we would expect the induct- ance per conductor computed from Eq. (2.79) to approach that of a single circuit. Considering one of the circuits of Fig. 2.16 alone, we obtain Deg = W/Pddd = 2d and Substituting the above values in Bq, (2.78), we find efeonductor (2.80) which is the same as Eq. (2.79) if the ratio g/f is one. Therefore, we may consider the ratio g/f'as a factor which accounts for the mutual effect of one circuit with flat spacing on a similar parallel circuit ‘The preceding discussion shows the application of the GMD method to the computation of the inductance of a flat-spaced parallel-circuit line. Equations (2.75) and (2.77) to (2.80) apply only to flat-spaced parallel lines. It is not practicable to develop the special equations for other arrangements because the equations are complicated. ‘The GMD method is applicable, however, to any cireuits electrically in parallel, °@———1'——9— tions (2.74) and (2.76) apply to any + multicircuit three-phase line if we remember that Dus, Dye, and Dey are mutual GMD values. Sg Example 2.5 A three-phase double-cireuit line is composed of 19-strand concentric, copper conductors of 300,000-circular-_ “Q————18 ———© mil cross-sectional area, ‘The Tine is ¥*, 2:17 Arrangement of conduetors arranged as shown in Fig. 2.17 and {" Bxample 26. iscompletely transposed. Find the 60-cycle inductive reactaney per mile, Solution From Table A.1, for the specified conductor, Distance from a to b in original position = Distance from a to bin original position = Distance from a to a’ in original position =“ ELEMENTS OF FOWER SYSTEM ANALYSIS. D, = (VIE V3 TF) = /OOIIET (26.9 x 21 X 26.9)" = 0.141 X 4.98 = 0.702 ft Da = Die = SOOT KAT? = 1.38 ft Da = VOBX 20) = 18,97 ft Day = Y/T188 X18 X 18.07 ~ 16.1 fe 16.1 ei T= 0.7411 log girij5 = 1.01 millibenrys/n Xz = 2x0 X 1.01 X 10 ./phase 0.38 ohm/mile/phase 2.15 Summary. ‘Tables are helpful in computing the inductance and inductive reactance of a transmission line. If the self GMD of the con- ductor is obtained from a table, we can find the indnetance of a single- circuit line by Faq, (2.57) if the line is single-phase, or by Eq. (2.73) if the line is three-phase. ‘These two equations are the same except that the numerator of the argument of the logarithm of Eq. (2.78) for the three- phase line is the distance of equivalent equilateral spacing, rather than a single GMD as in Eq. (2.57) for the single-phase line, When more elaborate tables giving reactance at 1-ft spacing for various conductors and inductive reactance spacing factors are available, the inductance can be found by adding two values obtained from the tables, For multicircuit lines tables may be used as described above except that additional calculations are necessary to apply the principle of GMD. PROBLEMS 2.1 A hollow, eylindrical conductor has an outside diameter of 1.100 in. and a wall, thickness of 0.130in, "Find the #ux density ata distance of 0.485 in from the een tor of the conductor when the current is 500 amp. Neglect the effect of the retarn circuit 22 | Derive the formula for the internal induetanee in henrys per meter of a hollow conductor having an inside radius r; and an ontside radius 2, In what units should rand rs be expressed? 23 Determine the formula for the inductanee in henry per meter of a single-phase line consisting of the hollow conduetors deserted in Prob, 2.2 if the spacing between conductors is D ft. In what unite should rand r, be expressed? Compare the formuala with the seif GMD of an annular area given in ‘Table 2.1 24 Compute the 6Oryele inductive reactance at I-t spacing in ohms per mile for the hollow conductor whose dimensions are given in Prob. 2.1 2.5 Find the self GMD of a seven-strand conductor if the center staf and replaced hy a strand of zero conductivity. Exprese the ne radius r of an individual strand 2.6 Find the self GMD af « threo-trand conduetor in individual strand 2.7 Find the self GMD of each of the unconventional 2.18 in terms of the radius r of an individual strand and 4, where 4 isthe total area of the composite conductor i ‘hat all strands have the same radius and the same o VERSION ADDS NO WA ena & J we rint-driNgINDUCTANCE OF TRANSMISSION LINES 4s Be 3 & Fie, 2.18 Cross svetional views of unconventional eonduetors for Prob, 27, 2.8 Compute the U-eyele inductive reactance at Lf spacing in obins per mile of & cable consisting of 12 equal strands around a noneondueting core. The diameter of feoeh strand is 0.0996 in, The outside dianteter ofthe eable is 0.470 in 29. The outside dinmcter of the single layer of aluminum strands of No. 6 AWG ACSR conductor is 0,198 in, ‘The diameter of each strand is 0.0661 in, Determine the GO.eyele inductive reactance at L-ft spacing. Neglect the effect of the center strand of steal, but compare tho result with the values given in Table A. 210 The Gileyele inductive reactance at Lit spacing of a solid conductor is (0595 ohin mile. Find the reactance for a spacing of 6 ft, and determine the sectional area of the wiee in eireular mis 2.11. The 6ddeyele inductive reactance por conductor of a single-phase line having. conductors spaced 4 ft apart is 0.791 ohm/mile, Specify the 25-eyele inductive Teactance at Lt spacing for the conductors, What is the eros-ctional area of the conductors in cinetar mila? 2.12 ‘The distance hetween conductors of a single-phase line is 10 ft. Pach con- Aluctor is composed of seven equal strands, ‘The diameter of each strand is 0.1 in, Find the inductance of the line in hentys per mite 243 A single-phase (0. he spacing b cyele power line ix supported on « horizontal erosearm. en conductors is ft. A telephone ine s supported on a horizontal rsrt Ot below the power line, ‘The eonduetors of the telephone line are No, 1 AWG wolid copper spaced 2 ft between centers, The conductors of the power line ture No, 2 AWG solid copper, Find the mutual inductanee betwen the cireuits and the voltage per mile indueed in the telephone line if the current in the power line is, 15amp, 2.16 If the power and telephone Hines deserihed in Prob. 2.18 are in the horizontal plane and the distance betsvoen che nearest conductors of th {60 ft, find the mutual inductance between the eireuits and the voltage po in the telephone line for 150 amp in the poser line 248 The conductors of a three-phase lin ductor isa solid wire having a diameter of 0.162 in, inductance per phase in millihenrys per mile 246 A threesphse line i designed with equilateral to hnild the lino with horizontal spacing (Dis ansposed. What should be the spacing betwoon adja ‘obtain the sume inductance as in the original design Find the Ye isteren’e REGISTERED ON ductors tn order to VERSION ADDS NO“ ELEMENTS OF POWER SYSTEM ANALYSIS 2.17 A single-phase cireuit consists of three conductors on one side of the line fand one on the other side. The arrangement is shown in Fig. 2.19, The three con Fic. 2.19 Arrangement of eonduetons for Prob, 2.17 ductors composing one side of the Fine are transposed. Find the inductance por mile of this line 2.18 Six conductors of 19-strand hard-alrawn eopper with an arca of 300,000 el cular mils are arranged ns shown in Fig. 2.17, Tho vertical spacing, however, s 18 f, the longer horizontal distance is 28 ft and the shorter horizontal distances are 22 Tt If the line is operated single-phase with conductors a, bande in parallel forming one side of the line and conductors a ”, and e' forming the other side, find the inductance mile of the line. Assume equal eurrent in all conductor 219 A 192ky three-phase double-ciroit poster line is arranged with the con- uetors of each eireuit in a vertical plane, ‘The distance between adjacent eon ductors of the same citouit is 12 ft, The horizontal spacing betwoun circuits is 24 ft ‘The conductors are 556,508 cielar-mil ACSR, 30/7. Compute the inductance per phase of the double-citeuit line. Compare the induetance of one conductor of the ouble-ireait line with the induetanee of one eonduotor of a single eireut alone with the same vertical spacing, 2.20 If the line of Prob, 2,18 is operated three-phase, find the inductance per phase and per conductor. 221 Each phase of a three-phase line consists of three solid conductors. ‘Fhe ameter of each eonductor is 0.26 in., and the specing is shown in Fig. 2.20. Phases Arrange are designated a, 2, and ¢, There is complete transpos to individual condietors in each phase, Find the indue per mile, VERSION ADDS NO WATERMARKCHAPTER 3 CAPACITANCE OF TRANSMISSION LINES 3.1 Introduction. ‘The potential difference between the conductors of a transmission line causes the conductors to be charged in the same manner as the plates of a capacitor are charged when there is a potential difference between the plates of the capacitor. ‘The eapacitance betwe conductors is the charge per unit of potential difference. Capacitance is a constant depending on the size and spacing of the conductors. For power lines less than about 50 miles long, the effect of capacitance is slight and is usually neglected. For longer lines of higher voltage, capacitance becomes increasingly important An alternating voltage impressed between the conductors of @ trans iission line causes the charge on the conductors to increase and decrease with the increase and decrease of the instantaneous value of the voltage. ‘The movement of charge is a current, and the current caused by the alternate charging and discharging of a line due to an alternating voltage is called the charging current of the line. Charging current flows in a transmission line even when it is open-cireuited. Tt affects the voltage drop along the line as well as the efficiency and power factor of the line and the stability of the system of whieh the line is a part 3.2. Electric Field of a Long, Straight Conductor. Just as the mag~ netic field is important in considering inductance, so the eleetrie field is important in studying capacitance. In the preceding chapter we saw (Fig. 2.1) both the magnetic and electric fields of a two-wire line. Lines of electric flux originate on the positive charges of one conductor and terminate on the negative charges of the other conductor, cleetrie fiux emanating from a conductor is numerically coulombs of charge on the conductor. Bleetric flux der flux per square meter and is measured in coulombyg Ifa long, straight, e out its length and is isolated from other charge uniformly distributed around its periphe equidistant from such a conductor are points of a ’ze through- ‘the charge is fS:REGISTERED © q fscpotential and have VERSION ADDS NO WATERMARK &, & * < Crint-drig“ ELEMENTS OF POWER SYSTEM ANALYSIS the same electric flux density. Figure 8.1 shows such an isolated con ductor carrying a uniformaly distributed charge. ‘The electrie fiux density ‘at # meters from the conductor may be computed by imagining a cylin- drical surface concentric with the conductor and x meters in radius ‘Since all parts of the surface are equidistant from the conductor, which bas a uniformly distributed charge, the cylindrical surface is a surface of equipotential, and the eleetrie flux density on the surface is equal to the flux leaving the conductor per meter of length divided by the area of the surface in an axial length of 1 meter. ‘The electric flux density is coulombs/meter? (3.1) where q is the charge on the con- ductor in coulombs per meter of length and 2 is the distance in meters 13.31 Lines of electric ux orginat- 0m the conductor to the point jing on the positive charges waiformly where the electric flux density is dstributed over the suriace of an o- eomputed. The electric field inten- sa ee sity, or voltage gradient, is equal to the electric flux density divided by the permittivity" of the medium. Therefore, the electric field intensity is & volts/meter 2) 4 Teak 3.3. The Potential Difference between Two Points Due to a Charge. ‘The potential difference between two points is the work in newton meters (joules) necessary to move a coulomb of charge between the two points, Blectrc field intensity is a measure of the force on a charge in the field. Electric feld intensity in volts per meter is equal to the force in newtons on a coulomb of charge at the point considered. Between ‘avo points the line integral of the force in newtons acting on a coulomb of positive charge is the work done in moving the charge from the difference between the two points Consider a long, straight wire carrying a positive meter, as shown in Fig. 3.2, Points P; and P, al In the rationalized mks system of units the pej 8.85 X 10° farad/meter. Relative permittivity & jl imittivity bof a material to the permittivity of free space, Len 31 SHERED'O oN VERSION ADDS NO WATERMARK &, & * < Crint-drigCAPACITANCE OF TRANSMISSION UNES 6 D, and Dz meters from the center of the wire. The positive coulomb oon the wire will exert 2 repelling foree on a positive charge placed in the ficld. Since the force repels a positive charge in the field and since Ds in this case is greater than Dj, work must be done on a positive coulomb to move it ~. from Pr to Py, and P, is at 9 higher potential than Ps. The difference in potential is the amount of work done. On the other hand, if the coulomb moves from P; to Ps, it expends energy, and the amount of work, or energy, in newton-meters is the volt- age drop from P, to P2, The poten- tial differenee is independent of the path followed. ‘The simplest way to Fic. 82 Path of integration between Compute the voltage drop between the {9 Doin extemal fos ylindscal two points is to compute the voltage futed positive charges hetween the equipotential surfaces passing through P, and P, by integrating the field intensity over a radial path between the equipotential surfaces. ‘Thus the voltage drop between Py and Ps is me a _4 Ds 2 3 where g is the charge on the wire in coulombs per meter of length. Note that the voltage drop between two points, as given by Eq. (8.3), may he positive or negative depending on whether the charge causing the potential difference is positive or negative and on whether the voltage \lvop is computed from a point near the conductor to a point farther away, ‘or vive versa. The sign of q may be either positive or negative, and the logarithmic term is either positive or negative depending on whether Dz is greater or less than Ds. 3.4 Capacitance of a Two-wire Line. In See. 3.1 the eapacitanee between the two conductors of a two-wire line was defined as the charge ‘on the conductors per unit of potential difference between the form of an equation, capacitance is Path of. inegraon Vi farads/meter RED |y “4 SB Qa Vi ie Se REGISTERED & q VERSION ADDS NO WATERMARK &, & where q is the charge on the line in eoulombs goad potential difference between the conductors in helween two conductors may be found by subst * < Crint-drig0 ELEMENTS OF POWER SYSTEM ANALYSIS expression for Vin terms of g found by Bq. (8:3). The voltage Vax etwveen the two conductors of the two-wire line shown in Fig. 3.3 may be found by determining the poten- . Ss tial difference between the two con é © ductors of the ine, first by computing > the voltage drop due to the charge ge k— con conductor a and then by computing Fo, 33 Cross wction of @ parallel the voltage drop due to the charge qs oe on conductor b. By the principle of superposition the voltage drop from conductor a to eonduetor b due to the charges on both conductors is the sum of the voltage drops eaused by each change alone. Consider the change qs on conductor a, and assume that conductor bis uncharged and merely an equipotential surface in the electric field created by the charge on a ‘The equipotential surface of conductor b / and the equipotential surfaces due to the charge on @ are shown in Fig. 34. ‘The distortion of the equipotential surfaces near conductor b is caused by the fact that Intearatin pth conductor js also an equipotential sur | (( e}-tematos face. Equation (3.3) was derived by as- 7 suming all the equipotential surfaces due to a uniform charge on a round conductor to be eylindrieal and concentric with the conductor. Such is actually true for the ease under discussion except in the region near conductor b. The potential of con- ductor b is that of the equipotential sur- face intersecting b. Therefore, in deter 5, gy \aohipnvntial qusfaces mining Vas a path may be followed from 5's portion’ the leerie hld conductor a through a region of undis- caused by a charged conductor a torted equipotential surfaces to the equi- net shown, Conductor } causes potential surface intersecting conductor. ometitorted Then, moving along the equipotential option surface to b gives no further change in E¥ee" « point ou voltage. This path of integration is indi- the conduetal cated in Fig. 3.4 together with the direct erent ig path. Of course, the potential difference **°"™ jis the same regardless of the path over which field intensity is taken.? By following the patl *8ee W. H, Timbie and V, Bush, “Principles of F pp. 519-820, Jobn Wiley & Sons, Ine., New York, 10 Sin of the % Pst REGISTERED hineering" 4th od, /ERSION ADDS NO WATERMARK &, & * < Crint-drigCAPACITANCE OF TRANSMISSION LINES 1 region, we see that the distances corresponding to D; and Dy of Eq. (3.3) are D and ra, respe in determining Vs due to qs. Similarly, in determining Ves due to q, the distances corresponding to Ds and Dy of Eq. (3.3) are nm and D, respectively. Considering both ge and q, we obtain ya = fein 2 a ein — volts 6) Var = fein + Hen i @ iar aaa and, since qa qs for a two-wire line a= gc (in2 In) vot 6) or, by combining the logarithmic terms, fn(2) volts Va = n(&) rt @n ‘The capacitance between conductors is 7. a L rads/meter Ca = FS = apie farads/met (38) Converting to microfarads per mile, changing the base of the logarithmie term, and assuming a relative permittivity of ky = 1, 0.0388 C8 og wD sf/mile 9) lire =r 0.0388 _ 0.0194 Tog Dit ~ Tog Dr Co ni /mile (3.10) eapacitance between the conductors of a ‘tance Equation (3.10) gives the two-wire line, Sometimes it is desirable to know the cap between one of the conductors and a neutral point between them, For stance, if the line is supplied by a transformer having a grounded center tap, the potential difference between ground is half the potential difference between the twa condf the capacitance to ground, or eapacitance conductor per unit of potential difference betweg ground. Thus, the capacitance to neutral for the' the line-to-line capacitance (capacitance betwog line-to-line capacitance is considered to be comp tances in series, the voltage across the line divides and the point between them is at the gro h conduetg 0 neutral, is wae Bis usice “aN h@etors). If the [S, REGISTERED 7) potential. WERSION ADDS NO WATERMARK2 ELEMENTS OF POWER SYSTEM ANALYSIS capacitance to neutral is that of one of the two equal series capacitances, or twice the line-to-line capacitance. 80, 0.0388 DOS ul/mile, to neutral — G11) on = Ch ‘The concept of capacitance to neutral is illustrated in Fig, 3.5 Equation (@.11) corresponds to Eq. (2.36) for induetanee. One difer- ence between the equations for capacitance and inductance should be noted carefully. ‘The radius in the equation for capacitance is the actual outside radius of the conductor and not the self GMD of the ¢ 5 conductor as in the induetance for- OF uta. Als, certain approximations a hhave been ‘made in deriving the capacitance equation which did not enter into the derivation of the in- a : 4 ductance equation. (OF 2 EO natation (6.3), from which Co Cu (3.5) to (3.11) were derived, is based (6) Peprecettin of intent capacitance ot| the assumption of uniform charge Fie. 35. Relationship between thecon- distribution over the surface of the cepisoflinetrlinecapacitanceandline- egnductor. | When other charges are ore ea present the distribution of charge on the surface of the conductor is not uniform, and the equations derived from Kg, (9.3) are not strictly correct, Noruniformity of charge distribution can be taken into account by considering the conductors as ‘equipotential surfaces, which they are, rather than as uniformly charged conductors. ‘Then, without much difficulty, the following equation is found: (a) Representation of ine-o-Snecopsctance -nDe ui/mile, to neutral? (8:12) ‘The formula is for capacitance to neutral for a tywo-wire Tine only. For any but the simplest configurations of conductors of paallel-cireuit or three-phase lines, the derivation of an equation to account for the actual charge distribution becomes too involved to be practical, ‘The assumption of uniform charge distribution leads to very slight erggps i + An alternative form of the equation, giving identical results, j 0.0388 x 2.803 __ 0.0804 cosh W(D/2)~ oul *(D/R) ‘The detivation may be found ia texts on ol J.C. Slater and N, H, Frank, “Bleetromagnetisn Company, Ine,, New York, 1017; W. B. Boast, “Prin Fields,” pp. 92293, Harper & Brothers, New York, 194 Cm ile VERSION ADDS NO WATERMARK &, & 7 < Crint-drigCAPACITANCE OF TRANSMISSION UNES 8 the spacing between conductors is large compared to their diameters, which is the actual case for open-wire power transmission lines. For & single-phase line the amount of error involved when the charge distribu- tion is assurned to be uniform may be seen by comparing the capacitances computed by Eq. (8.11) to those computed by Eq. (8.12) for different ratios of D/r. Table 3.1 shows the error occurring when Bq. (3.11) is used instead of Bq, (3.12) ‘Tamun 3.1 Ramon Causnp wy Asseaune Usivont Cuanen Distrmcnon 18 Compunina Caracttaxce oF 4 TWo-WiRe Las Ratio D/r Per Cent Exvor in Ba, (11) 10 oad 2 0.084 50 0.010 100 0.002 200 0.0005 A question arises as to the value to be used in the denominator of the argument of the logarithm in Bq. (8.11) when the conductor is a stranded cable, since the equation was derived for a solid, round conductor. Since electric flux is perpendicular to the surface of a perfect conductor, the cleetre field at the surface of a stranded conductor is not the same as the field at the surface of a eylindrical conductor. ‘Therefore, the capacitance calculated for a stranded conductor by substituting the outside radius of the conductor for r in Bq. (3.11) will be slightly in error because of the difference between the field in the neighborhood of such a conductor and the field near a solid conductor for which Eq. (8.11) was derived. The error is very small, however, since only the field very close to the surface of the conductor is affected, ‘The outside radius of the stranded conductor is used in calculating the eapacitance. After the capacitance to neutral has been found, the capacitive react- ance existing between one conductor and neutral is found as follows: x10%16g 2 ahn/mite to nent) Some tables list capacitive susceptance at various spacings for the com- mon conductors or give the outside diameter from which capy reactance and its reciprocal, capacitive suseeptance, ean bi Other tables, ns suggested by W. A. Lewis, list cape eft spacing for the common conductors. Such table junetion with tables of eapacitive reactance spaci in Eq. (3.13) are in feet, capacitive reactance at term, and capacitive reactance spacing factor the equation is expanded as follows: RED a ikon ve ete AN “REGISTERED 9 q VERSION ADDS NO WATERMARK &, & S * < Crint-drigPs FLENENTS. OF POWER SYSTEM ANALYSIS Bc totog! + 188 5 oe b ohms/mile, to neutral (8.14) ‘The sum of capacitive reactance at 1-ft spacing and capacitive reactance spacing factor, as given by Eq, (3.14), is eapacitive reactance to neutral. Tables A.1 and A.2 in the Appendix list capacitive reactanice at I-ft ing, and Table A.1 lists values of capacitive reaetance spacing factor. ‘The use of capacitive reactance tables is similar to that of inductive renetance tables discussed in Chap. 2 Example 3.1 Find the capacitive susceptance per mile of a two-conductor single- phase line operating at 60 eps. ‘The conductors are each No, 1/0 seven- strand hard-drawn copper wire spaced 18 ft between centers. ‘This is the line described in Example 2.3. Solution ‘The outside diameter of the conductor is 3X 0.1228, oe 0.368 53 The radius r = yyy = 0.0158 ft ee 1098 90 toe 18 ; AED py Xe = SER? x 108 log g.grsg = 0-210 X 10 ohm/mille, to neutral be = 1/Xe = 4.76 X 10- mho/mile, to neutral bles of capacitive reactance at I-ft spacing and capacitive reactance spacing factor give 4.093 sae op —! 6p % 10" 108g 0.12 Capacitive reactance at I-ft spacing = 10* ohm/mile 4.08 Capacitive reactance spacing factor = 2553 x 10* log 18, 0.086 X 10° ohm mile Capacitive reactance to neutral = 10%0.124 + 0.086), 0.210 X 10% ohm ‘mile from whieh, _d = p2T0 M10 pacitive reactance and suseeptance from line to line Xe = 2X 0.210 x 10° Do = 4.76 X 10-4/2 = 3.5 Potential Difference between Two Co Chorged Conductors. If a number of eonduet be 1.76 X 10-* mho/mile, to neutral © My REGISTERED VERSION ADDS NO WATERMARK‘CAPACITANCE OF TRANSMISSION LINES 55 they are parallel to each other, the voltage between any two of them ean be found by applying Eq. (8.8) repeatedly to determine the voltage between the two conductors in question due to the ductor in the group indepencder ‘The voltage drop between the two conductors is the sum of the voltage drops due to exch charged vonduetor. Such a group of conductors is shown in Fig. 3.6. If we assume that there are no other charged surfaces in the vieinity, the sum of the charges on the conduetorsis zero, Ii the ground is far enough away to have negligible effect, and if we assume further that the spacing between conductors isp. gg large compared to the radius of any Sonducion, NP °F Peale! shersed one so that the charge distribution over the surface of a conductor will be uniform, repeated application of Eq, (3.3) will yield accurate results, $0, from conductor a to conductor », the voltage drop is Arge on each con 1 ne ven gi (un ton tem tombs) ain Each term in Bq. (8.15) is the potential drop from a to b due to the charge on one of the conductors in the group. In a similar manner the voltage drop may be found between other pairs of conduetors in the group. For example, 1 (win Be + am Peat SERED [ eo “aN [S.REGISTERED J) VERSION ADDS NO WATERMARK &, & If the voltages between conduetor a and the other ¢ the group of simultaneous equations expressing terms of charges may be solved for the charges sinusoidal and expressed as complex quantities, and are expressed as complex quantities. Simil ; < Crint-ari56 ELEMENTS OF FOWER SYSTEM ANALYSIS. for the complex values of flux linkages where the currents were expressed in complex form, ‘The equations above are seldom solved to find charge but are used in the derivation of capacitance formulas for polyphase circuit, 3.6 Capacitance of a Three-phase Line with Equilateral Spacing. ‘The three identical conductors of radius r of a three-phase: line with equilateral spacing are shown in Fig. 3.7. Figure 3.8 is the phasor diagram of voltages for this line. ‘To solve for the capacitance to neutral, wwe first write the expression for the voltage drops from conductor @ o Dy > e D e Fro. 3.7. Cross section of a three-phase asor diagram of the bale Tine with equilateral sparing. anced voltages of a three-phase line to conductor b and from conductor @ to conductor ¢, ‘Thus, from Eqs. (3.15) and (3.16) 1 Deow teem? ais Van ge (ein? + ming +0 np) volts.) and Adding Eqs. (3.18) and (3.19) gives Vac = If we assume there are no other charges in the vicinity, thgggamy charges on the three conductors is zero, and we Eq. (8.20) for q + ge and obtain ‘i = Be jy, P From the phasor diagram of Fig, 3.8, we obt between the line voltages Vs and Va. and thy Si REGISTERED ON Yee Vee from line VERSION ADDS NOCAPACITANCE OF TRANSMISSION LUNES s to the neutral of the three-phase circuit: Ven = V8 Von(0.866 + 50.5) (22) Ver = Vn = V3 Voq(0.866 — 30.5) (3.23) Adding Eqs. (3.22) and (8.23) gives Ves + Voc = 3V¥ an (3.24) Substituting 3Vs, for Vas + Vos in Bq. (8.21), we obtain te iD fn 2 volts Since capacitance to neutral is the ratio of the charge on a conductor to the voltage between that conductor and neutral, Qe _ Oak Vi, ~ in DF farads/meter, to neutral (8.26) For a relative permittivity of ky = 1 2 = 2288 ie to net an Cumpaion of Bg. (27) and (1) shows hat evo ae ene culty spaced treopanc ne repeesrey. We mn Chap 2 pice and cautery pact troephane ne expan of ie Fora shepae i the sharing erent ihe produto the tineotne tla and he faeto ine snepanee Taig = jaCaVen (8.28) For a three-phase line, the charging current is found by multiplying the voltage to neutral by the capacitive susceptance to neutral, ‘This gives the charging current per phase and is in accord with the caleulation of balanced three-phase circuits on the basis of a single phase with neutral return, ‘The charging current in phase a is Tote = ja0nVon 37 Capacitance of a Three-phase Line with Uns, When the conductors of a three-phase line are n the problem of ealeulating the eapacitance bear such a line is not transposed, the capacitances are unequal, and, if the line is transposed, the} VERSION ADDS NO WATERMARK8 ELEMENTS OF POWER SYSTEM ANALYSIS. phase to neutral is different for each position occupied by the conductor in the transposition eycle, In the transposed line, however, the average capacitance to neutral of any phase for the complete transposition eyele is the same as the average capacitance to neutral of any other phase, since each phase oceupies the same position as every other phase over an equal distance, The dissymmetry of the un- transposed line is slight for the usual configuration, and capacitance caleu- lations are carried out as though all Tines were transposed. Equation (3.15) may be applied to fy Dx '3 the line shown in Fig. 3.9 to compute 3.9. Cross section of a three. Vas due to the charges on all three con- line with unsymmetrical ductors, ‘Three equations are found for Va for the three different parts of the transposition eyele. With phase a in position 1, bin position 2, and c in position 3, 2 pha Spacing. Y= she (gen 28 + asin Fe toms) vols 6.30 With a in postion 2,6 in position 3, and e in postion 1, Va= oy (Ce Pin ge tain pH) volts B21) and, with a in position 3 6 in position 1, and e in position 2, Yau = gh (ooin Pe + asim gh tac B2) volte 22) Equations (3,30) to (3.2) ave sina in form to Bas. 2.66) to 2.08) forthe fx Hokage of one conductor of transposed line. Bach ofthe later equations ves the vale of tix Knkage in oe part ofthe tant position cyclo An average value of flux Hnkayes over the complete itanspetioneyse was found by noting that the eaten ia any ph iit ength onthe eondtor af each phas the of that phase as for any other positon in the tans sarod an 0) 1.2) wld une LEN Capecttance ta neutral of phate in one pert g yen ‘yele Os is in parallel with the capacitances to neutral ? LSREGIST! ERED © q VERSION ADDS NO port te eanapeation epee, Therere i WATERMARK &, & * < Crint-drigCAPACITANCE OF TRANSMISSION UNES ” along the line, the voltage to neutral of a phase in one part of the eycle is equal to the voltage to neutral of that phase in any part of the ¢ Honee, the voltage between any two conductors is the same in one part of tho transposition eyele as in other parts between the same conductors Since the voltage is the same anywhere in the transposition eyelo, it follows that the charge on any conductor must be different when the position of the conductor changes with respect to the other conductors A treatment of Eqs. (8.80) to (8.32) analogous to that of Eqs. (2.60) to (2.68) is not rigorous, Equations (3.30) to (8.82) have 10 unknowns, the voltage Vay and nine charges, for the charge on each of the three conductors is different in each of the three positions occupied by a eanduetor in the three parts of the transposition cycle. ‘Thus, a rigorous solution for capacitance in terms of the ratio of charge to potential difference requires six more equations in order to eliminate all the unknowns exeept one voltage and fone charge, ‘Three additional equations, similar to Eqs. (3.30) to (8.32), may be written for the voltage Pre, and the latter voltage may be expressed as Vin(—0.5 ~ 0.866) if the voltages on the line are assumed to be balanced. ‘The other three equations required may be obtained iy equating the sum of the charges in euch of the three parts of the trans position eyele to zero. ‘The rigorous solution for capacitance is too involved to be practical except perhaps for flat spacing with equal distances between adjacent conductors. With the usual spacings and conductors, sufficient aeeuracy is obtained by assuming the charge per unit length on a conductor to be the same in every part of the transposition eycle. When the above assumption is made with regard to charge, the voltage between a pair of conductors is different for cach part of the transposition eyele. ‘Then an average value of voltage between the conductors ean be found, and the capacitance calculated from the average voltage. We obtain the average voltage by adding Bqs. (8.80), (8.31), and (8.82) and dividing the result by three. ‘The average voltage between conductors a and 6, hhased! on the assumption of the sume charge on a conductor rogardless of its position in the transposition eyele, is 1 DisDssDa1 ca ink [+ i (PabePe ' ao (oxtsm;) Vos aay SRN REGISTERED 9) VERSION ADDS NO WATERMARK &, & Sang) vot where Day = VDDD, * < Crint-drigry ELEMENTS OF POWER SYSTEM, ANALYSIS Similarly, the average voltage drop from conductor @ to conductor cis Vue = shy (ae n Pet ira ole 35 van sh (nin Betetne) wis as ek 4) to find the voltage to neutral, we have Da tain ) volts (3.36) Applying Ven = Vas + Voc Since qe + 4s + ge = 0 in a balanced three-phase circuit, Ven volts (37) and eek Vou” In Delt For a relative permittivity of hry 0.0388 0g Dev Equation (3.39) for capacitance to neutral of a transposed three-phase line corresponds to Eq. (2.71) for the inductance per phase of a similar line. C= farads/meter, to neutral (3.38) C. af /mile, to neutral (3.39) Example 3.2 Find the capacitance and capacitive reaetance per mile of the line described in Example 2.4. Tf the line is operated at 22,000 volts, find the charging current per mile Solution i Po ass oe Co = ag Be a sy ~ 01488 f/m, to nour Xe= Z = 0.185 x 10* chm /mile, to neutral Fe XW X TOTS or from tables Capacitive reactance at 1-ft spacing Capacitive reactance spacing factor for 5.45 ft Capacitive reactance = [=.REGISTERED ON VERSION ADDS NO WATERMARK‘CAPACITANCE OF TRANSMISSION LUNES * The magnitude of the charging current is 2 X 60 X 0.01438 x 10-* x 22,000/+/3 = 688 X 10°? amp/mile Tose 3.8 Effect of Earth on the Capacitance of Three-phase Transmission Lines. Earth affects the capacitance of a transmission line because its presence alters the clectrie field of the line. If we assume the earth to be a perfect conductor in the form of horizontal plane of infinite extent, we realize that the electric ficld of charged conductors above the earth is not the same as it would be if the equipotential surface of the earth were not present, ‘The electric field of the charged conductors is forced to conform to the presence of the earth's equipotential surface. Consider a circuit consisting of @ single overhead conductor with a return path through the earth. In charging the conduetor, charges come from the earth to reside on the conductor, and a potential difference exists between the conductor and earth, ‘The earth has a charge equal in magnitude to that on the conductor but of opposite sign. Blectrie flux from the charges on the conductor to the charges on the surfac of the earth is perpendicular to the earth’s surface, since the surface assumed to be a perfect conductor. Let us imagine a fictitious conductor of the same size and shape as the overhead conductor lying directly below the original conductor at a distance equal to twice the distance of the conductor above the plane of the ground. The fictitious conductor is helow the surface of the earth by a distance equal to the distance of the overhead conductor above the earth. If the earth is removed and a charge equal and opposite to that on the overhead conductor is assumed on the fictitious conductor, the plane midway between the original conductor and the fictitious conductor is an equipotential surface and ‘occupies the same position as the equipotential surface of the earth. ‘The clectrie flux between the overhead conductor and this equipotential surface is the same as that which existed between the eonduetor and the earth, ‘Thus, for purposes of calculation of capacitance, the earth may be replaced by a fictitious charged conductor below the surface of the earth by a distance equal to that of the overhead conductor above the earth, Such a conductor has a charge equal in magnitude and opposite in sign to that of the original conductor and is called the image g ‘The method of calculating capacitance by replacing the image of an overhead conductor can be extended to conductor. If we locate an image conductor forg ductor, the flux between the original conductors' perpendicular to the plane which replaces they is an equipotential surface. ‘The flux above the when the earth is present instead of the image eg images is that, plane Sa "REGISTERED ©, q VERSION ADDS NO WATERMARK &, & * < Crint-ari2 ELEMENTS OF POWER SYSTEM ANALYSIS hase line, refer to Fig ed and that conductors a, b, and ¢ carry the 4 q. and occupy positions 1, 2, and 3, respectively, in the first ps transposition eycle, ‘The plane of the earth is shown, and below it are the conductors with the image charges —q., —g, and —g-. Equations “6 REGISTERED 9) ERSION ADDS NO WATERMARK & oe Op; $ Crint-drigCAPACITANCE OF TRANSMISSION UNES 6 With conductor a in position 2, b in position 3, and c in position 1, )} es and, with a in position 3, b in position 1, and ¢ in position 2, Do _ yy Hs ty Hh [2e(1m 22 — 1m Be) toon gf ~ nf) Distant) a(uBe—nls)] ee If the approximately correct assumption of constant charge per unit length of conductor throughout the transposition eycle is made, an aver= age value of Vp for the three parts is {(osahans) (arias) |} DiDat Similarly, Vac = sh n(n Bet — mn Ss ‘m= Ok | r VHA, a WH + ge(In 2 = In-Y SE )] (3.45) ol 5,-" Sree) 6 and 1 D. TI Von = gh [2m (In Pet ~ 10 Yatra) HV ata 20 (0 VHsHally r YH al w+ ge) (In — in YE )| +t a(t Da" VM Since qe + q+ 9 = 0 y 3 Dea VA H all and SS Dagit — log (X/HiHasH / W/HalT REGISTERED Dio neutral (B48) VERSION ADDS NOa ELEMENTS OF POWER SYSTEM. ANALYSIS Comparison of Eqs. (3.39) and (3.48) shows that the effect of the cearth is to inerease the capacitance of a line, ‘To account for the earth the denominator of Bq. (8.89) must have subtracted from it the term log (/Hullslls/W/HIGH,). Uf the conductors are high above ground compared to the distances between them, the diagonal distances in the numerator of the correction term are nearly equal to the vertical dis tances in the denominator, and the term is very small. ‘This is the usual case, und the effect of ground is generally neglected for three-phase lines except for calculations by symmetrical components when the sum of the three line currents is not zero. Calculations of capacitance for this condition will be cousidered in Chap. 12. 3.9 Parallel-circuit Three-phase Lines. Let us consider two special arrangements of parallel-cireuit lines, the double-cireuit line with hex- agonal spacing and the double-cireuit line with flat, vertical spacing. The equation for capacitance of each of these lines is relatively simple Many double-cireuit lines have flat, vertical spacing, and the spacing of most other double-circuit lines is intermediate between flat spacing and hexagonal spacing, Consider first the double-cireuit line with hexagonal spacing shown in Fig. 3.11. Phase a is composed of con- ductors a and a’, phase b of condu tors b and 6’, and phase c of conductors, cand e’. ‘The two eonduetors of each phase are electrically in parallel and have the same charge. Because of the symmetrical arrangement the phases are balanced, and the condue- tors of each individual phase are also balanced, if the effect of ground is neglected. Therefore, transposition of the conductors js not necessary to balance the phases. ‘The equations for voltage drop may be written in the usual manner, and the derivation of the expression for eapacitance proceeds as follows: Fro, 31 Double-cirenit three-phase line with hexagonal spacing. ronifa(nt au) raGus +0 8) +1.(n aig te — 0») In (YP) a (3.50) 5. REGISTERED VERSION ADDS NO = tea n(CAPACITANCE OF TRANSMISSION UNES 4s Van + Vue = Won = 3 One — a — 29 0 (YEP) votts 6.52) and, since qa + q@ + Ge = 0, 5D) vats (8.58) then te Del a5 og = gfe = PAE farade/meter/eonductor, to neutral (8.54 On BL A bray farmn/meter/cond tral (3.54) 0.0388 uf/mile/conduetor, to neutral log (V3 D/2r) Equations (3.64) and (3.55) give the capacitance from one conductor to neutral, not from one phase to neutral. The expression for capacitance was found by taking the ratio of the charge on only one of the two con ductors of a phase to the voltage to neutral. ‘To find the eapacitance to neutral per phase, we note that each phase consists of two identically charged conductors in parallel. ‘Therefore, the capacitance to neutral per phase is twice the capacitance to neutral of one conductor, or 0388 0, = 2X — log (V3 D/2*) uf/mile/phase, to neutral (3.50) We recall that the inductance of parallel-cireuit three-phase lines was calculated by using the method of GMD. Let us apply a modified method of GMD to the ealeulation of the capacitance of a hexagonally- spaced double-cireuit line. In applying the method to capacitance cal culations, actual radii of the individual conductors of a phase will be used to obtain the modified self GMD of a phase. We speak of the modified self GMD and the modified method of GMD in connection with eapaci- tance ealeulations beeause we are not following the mathematical concept of GMD discussed in Chap. 2 when we use the actual radius of a conductor composing one of the circuits instead of the self GMD of that conductor. We must use the radius of a conductor rather than its self GMD bg all the charge resides on the surface. The idea of self inductance calculations because of the internal flux ductor, We shall still combine the parallel conduct the method of geometric mean distances, and we # in all other respects. For the hexagonally space ‘modified self GMD for all phases is the same differs from the self GMD by the substitution q ERED 1s Cia) re method AN Fig. 3.11 the = REGISTERED 9 q VERSION ADDS NO WATERMARK &, & 3 <. Crint-drigrs ELEMENTS OF POWER SYSTEM ANALYSIS. D, GMD atob MD b toe GMD c toa 5D Dey = W/DasDicDay = V3 D ing r in Eq. (8.39) by Day 8 0.0388 B_ 0.0888 __ 5, __0.0888_ log Deg! D. log (W/3 D//2rD) log (V8 D/2r) wi/mile/phase, to neutral which is identical to Eq. (8.56) and shows that our modified GMD method is valid in this ease. We note that the GMD method always gives per phase values, rather than per conductor values, because it eom- RO RC NC RONG (2) Phase ain e Pseie (e) Phase ain Dratton | postion 2 fostion 3 12. Arrangement of the eonduetors of a double-cirenit three-phase line in the tree parts ofa ransposition eye bines the conductors electrically in parallel in any one phase in computing distances, Now consider a double-circuit line with flat, vertieal spacing, as shown in Fig. 3.12. Such a line is not balanced without transpositions, ‘There- fore, the derivation will be made for a transposed line, and again charge on ® conductor per unit length will be assumed to remain the same ‘throughout. the transposition eycle. Since the conductor of each phase of one circuit is in electrical parallel with the conductor of the same phase in the other cireuit, a conductor in one cireuit has the same charge as the conductor in the other circuit with which it is in parallel, Tn order to compute the voltage drop from a to h, the charge 9. on conductor qgand. aan identical charge qe on conductor a! must be considered on the conductors of the other phases are treated in q j)ro(og ent JSPs RecisTereD 9) /ERSION ADDS NO WATERMARK &, & * < Crint-ariCAPACITANCE OF TRANSMISSION LUNES ” With a in position 2, ee 4 inf)ea(ing + inf Va ~ shy[se(imd +f) 4 a(wg ent) ea eeemtE\ eee Fa(in2t +in4)] @.39) and, with a in position 3, LE (ain aclmt owt a= hale (m2! +n) + (ing +m0f) 441.9) oso +a(ntemd)] eso ‘The average value of Vas in the three positions is soe Lan (n 24 42) — (m2 4 1n)] ¢ Vo gyn (n2 + inf) — o(n2t +i0%)] eon Similarly, Ya = ot 2 nt) a(n ame)] es Vu = gy [oe(n 2 +n) — 2 (m2 +m) cov and a Bae (2: Vat Ben (2) = fo P24@)] vos Ge sf/mile, one conductor to neutral (8.61) rl ‘The capacitance to neutral per phase (two conductors in parallel) is 0.0385 t/mile/phase, to neutral (3.65) wel) Now let us apply our modified GMD method to the derivation of the expression for capacitanee to determine whether the method is valid. By comparison with Eqs. (2.75) and (2.77), C,=2% gh and where r replaces r' since we are dealing wit inductance. ‘Then ADDS NO WA TERMARK & fg Crint-drig“ ELEMENTS OF POWER SYSTEM ANALYSIS. 0.0388 _ 0.0388 9.0388, log Dec/De ( yO] 7 f/mile/phase, to neutral (3.67) Since Eqs. (8.05) and @. asn are identical, the modified GMD method holds to the same close approximation as Eq. @.65). Both equations are slightly in error because of the assumption of the same charge on a conductor in any position of the transposition cycle, because of the noglect of the effect of earth, and because we have assumed in all our derivations a uniform distribution of charge over the surface of the conductors, All these differences are negligible for the usual overhead line. Since the modified GMD method has been shown to be valid for hexagonal spacing and for flat, vertical spacing, it is reasonable to assume that it may be used for arrangements intermediate between the two. (3.66) we Cn = 2X Example 3.3 Find the capacitance and the 60-cycle capacitive suseeptance to neutral per mile per phase of the double-circuit line described in Example Solution From Example 2.5, Deg = 16.1 ft ‘The ealeulation of the modified D, is the same as in Example 2.5 except that r is used instead of r'. ‘The outside diameter of 19-strand, 300,000-cireutar-mil conductor is 0.629 in, 0.629 _ p 994 Pe = 0.026 1 = (V26.9 X 0.026 V 21.0 X 0.026 20.9 X 0.036)" -V0.026(26.9 x 21.0 X 26.9) = 0.808 ft 0.0388 Cu = TE TG-TOROR) 7 00299 af/mile/phase, to neutral b, = 2efC = 2x X 60 X 0.0299 = 11.27 mieromhos/mile/phaggs to 3.10 Summary. ‘The capacitance of a single-cireuit by Eq. (8.11) if the line is single-phase, or by Eg three-phase. ‘These two equations are the same A ator of the argument of the logarithm of Eq. (3 equivalent equilateral spacing of the line between the two conductors of a single-phase J [S-REGISTERED 9, q Sh both equations, VERSION ADDS NO WATERMARK &, & * < Crint-ariCAPACITANCE OF TRANSMISSION LUNES. % rr is the outside radius of the conductor. For parallel-cireuit lines, a modified method of GMD is used wherein the outside radius of a con- ductor enters the computations in place of the self GML of the conductor found in inductance calculations. Several approximations are made in deriving capacitance formulas, but the importance of all of them, inelud- ing the effect of earth, is usually very slight PROBLEMS: 8.1 A three-phase transmission line has Bat, horizontal spacing with 6 ft between adjacent conductors, At a certain instant the charge on one of the outside eon ductors is 0.1 X 10 eoulombs/mile, and the charge on the center conductor and on the other outside conductor is —0.05 X 10°* coulomb/mile. ‘The radius of each conductor is OL in, Neglet the effect of ground, and find the voltage drop between, the two identically charged conductors at the instant specified. ‘82. The 60-cycle capacitive reactance to neutral of @ solid conductor, which is one conductor of a three-phase line with an equivalent equilateral spacing of 4 ft, is 186 X 10 ohms/mile. What value of reactance would be specified in a table listing the capacitive reectance of the conductor at: Ift spacing for 25 eps? What is the crosesectional ares of the eonduetor in circular mils? ‘33 Derive an equation for the capacitance to neutral per mile of a single-phase Iine, taking into account the effect of ground. Use the same nomenclature ws in the ‘equation derived for the capacitance of n three-phase line where the effet of ground is ropeeeented by image charges 8.4. Calculate the capacitance to neutral per mile of a single-phase line composed of two No, 2 single-strand conductors spaced 10 ft apart and 25 ft above ground. Compare the values obtained by Eqs, (.11) and (3.12) and by the equation derived in Prob. 8.3. ‘36. Derive w formula for the capacitance between the single inner conductor and the concentric outer sheath of a power eable, Assume that the radius of the inner conductor is @ and that the inner radivs of the sheath is b ‘2.6. single-conductor power cable has a conductor of No.2 solid copper. Paper insulation separating the conductor from the concentric lend sheath has 2 thickness of {gg in, and a relative permittivity of 2.7. The thiekneae of the lead sheath is S¢q in. Find the capacitive reactance per mile between the inner conductor and the lead sheath, 3.7 "A three-phase transmission line has two conductors 8 ft apart in « horizontal plane, “The third conductor is $ ft above the plane of the other two and midway between them. ‘The conductors are solid, round wires with a eapacitive reactance at it spacing of 0.1345 megohm mile at 60 eps. Find the eapacitive reactance to neatral per ile of line at 60 eps, and find the radius of the wire 38 A three-phase GO-eycle transmission line has fat, horizontel spac between adjacent conductors. The conductors are No. 2/0 hard copper. ‘The voltage of the line is 110 kv. Find the enpacitive ckctenttoeememe: Acne ° S Titan tie rfthecyrenactne ose Ste REGIST ERED ON VERSION ADDS NO WATERMARK & per mile per phase and per conductor at 182 kv ancl ee ‘ON GISTERED © Fiat RSION ADDS NO WATERMARK &, & where I. is the current enclosed by the tubular ol Tis the rms value of the current, His the rms vq We shall find it necessary to deal with instantan it is desirable to express Bq. (4.4) in instantangl 7 < Crint-drigon LEMENTS. OF POWER SYSTEM ANALYSIS al a where the symbol sie means “real part of.” It is customary to omit the symbol e, so that taneous value of the field intensity is Se [Haauat] = Be [! ») Lane Haast (46) and, upon dividing both sides of the equation by +/2 to convert from. maximum to rms values, we have 1 Hac = a7 whieh is the expression of Hq, (4-4) in instantaneous form. ‘To find the current J_ in terms of eurrent density, let J, be the current density at # distance x from the center of the conductor. ‘Then the current in the walls of the tubular element of radius 2 and wall thickness de is 2rxJ de, and the current enclosed by the tube (that is, in the cylinder of radius 2) is fo [f2ead te as) Introduetion of the factor «i follows an accepted convention of notation, ‘The instantancous value of a current which varies sinusoidally may be expressed by f= Manl com ot Where (Foal is the magnitude, or absolute value, of the maximum current. Another way of expressing the same current is By Euler's formula, oH = 08 ot + J sin ot and Sie [om = 008 at “Therefore, Ge (Mma = [nl 608 et 1 £ = Haul 008 (ut +4) we can ay Ge [male] = Re [uae ‘and, letting Tom Mase = Hal de we have ERED va hr Zoi comple Py petog Ee hy ths notation Porta dacoaen a te mits Nea WoL py 7005 Joho Wie & Som fae EGISTERED 9) VERSION ADDS NORESISTANCE AND SKIN FECT 7 Substituting in Eq. (4.7), we obtain ent ast = [ana do as) ‘Taking the partial derivative with respoct to x yields de [> A cree) + met] = Deedee .10) deze tHe att m Dea aay Dividing by 2rze, we obtain aul, ae Je (412) Enuation (4.12) contains both JT, and J. as variables dependent upon 2, If we can find another relation between /, and J, we ean el H, and oblain a differential equation having Jz as the only variable dependent on 2. Such a relation ean be found by applying Nirehhoft’s voltage law to the voltage drops around the closed path a'l’ba shown in Fig. 4.86. ‘The voltage drops consist of ohmie drops on the paths ab and a'l/ and of a voltage drop caused by the changing ffux linking the closed loop a’¥’ba, ‘The instantaneous ohmic drop from a to D i J.mse!p Al where Jame" is the instantaneous current density on the path al and is the resistivity of the conduetor.? Similarly, from a’ to b’ the ohmic drop is [vena + 2 aan da] pat (4.18) paths. Around the loop a’b'ba the total ohmie drop is ERED pp 'S REGISTERED 9) VERSION ADDS NO* LREAENTS OF POWER SYSTEM ANALYSIS at[ Foam + & Gamat) d2| ~ pdm = MZ Vamnelt) dz (Ho14) ‘The voltage drop clockwise around the loop a’b’ba due to the changing, flux in the loop is —a@/at. ‘The negative sign is necessary because increasing flux due to current in the direction shown induces @ voltage rise in a clockwise direction. ‘The voltage drop is the negative of the voltage rise. Rquating the voltage drops around the loop to zero, in accordance with Kirchhoft’s law, gives ul sist) dg — 2% Z Canmeia) de — 8 pal -0 (4.15) The flux ¢ linking the path a’b/ba is in the tubular element of thiek- ness dr and is concentrie with the tube. It is a function of both time ¢ and distance « from the center of the conductor. Equating the instan- taneous flux to the product of instantaneous flux density and ares gives 8 Brot Md = pHayace Adz (4.16) ‘Therefore, upon substitution of Eq. (4.16) in Eq. (4.15), we obtain p31 saab de ~ 2 amb de =0 ‘Phen converting to rms values, assuming constant permeability, expand~ ing the partial derivative with respect to f, and noting that Heme is not a function of ( and that is not a funetion of x, we have al, p Ale’ Jes Ale" ang = 0 (4.18) In interpreting Bq, (4.18) we must remember that Hand J, are complex and that in taking the partial derivative with respect to f we have omitted the symbol ie. The student ean show, by replacing &*by eos af + jsin of that Bae tne] = &e | 5 from which it follows that, J Be Unmet) = te Ljataet and, omitting Sate) = jal whieh ie the expected result for the derivative of the tnd imaginary components of Zysi, We must rement taneous value of 0/4¢ is the horizontal projection (eeal py fs yeeistered o fe), VERSION ADDS NO WATERMARKRESSTANCE AND SKN EFECT 7 Dividing Bq. (4.18) by & and by «V to obiain rms values, we have ip ad, Mem oe dz (4.19) Substituting H, from Eq. (4.19) in Eq. (4.12), we obtain (4.20) or, multiplying by jeu/a and noting that J. (a complex number and fune- tion of 2) is not a function of t, (4.21) The change from partial derivatives to total derivatives is possible in Bq. (4.21) because the only independent variable is z. Eq, (1.21) is the second-order differential equation relating the rms value of the eurrent density to the distance from the center of the conductor. Bagstion (423) i a spot form of the longrocymied Basel equa von’ Te may be ttn in more enti form as flows ay Ady yas, 29 Steet eyo (422) To solve Eq. (4.2: assume a solution in the form of an infinite series, or Y= a9 + ae + get faye + + tage + (4.23) Then re ae eee te 1a 2 4 bay + Base + dove! + Base? + Caer! 425 Fae mE + Bae + Bast + dows? + Bat? + Bert + (4.25) and By = Way + Bose + Mast + Bast! + ae! + + (4.08) ‘The solutions to Eq. (4.21) are called Bessel functions of zero order, The Bessel ‘quation having solutions of the nth order is y=0 SMERED 1/7 SB = REGISTERED 9) VERSION ADDS NO WATERMARK & 7 = REGISTERED ON VERSION 2D ADDS NO® ELEMENTS OF FOWER SYSTEM ANALYSIS. Rem, % 1,000 ohms/mito (452) Substituting Ro/1,609 for p/xr? in Eq. (4.51) gives NR where yu, = relative permeability of the wire f = frequeney, eps Ry = dee resistance of the wire, ohms/mile ‘Tabulated values of the ratio of effective resistance to d-c resistance calculated from Eq. (4.48) have been published by the U.S. Bureau of Standards. The resistance ratios plotted in Fig. (4.4) this source, which lists the ratios for values to mr = 100. At fre- quencies of 60 eps or less, stranding has negligible effect on the ratio of cifective to d-e resistance of con= centrically stranded conductors, and effective resistance may be found by multiplying the d-c resist- ance of the stranded conduetor by the ratio read from Fig. (4.4) for a 46 Resistance from Tables of Conductor Characteristics. Some of the factors considered in the dise cussion of resistance and skin effect can be verified by referring to the tables of conductor characteristics mr = 0.0636 ./¥! (4.58) Pio, 44 Ratio of sor resistance to d resistance for a. eylindrieal conductor faving « uniform magnetic field aronndd the periphery. The ratio is plotted aa in the Appendix. ‘The increase of function of mr, here cee resistance caused by stranding me = (40036 V/s 77 is illustrated by a hard-dyawn, tnd Ios the dhe resistance in ohms pee per conductor with a tional area of 66 and composed of 1, 8, or 7 strands, ‘The d-e resjgtay single strand is listed in Table A.1 as 0.864 ohm: eeeO “es QS cesotion of Sy PS REGISTERED ON VERSION ADDS NO WATERMARK &, & Sco F. B. Rom and F, W, Grover, “Formulas and Mutual and Self Inductance,” Seientific Paper 168, ul ‘Table XXIF, 1912, * < Crint-ariRESISTANCE AND SKIN EFFECT 6 the d-c resistance is 0.873 ohm/mile, and for 7 strands the value is 0.881 obm/mile, Note that the resistances for the 3- and 7-strand conductors ‘are 1% and 2%, respectively, above the resistance of the solid conductor This is in agreement with the prineiple stated in See, 4.1 ‘The values of 0.864 and 0.945 ohm/mile at 2, respectively given in Table A,1 for the 66,870-circular-mil copper conductor, are verified by Eqs. (1.2) and (4.8). ‘The dee resistance at 20°C of a solid conductor is 10.66 5,280 _ R, 0.848 ohm/mile = 0.864 ohm/mile = 0.945 ohm/mile mination of the tables shows that skin effect at frequencies up to 60 eps is negligible for the smaller conductors, ‘The 60-eycle resistance of the 66,870-circular-mil conductor is equal to the d-e resistance, Skin effect becomes appreciable, however, at power frequency for the large con~ Uuetors, For instanee, the d-e resistance of 500,000-cireular-mil hard~ drawn copper conductor with either 19 or 37 strands is 0.1280 ohm /mile at 50°C, but the GO-eycle effective resistance is 0.1303 cbm/'mile. For this conductor, stranding does not appreciably aller the ratio of effective to d-c resistance computed from Fig. (4.1). From Eq, (4.53) [wo \ and from Fig. (4.4) the resistance resista mr 1.38 0.0036 ratio is 1.02. ‘Then the 60-cyele R = 1,02 X 0.1280 = 0.1305 ohm/mile 47 Skin-effect Inductance Ratio. ‘The imaginary component of the internal impedance of a conductor is the inductive reactance due to internal flux linkages. Rationalizing the expression for internal imped- sanee given by Eq, (4.45) and disearding the real par expression for internal inductive reactance: pm bei mr beif mr -+ ber mr ber’ mr ar (Dei mr)? + (her mir) Equation (2.18) gives the internal flux linkages of a Dividing Bq, (2.18) by the q | inductance at frequencies s¢ ant is valid. ‘Thus ols = hg uniform cs Lio which, is ‘REGISTERED ©, q VERSION ADDS NO WATERMARK &, & current densit is the inte uniform cur * < Crint-drigu REAENTS OF FOWER SYSTEM, ANALYSIS Lig = g- —henrys/meter (4.55) and from Eq. (4.54) the ratio of intemal induetance of a wire at any frequeney {o internal inductance at zero frequency is, 4 [os ime bei! me + ber mr ber’ mr | me (bei mir)® F (her? mr) ‘The ratio approaches unity as the frequency approaches zero and sub- ame cael of partial flux linkages in deriving Tia G8) ae frequency increas the mtlo becomes salle, foram fect causes th current to er toward the surface of the wire and thereby reduces the number of inter- nal flux linkages. Tabulated values of the skin-effect inductance ratio are available in the paper by Rosa ‘and Grover cited in footnote 6 of this ‘hapten Theva ate sloted in Fig. 4.5. Internal inductance caleu- ie eae cs ese should be corrected by the ratio read from Bg These GMD (GN) of by Ba, (230) or aml equations, the waive ned for Dt may be al juste to accoun fur kin eet T this is done the frequeney for which the value of D, is listed must be specified. In Tables A.1 and A.2, ‘Fic. 4.5 Ratio of actualinternalindue- for instance, the values listed for self eet ine oe esa ee oy net ates deem agenda, wed in Bq. 2.57 give induggnce ‘the periphery. The ratio is plotted as values which clude ine = 0.0636 nfs ft spacing listed ‘per mile, account at thy rae 48 Other Losses. The high electric field it (4.56) Joss in the trans. Lt VERSION ADDS NORESISTANCE AND SKIN EFFECT as mission of power, ‘The high voltage gradient at the surface of a wi sometimes aceclerates electrons in the air sufficiently to ionize air mole cules by collision. If the voltage gradient at the wire exceeds a certain critical value, the process of ionization becomes cumulative and results in appreciable loss of energy. ‘The ionization is characterized by & faint glow surrounding the wire and is called corona. ‘The critical voltage depends on wire size and spacing and on atmospheric conditions. Corona is most likely to occur when the diameter of the conductor is small compared to tho distance between wires. High voltage, small wires, and close spacing contribute to a high voltage gradient which may induce corona. Damp weather increases the loss from corona, and rough ot dirty surface on a conductor increases the probability of the ‘occurrence of corona. Empirical methods for the calculation of corona loss are available in the literature.’ When a line is designed, the effects of corona are considered, and the design is modified, if necessary, to reduce corona loss to a minimum, usually below 2 kw/mile for a three-phase line under normal conditions. Radio influence due to corona must also be con- sidered and may be more important than line losses. As discussed in See. 1.2, tests on the 500-ky Tidd experimental line included an extensive investigation of corona as a source of loss and as a radio influence factor.* Another loss oceurring on transmission lines is caused by the leakage of current at the insulators which support the lines at the towers. Tt differs from leakage through the insulation of cables because it is lumped at the insulators and not uniformly distributed along the line. Even so it would be computed as though uniformly distributed by representing it as. conductance, but in overhead lines the leakage is negligible. Since leakage at insulators of overhead lines is negligible and corona loss is “The empirical equations for corons loss considered most accurate are those given by Peterson and by Carroll and Rockwell, See W. 8. Peterson, diseussion of the paper by J. 8. Carroll and B. Cozzena, “Corona Loss Measurements for the Design Of Transmission Lines (o Operate at Voltages between 220 Ky. and 330 Kv.” Trans. ATER, vol. 62, pp. 62-63, March 1933; J. 8, Carroll and M. M, Rockwell,“ Fmpiriea Method of Calculating Corona Loss from High-voltage Transmission Lines," Trams AIEE, vol. 56, pp. 558-565, May, 1987. * Seo I, W, Gross, C. F. Wagner, O. Naef, and RL. Tremaine, “Corona Invg tion on Extra-high-voltage Lines 600k Test Projeet of the Ameri Fleetric Company,” Trane, ALEE, vol. 70, pp. 75-1, 1951; ee G Pakala, 8. C. Barllett, and C. D. Pahenkopf, “Radio Inf Laborstory—300-kv Test Project of the American Gas Trans. AIBB, vol. 70, pp. 251-264, 1951. For the minizna having tolerable radio interference at the higher trensmis Rorden and R. 8. Gens, “Investigation of Radio Noise of High-voltage ‘Transmission Lines,” Trans. ATEB, va 1052, ‘ a Y fe ren “aN VERSION ADDS NO WA ena & J we rint-driNgrs ELEMENTS OF POWER SYSTEM ANALYSIS usually small in a line which is properly designed, the conductance between conductors of an overhead line is assumed to be zero. mission line there is a nonuniformity of current distribution in addition to that caused by skin effect. In a two-wire line, slightly fewer lines of flux link the elements nearest each other on opposite sides of the fine than link the elements farther apart. ‘Therefore elements in the near sides have lower inductance than elements on the far sides ‘The result is « higher current density in the elements of adjacent con duetors nearest each other than in the elements farther apart. ‘The effective resistance is inereased by the nonuniformity of current distri- bution. The phenomenon is known as proximity effect. ‘The increase in resistanee depends on the frequency, distance between conductors, conductor size, and permeability. Proximity effect is present for three phase as well as single-phase cireuits, Even at very high frequencies if the ratio of spacing between wites to the radius of the wires of a fwo- wire line is greater than 15 to 1, the increase of resistance due to proximity effect is only 19. For the usual spacing of overhead lines at 60 eyeles, the proximity effect is much less than the probable error in determining, the resistance and is neglected. PROBLEMS 4.1 Compnte ber 1.8 and bei 1. 2 Compute ber’ 18 and bei’ 1, 43 Caleulate 2/Mte and L,/Lie for mr = 18, 44 Find the d-e rusistance per mile fora wire having mr = 1.8at 60 eps 46 Compute the GO-cyele resistance at 50°C and the b-eyele inductive reactance at Lft spseing for round heed-drawn copper wice having a erosescetionsl area. o £800,000 rireular mils, Compare the result with the values given in Table A. for the Bz-strand copper conductor of the same area. Explain the reasons for any dis: crepancy. Use Figs, 44 and 45, 146 Specify the two steclrwinforved aluraininm exbles which have approximately the aume G-eyele resistance af 25°C aa & 10esteand, 300,000-cireular-nil hard-drawn, copper conductor. » Would the copper conductor oF one of the ACSR conductors be expected to have the lowest rorona loss for the same spacing between conductors? Give reasons 47 Calculate the d-e resistance per mile at 25 and 50°C for a solid 400,000- tireular-mil harddrawn copper condnctor. Cotapare the result with the d-c ggait ance listed in Table A.1 for a strand eonductor, and explain any diso ‘48 Calculate the O0-eyele resistance por mile at 25°C for a say til handcalrawn copper conductor. — Acconint for the difference b the value listed in Table A.L for a Westrand conductor, Dae resistance for the H-strand conductor agree with the val the det resistance of the solid conductor hy 2% and ratio sed in determining the O0-eyele resistance of the VERSION ADDS NO WATERMARK & ! Din por unit = D rosh 4 Upon substituting the expression of Eq. (6.22) for yf in Eqs. (6.27) and 2.06 X 10-3, we obtain for the constants in per unit letting Werle of |Z. 4 = cost (2.08 x 10-4 p — [inn (206 x 10-1 B= [sim (200 x 104 J f © = [sion (2.06 x 10-4 Dea (6.28) ‘The angle ¢ associated with the characteristic impedance is also a function of R/X, since (6.20) (6.30) and With ¢ expre ed in terms of H/X, the ABCD cong given in Eqs. (6.28) are functions only of the ratio # nf the line |. The curves plotted in Figs, 6.7 to parameter and show the variation of the ABCD length. Since the constants are complex, the 1 ponents are plotted separately, and the constat EDD ren x nN [= REGISTERED 9) VERSION ADDS NO expressed in per unit ofELEMENTS OF POWER SYSTEM ANALYSIS Am Art jae B= By +iBs (6.32) C=O Fie Sometimes the constants may be useful in their persunit form. At other times it may be expedient to convert them to the form required in rooop ET TTT TTT) PANETT ‘ ‘Ay inper unit L L 0 50 109 150 200 250 300 «350 400 Line tength in miles Fie. 6.7. The transmission-line Ay constant as a function of Kine length in miles (From D. J. Povejsl and A. A Johnson, “A Bersinit Interpretat ‘Transmission Line Constants,” Trans, ATEE, vol. 70, pp. 194-200, 1951, by pera equations involving volts and amperes. No conversioggis the constant A. The B and constants are converted which is the inverse of that used to obtain Eqs. (@ Bin ohms = |Z. X Bin pg Cin mhos = © Ber unit ADDS NO WATERMARK < Crint-ariGENERALIZED CIRCUIT CONSTANTS, Ww on os |— + ¥ ooel tees a oe el 3 Ze rr | ye iy aa ae Bae Line feng in niles Pio. &8 The transmissionline 4 constant a a function of line length ia miles, From D. J. Povejsil and A. A. Johnson, A Por-unit Interpretation of Transmission Line Constants,” Trans. ATBB, vol. 70, pp. 194-200, 1951, by permission.) BOEBR : —Lpxles B, in per 50 100 180 200 +250 «800 ay ine length in miles ansmissioncline 2 constant in per unit of 12 d ‘Povejsl and A. A. Johnson," A J ine Constanta,” Trans. ATE, vol. Fro, 69 1 Feneth in mile tion of ‘Transt permission.) VERSION ADDS NO WATERMARK &, & <5 we rint-driNgne ELEMENTS OF FOWER SYSTEM ANALYSIS. aco ; 0.700 —— t | osc} +} 0500 |— > = TT zg | io #000 | a 7 “| © 0300 7 co i rT 20 it ox00}—+ | | r | 1 30100 —1s0_ 700 —2s0- 300 380 00 Line lth in ies Fis, 6.10 The transmisiontne MB constant in per unt of {| aa fonction of Teneth in mls, Pom DJ. Fora and Ac dohnsn, “A Pera Tepe tof Tratamason Line Constance” Pons AVE, va, py. 104-200, 13, by pernisbn) =0038 0 80~«100—«ts0.—< 200-280 Line fength in miles Fic, 6.11 The transmission-ine C, constant in per wnit of length in miles. (Prom D. J, Povejsil and A. A. dohns tion of Transmission Line Constants,” Prans, ALEE, vo permission.) Y & Gvtion of ine “TN SIREGIS TERED 9} VERSION ADDS NOGENERALIZED CIRCUIT CONSTANTS. ue 080 070 060 050 040 Ce in per unit ofl 0.30] 020 a0 a | | al 080 TO 159. “Zoo 7503008 OO tine ecg in mies o tranamitsontine Cz constant in per unit of 2 a» faneton of tine Fic. 612 length in miles, (From D, J. Povejsl and A. A. Johnson, “A Per-unit Interpreta~ {don of Transmission Line Constants,” Trans, ATBE, vol. 70, pp. 104-200, 1951, by pertnission, If the product of L and C is the same for all lines, the magnitude of the characteristic impedance of any line ean be read from a chart, for if we assume that Vote 2.06 X 10-* (6.34) we ean solve for aC’ to obtain 4.24 X 10-6 whe Upon substituting this value of #( in Eq. (6.29), we ob REGISTERED 9) VERSION ADDS NO120 ELEMENTS OF POWER SYSTEM ANALYSIS Equation (6.37) shows that Z, can be expressed in terms of the inductive reactance per mile and the ratio 500 Te of the resistance to the inductive |_| |_| 08 | reactance. Figure 6.13 provides curves from which [Z| can be read. ‘The method of obtaining the ABCD constants from the charts is as follows: 1. Calculate, or obtain from tables, R and X in ohms per mile. | 2, Caleulate the ratio #/ | 3. Enter Figs. 6.7 to 6.12 with | the line length and the ratio R/X, and read the real and quadrature 2m, + components of each constant in per oz cm 06076868 4. If Band Cin ohms and mhos Surge impedance [Z,lhs § Inductive eadancechmiinile gre desired. Fig 614 Tho magnitude ofthe carne (a) Read the magnitude of Z. mnision ince as a fanction of fine inde from Fig. 6.13, or caleu- tive resclance per mile. (tom D. d late Za Povejal and A. Johnson, "A Perunit b) Obtain the required values Tnustetaton’ cf “Rusanieton Line (© Obtain the requ Const Trans, AIEE, vol. 70, pp. bl ane ane 191-200, 195, by permission.) (6.33) Example 6.1 Obtain the ABCD constants of the line sending-end quantities found in Example { Example 5.1, and check the 2, Solution From Example 5.1 R = 0.172 ohm ‘mile X = 0.824 ohm/mile Then As = 0.89 4, = 0.022 BI VERSION ADDS NO WATERMARKGENERALIZED CIRCUT CONSTANTS a and upon combining the real and quadrature components we have As + jade = 0.895 + 70.022 = 0.895/1.4° By + jBs = 0.089 + 70.440 = 0.450/78.0° per unit ©, 45C: = 0.0085 + j0.451 = 0.151/90.47° per unit A a D The constants B and C when expressed in per unit are equal in magnitude but not in angle. ‘The difference between the magnitudes of B and Cin per unit here is the result of variables which enter in reading the charts and in manipulating the slide rule. The magnitude of the characteristic impedance ean be read from Fig. 6,13 or taken from the solution of Example 5.1, From the example 2. = 405/—5.9° ohms Then from Eqs. (6.33) B = 405 X 0.450/78.6° 82.5,/78.6° ohms O51 g,47° = 9111/9047? mo Ve 200/0° volts and Tx = 361/0° amp Vs = 0.896/14° X 115,200 + 182.5/78.6° x 361/0° 103,000/1.4° + 65,800/78.6° e000 32.520 + 13,000 j04,500 and ry = 0000111/8049° x 115,200 + 0895/14 x 361 -1 = 39 WAT? + + j136 = 350, [LA = ~ 1.08 + j128 + 323 + 2.0" amp At the sending end Line current = 360 amp Line voltage = V3 X 134 = 232 kv Power factor = cas (30° — 22.9°) = 0.99 V3 x 232 X 350 X 0.9928 =, REVERED YN oP 45 [REGISTERED ©) VERSION ADDS NO WATERMARK Example 6.2 Find the equivalent-r circuit for the line af from the ABCD constants obtained from the ch]12 ELEMENTS OF POWER SYSTEM ANALYSIS Solution From Eqs. (6.19) and the constants found in Example 6.1 = B = 1825/78.6° ohms ¥ _ A=1 _ 0.895 + 70.022 ~ 1 _ -0.105 — Ter 778.0° 182.5778. = 0,000588/89.6° mho 6.5 Constants of Combined Networks. When a power system con sists of series and parallel combinations of circuits whose ABCD con- stants are known, it is convenient to find the constants of the circuit which is equivalent to the several component networks combined. This is a form of network simplifieation, In some cases it may consist of the inclusion of the characteristics of the transformers at the terminals of a Js ly In =F As = a % Ae Ba Co Dy ve As ByCyDs | Ve if eens mel Fra, 6.14 Two four-terminal networks in serie. line in the ABCD constants of the line itself in order to analyze the over-all, ‘operation of the line with its terminating transformers. At other times it, may be desirable to know the constants of a circuit equivalent to two Jong lines of different characteristies but operating electrically in parallel Consider two circuits in series as shown in Fig. 6.14. ‘The two net~ works can be combined into a single equivalent network by writing equations for each one separately and eliminating the voltage and cur- rent at the junction of the two networks. For cireuit a of Fig. 6.14, the voltage and current at the junction are Vx = D.Vs ~ Bels (6.38) Iz = —CiVe + Aals (6.39) and for eireuit b the same voltage and current are 4Ve + Bile Va + Dida Bliminating Vx from Eqs. (6.38) and (6.40) and 7} (6.1), we have YU 2.0) and AN REGISTERED ©, (63) Lt VERSION ADDS NO WATERMARK D,Vs — Bele = =C.Ve + Aals = CVs 4(GENERALIZED CIRCUIT CONSTANTS wa Multiplying Eq. (6.42) by A, and Eq. (6.43) by B, and adding the result ing equations, we obtain (AsDy — BC)Vs = (Aude + BiCi)Vn + (AoBe + ByDi)Ie (644) ‘Multiplying Eq. (6.42) by C. and Eq, (6.43) by Dy and adding the result~ ing equations, we obtain Da ~ Bla) = (Ais + CyDa)Vn + (BiCa + DaDi)In (6.45) For the network equivalent to networks a and b in series, from Eqs. (6.44) and (6:45), since A.D. — BuCa = 1 1e = Aady + BCs B, = A.By + BaDy Cy = AvCa + CrDa (6.48) Dy = BCs + D.Dy HE network b is at the sending end and a is at the receiving end, subseripts and b must be interchanged in Eqs. (6. wfe=Dh. cned] | \. Ay By Cy Dy a Fie. 6.15 ‘Two fourterminal networks Fi, 6.16 Two equivalent-r eireuits in in parallel parallel. If two networks are connected in parallel as shown in Fig. 6.15, a convenient way to derive the 4 RCD constants of the resultant network is to consider their equivalent-r circuits, shown in Fig. 6.16. The param- eters of the resulting single equivalent-r circuit are obtained by adding the admittances which are in parallel and by finding the equivalent impedance of the two impedances in parallel in Fig. 6.16. ‘The param- eters of the resulting equivalent-r circuit in terms of the para the circuits equivalent to networks a and b are Ye= Ynot Yo Upon substituting in Kgs. (6.47) the appr opciones. gf retaereReD oN VERSION ADDS NO WATERMARK4 ELEMENTS OF FOWER SYSTEM ANALYSIS. of the » cireuits equivalent to networks a and b, we obtain De=1, De= 1 Bt Bs BoB Bot Be (68) When the values obtained in Eqs, (6.48) are substituted in Eqs. (6.17, the constants 44, By and D, of the resulting equivalent network are found to he (6.49) By ByD. + BaD But By Dem ‘The fourth constant C, may be found by substituting the above expres- sions for the other three constants in the relation A.D, — B.C. = 1 ‘Thus (Ag = As)(Dy = Da) Bet Be C= OOH (6.50) 8.6 Measurement of Generalized Circuit Constants. When a trans- mission line is being designed, the gencralized circuit constants must be determined by methods previously discussed, all of which depend on computing the parameters of the circuit, ‘The aceuruey of such eompu- tations depends on how closely the assumed data approach actual condi~ tions. If the line is already built, the ABCD constants can be measured by making a few simple tests on the line. In Sec. 6.1 the constants were shown to be ratios of either voltage or current at the sending end of a general fout-terminal network to voltage or current at the receiving end fof the network with the receiving end open or short-circuited. If the network is a transformer, generator, or some circuit having lumped parameters, measurements of voltage and current at both ends of the line ean be made, and the phase angles between sending- and reccjg end quantities can be determined. ‘Thus the ABCD const determined as indicated in See. 6.1. It is practicable; the magnitudes of the required voltages and cugreyy at both ends of a transmission Tine, but there is no the difference in phase angle between quantities Fine. Phase difference is important beca stants are comples, By measuring two imped] P determine fo ends of the FGISTERED SN VERSION ADDS NO WATERMARK &, & RED “4 S Op; $ Crint-drig[GENERALIZED CRCUIT CONSTANTS 135 transmission line, however, the ABCD constants ean be computed, The impedances to be measured are Zao = sending-end impedance with the reveiving end open-cirenited Zss = sending-end impedance with the receiving end short Zao = receiving-end impedance with the sending end open-citeuited Zns = receiving-end impedance with the sending end short-circuited reuited ‘The impedances measured from the sending end can be determined in terms of the ABCD constants from Eqs. (6.4) and (6.5). With Te = 0, the equations give (6.1) and with Vx = 0 Las (6.52) To find the impedances meusured from the receiving end, Fas. (6.8) and (6.9) must be modified by changing the signs of all the current terms The change in sign is necessary because, with the voltage applied at the receiving end rather than at the sending end, the direction of current flow assumed to be positive when measuring impedance is opposite to the direction shown in Figs. 6.2 and 6.4 to whieh Eqs. (6.8) and (6.9) apply The equations become Ve = DV s+ Ble (6.53) Te = CVst Als (6.54) From Bos. (6.58) and (6.54) with Is = 0 D sera] 0 7, -¥e 8B ten Ta ‘The expression for Zyo and Zs ean be determined by another method, If the positions of Ys and Ye are interchanged in the unsymmetrical-r cireuit shown in Fig. 6.5, the impedances measured at the sending end of the modified circuit are equal ta the receiving-end impeda; original unsymmetrical x. Equations (6.19) show that Yd terms of the ABCD constants are the same except that where D appears in the other. ‘Therefore the sui and of A for D in Eqs. (6.51) and (6.52) gives th ances of Eqs. (6.55) and (6.56), The values of the ABCD constants in term ances are found as follows: VERSION ADDS NO WATERMARK128 ELEMENTS. OF POWER SYSTEM ANALYSIS. (6.57) (6.58) (6.59) After A is computed by Kas. (6.59), the other constants may be found by Eqs. (6.51), (6.55), and (6.56). It is good practice to take enough addi- tional data to check the magnitude of each constant by the ratio of a sending-end quantity to a receiving-end quantity, as deseribed in See. 6.1 6.7 Advantages of Generalized Constants. While it may seem unnecessary to introduce generalized circuit constants into a discussion ‘of power systems, the advantages gained by their use and their wide ‘ucceptance by the power industry make the understanding of them essen- tial to the engineer. Often they result in more concise expressions for the equations relating voltages and currents, especially where hyperbolic functions are involved. ‘The greatest advantage to be gained is the increased generality of the expressions that are derived in terms of the ABCD constants. In the next chapter we shall discuss circle diagrams of transmission systems for which ABCD constants may be found. "The derivations will be made in terms of the ABCD constants of the equivalent circuit. The ABCD constants may apply to only one piece of apparatus such as a transformer, toa line alone, or to a line plus its terminating transformers and other apparatus. ‘The constants may also apply to any mumber of series and parallel combinations of lines with their terminal equipment, pro- vided the system resulting from these combinations has only one location for power entering the system and one location for power leaving the system in addition to points where power entering and leaving ean be simulated by fixed impedances without emfs and ean be included in the ABCD constants PROBLEMS reactance the src arm mere the soning en, 20 ohne ERED YA ‘the series arm nearest the receiving end, and 1,000 ohms of « Dance in the A” Sapte Mince ei nena he Mae S Fea nae are ea ree [S-REGISTERED ON VERSION ADDS NO WATERMARK &, & * < Crint-ari(GENERALIZED CRCUIT CONSTANTS w (6) Determine the rotio [Vs if the Tne it 75 miles long andl open at the reeeiy= ing end. {) Determine the rato [sir ifthe line is 200 miles long and open at the receiv ing end, 4. ‘The A and B constants of a three-phase transmission line are 0.96/1.0° and 100/80° ohms, respectively. If the line-torline voltages at the sending and receiving tends are both 110 ky and the phase angle between them ia 80°, find Zand the poxeer tnd power factor of the load, 8.5 Calculate the ABCD constants of the nominal-r and equivalent-r eiruits of ‘ronsmission line of Prob. 5.11. Do not rofer to Fig. 6.7 to 612. 8.6 Determine the series impedance and shunt admittances of the equivalent-r circuit ofthe line of Prob, 5.1 from the ABCD constants found in Prob. 6.5 6.7 Find the ABCD constants of the line of Prob. 5.11 by the charts of Figs 6.7 to 6.2, 68 A Gl-cyele three-phase transmission line has an inductive reactance of 08 cohm/mile, and the ratio of the resistance to the inductive reactance is 0.20. ‘The line 2130 miles long. Find the ABCD constants from the charts 69 The snding-end voltage of the line doseribed in Prob, 6.8 is 230 ev. F the open-cirenit voltage at the recciving end 6.410 Find the voltage regulation of the line of Prob. 6.8 for «load of 100,000 kw ata power factor of 0.8 lng in parallel with synchronous condensers of 100,000 ky asin that the load voltage is 210 ky for this condition, and that the sending-end voltage is held constant at the value requited to maintain 210 ky at the receiving end for the load deseribed Git Find the ABCD constants of the fouscterminal network resulting when a resistance of 10 ohms is connected in eerice at the sending end of the four-terminal id network whose constants are as follow: 1 = 0.96 /0° 1B = 40,070" obs = 0.002045 90" rho D = 0920" 6.12 Find the series impedance and shunt admittances of the equivalent-r cireuit whose ABCD constants are given in Prob. 6.11 6.13 Measurements on a four-terminal network yield the following value Zsp = eo = 20 chins, pure resistance; Zss — Zes = 5 ohins, pure resistance, Find the ACD constants of the network and the parameters ofits equivalent 6.44 Find the ABCD constants of the transmission line for whieh Zany = Zso = 1,415/—80.25° ohms and Zee = Zas = 119/68:95" ohms, Find the parameters of the equivalent x of the eieuit. ERED | “4 d REGISTERED 9) VERSION ADDS NO WATERMARK & (623) connected to the node o in the original star circuit+ When a number of generators are connected through a newror taining several nodes and the emf of each generator is known| of each ean be found by eliminating all the nodes in ti emt connect day every oer ent throu PP REGISTERED VERSION ADDS NO WATERMARK & very other emf throu m z See Richard H. Prozie, “Elementary Electric Cired Hill Book Company, Tne,, New York, 1945 , <0 Crint-drigv0 ELEMENTS OF FOWER SYSTEM ANALYSIS. potential between the two terminals of the impedance divided by the impedance. Example 8.5 Four busses labeled a, 5, ¢, and d are interconnected as shown by the one-tine diagram of Fig. 8.18. Generators connected to busses « and b supply a synchronous motor load at bus d. For purposes of analysis all the machines at any one bus are treated as a single machine and rep- resented by a single emf and series resetance, The reactance diagram, , Oo} tod 5 & ats 0.9 Flo, 8.14 Reactance diagram for Example 83 with reaetances specified in per unit, is shown in Fig. 8.14. Simplify the reactance diagram by eliminating the nodes at each bus andgcon- verting the resulting cireuit to a mesh to whose terminals are connected the emis of the machines. Solution ‘The successive steps in the reduction of the Fig. 8.15. ‘The node at b is eliminated by tram delta the Y-connected reactances from @, ¢, Figure 8.15q is thus obtained from Fig. 8.14. VERSION ADDS NOREPRESENTATION OF POWER SYSTEAS m $03 Xj0.3 +03 XJ01 4503 XJ01 0.18 oy 5 oe uni Wd joni ~ FSperunit j0.5 per unit 0.5 per unit, ji jas jose Ne Es @ Fra, 8:15 Suovessive steps in the network reduetion of the reactance diagram for Example 8 Combining the impedances in series between ¢ and a, between ¢ and f, and between d and g gives 1.0 430.1 = jd 10.9 + 70.05 = j0.95 0.9 + 70.1 = 1.0 and combining the parallel impedances in Fig, 8.15a bet{ between ¢ and d gives I> RE VERSION ADDS NO WATERMARK from which Fig. 8.15) is obtained, * < Crint-drigwa ELEMENTS OF FOWER SYSTEM ANALYSIS ‘Transformation of the A connecting points a, ¢, and d in Fig. 8.156 into its equivalent Y gives Fig. 8.15c, ‘The computations are 0.1875 X j1.5 = 0.15, __g0.1875 x 50.1875 FL + j0.1875 + 70.1873 Ba 0.0188 Figure 8.15d results from combining the series impedances of Fig, 8.15 The required mesh, in this ease a 4, is obtained by a final Y-A transforma tion, and the result is shown in Fig. 8.15e. ‘The computations are Boy = EBS X H.1O + 110 Xj1.02 + j1.02 XJ 377 soy oS j1.02 jor 7337 93.02 saa Example 8.6 If the intemal emfs in per unit at stations 1, 2, and 3 of Example 8.5 are By = 1.5/0°, By = 1.5/15°, and By = 1.5/—36.9°, find the per-unit power outputs from stations I and 2, and find the per-unit. power input to station 3. Solution ‘The currents in the A of Fig. 8.15¢ are === = +0.426 — 70,088 per unit 0.263 + 70.088 per unit ‘The currents at the terminals of the delta are From generator station 1 Te = Leg — Toe = 0.158 ~ 30.102 = 0.188/—82,8° per unit From generator station 2 Ty = Typ — Tey = 0.531 — 0.069 = 0. Into motor station 3: Te = Lia = Toe y “a f= "REGISTERED 9 VERSION ADDS NO WATERMARK 689 — j0.171 = 074 ‘Lhe required power values areREPRESENTATION OF POWER SYSTEMS vs Output from station 1 1.5 X 0.188 cos 32.8° = 0.287 per unit ion 2: Output from stat 1.5 X 0.536 eos (15° + 7.4°) = 0.748 per unit Input to load 1.5 X 0.710 cos (86.9 — 13.9°) = 0.985 per unit ‘Phe total three-phase power at each station is the product of the above per-unit values and the base three-phase kva. ‘The amount of work involved in reducing 9 complex network found ‘a typical power system to a minimum number of impedances is barely tii Liid aff afl aff of taf I 7 1 (a) (b) Fig, 816 Lincar network with n terminals at which emf may he applied. indicated by the simple examples above. Where more nodes are present ind where the general star-mesh equations are required because of the mmber of branches terminating on one node, the work is inereased eon siderably. If the impedances are not pure reactances, the complication of using complex numbers adds enormously to the work. Evenso, network reduction is much preferred to the solution of simultaneous equations which would otherwise be required in any analytical solution, time required for an analytical solution Jed to the development of calculating hoards, 8.8 Driving-point. ond Transfer Admittances. driving-point and transfer admittances is a. convenient networks having several points where power enters ang work. Figure 8.16a shows several emfs connect ated by the rectangular box. ‘The internal imped: whose emfs are shown are incorporated in the point admittance at any terminal is the ratio of The NE ini he machines / “eN PS REGISTERED ON VERSION ADDS NO74 ELEMENTS OF POWER SYSTEM ANALYSIS terminal to the short-cireuited, eee fetes nation aleurone ‘aio of the coment Toaing the follageepptol atthe other te mw a othr ems chore circuited, % va | "Ture 8.105 sho che method ; oc eet the in pt elt. fs : ., made as shown, the driving-point + tan | admittance at @ is Too/E», and the win Ta/By Mr the noite we REST Grutor, eliminated from tho netork 0 Pig, 8.16 by" its equivalent mesh. that the terminals are intereon- Moved ha mes the reuling roel! to that of Fig Sah inshown a Fig 8% Wee ha Yas Be Tug © Bul¥ou + Vor # ++ + Yan) (8.24) from which we conclude that the driving point admittance at any point is the sum of the admittances terminating at that point when the civeuit has been reduced to the simplest mesh. ‘The current leaving the cireuit at nis BY on fom which we conclude thatthe transfer admittance between two points the admittance betwoon the theo. paints aftr the ereuit has’ been reducod to the simplest mesh, If the driving-point admittances are determined at all terminals, and transfor adinittanees are determined between all point, the currents entering the network at the terminals can be found by applying the familiar superposition principle. ‘The superposition principlgtate the current resulting from several voltage sources in a lin is equal to the sum of the soparate currents resulting lone. with all other voltages. short-circuited. entering a point caused by each emf alone gives Ty = Ee¥au — EaYou ~ BY oo — Ty = EX — Baos — EeVin — + VERSION ADDS NO WATERMARK & o ¢ < Crint-ariREPRESENTATION OF POWER SYSTEMS ws and similar equations where Yeo Ya, + Yeu = driving-point admittances at points a, 6, ¢, 1m, respectively Yau: Yer Yan Vis = transfer admittances between points a and b, between points b and ¢ between points a and n, and between points b and in, respectively Example 87 Determine the current entering each terminal of the network of Example 8,5 for the emfs specified in Example 8.6 by the method of Ariving-point and transfer admittances. Compare the results with the values found for the same currents in Example 8.6, Solution ‘The network reduction carried ont in Example 8.5 enables us to ealeu- late the required driving-point and transfer admittances, as follows = 0.270 — 70.292 = —j0.562 per unit = 0.270 — 70.381 —j0.601 per unit 0.331 — j0.292 = ~j0.628 per unit = 0.270 per unit = = 70.3 per unit BO ieee Tae 7 FB The currents entering at terminals ¢, f, and g are Yn = ~ 30.292 per unit Tp = BW — EnV eg — Ex¥ qe = (1.5 + J0)(—j0.562) (145 + 70.888)(—j0.270) — (1.2 — j0.9)(—70.202) = + j0.392 — 0.105 + 0.350 + 0.263 = 0.158 — ji [yp = BV yp = BV ~ EsV pg = (145 + j0.388)(=30. ~ (5 + 70)(—J0.270) — (1,2 = j0.9)(—j.33 + 70.405 + j0.397 + 0.298 = d = B¥ee ~ BsV rg = (1.2 ~ 50.9)( — (1.5 + j0)(—j0.292) — (1.45 + 70.388) 0.560 + j0.438 + j0.480 ~ 0.129 = Gio per unit ERED | FS REGIS TERED © q FX + 90.177 per unit, VERSION ADDS NO WATERMARKcase together with an ammeter and voltmeter. A. (ERED sa ete SS Soak eS TERED VERSION ADDS NO WATERMARK &, & Op; $ Crint-drigREPRESENTATION OF POWER SYSTEMS 7 ported, the portable boards are not us accurate or as convenient to oper ate as the permanent installations with calibrated resistors ‘To study short-circuit currents on a d-e calculating board, a reactance diagram is prepared from a one-line diageam, such as that of the local generating system of an industrial plant shown in Big. 8.18. Of eourse, information must be available from which to obtain the reactances of the lines, the transformers, and the generators and motors, If the local system is connected to an external power system, as shown in Fig. 8.18, the external power system must be represented on the d-e board by a. connection to the d-e supply through a resistance determined by the ge To poner te 888 Fie, 8.18 Oneline diagram of a typical industrial power syst amount of curzent the power system would supply to a short cireuit at the point of conneetion. That the external power system is correctly represented by voltage and series impedance follows ftom the HelmholtzThévenin theorem.* ‘The theorem states that a linear network terminating on two points and b and containing any number of emis may be replaced by a single emf and a series impedance between a and b, ‘The emf is equal to the ppenceircuit voltage measured between a and b. The series impedance is the impedance of the network measured between @ and_b wi emfs short-circuited, If the emfs are constant, the imp open-circuit voltage between a and b divided by the ¢ short circuit applied hetween a and 6, Power compy xiving the expected short-circuit current at po systems, On a d-e board the impedances of the RED Cen eR “es Are considered 'S [=.REGISTERED 9) VERSION ADDS NO "See W. R. LePage, “Analysis of Alternating-curn MeGraye-Hill Book Company, Tne, New York, 1952we ELEMENTS OF POWER SYSTEM ANALYSIS to be inductive reactances or impedances having equal phase angles and, of course, must be represented by resistances. If the external power system is lange compared ¢o that of the industrial plant, disturbances within the plant do not affect the voltage at the point of connection. In such a ease the external power system is said to be an infinite bus and is represented by a constant voltage having no internal impedance. On the assumption that the external power system is an infinite bus, the reactance diagram of the industrial power system of Fig. 8.18 is shown in Fig. 8.19 with reactances of the components of the system assumed to have the values given there in per unit. on whatever base is selected. Reference bus Fro. $19 Reaetance diagram of the industrial power system of Fig. 8.18, React ances are matked in per unit Figure 8.20 shows how the system is set: up on a d-e ealeulating board. If 10 volts is selected to represent. one per-unit voltage on the board and 1,000 ohms is one per-unit impedance, 10 ma is the base current. All the internal ems of the machines of the plant system and the emf repre- senting the external power system are assumed to have the same per-unit value, and this per-unit voltage on the base used for the calculating board is connected between the positive and the negative busse ‘The interconnected resistances in the eireuit of the ealculating ‘one per-unit resistance equal to 1,000 ohms, are mark4 to correspond to the per-unit reactances of the wy three-phase short circuit occurs in a system if thr nected in Y to the three lines and if the impag to zero. Therefore, a three-phase short cird single-phase equivalent circuit by a short cirg Qeé then reduced © Sn REGISTERED ©) VERSION ADDS NO WATERMARK &, & * < Crint-ariREPRESENTATION OF POWER SYSTEMS v9 circuit to the common return. On a dec board this is aecomplished by connecting the negative bus to the junction between resistors correspond ing to the point at which the short circuit occurs. In Fig. 820 a three- phase short cireuit on bus © is simulated by closing switeh 8, which corresponds to connecting point C’ to the common return in Fig. 8.19. Until such a connection is made to the negative bus of the d-e board, 10 current flows in the circuit. Insertion of an ammeter in any branch of the circuit indicates the current in that branch due to the fault, ‘The current may be recorded in per unit ss read from the ammeter and con- verted later to amperes by multiplying by the base current, or the Postve A 3000 Negative Be Fic, 820 Cirvuit of a de caleulating hoard to represent the ayatem of Figs, 8.18 and 819, amperes flowing in the branches of the d-e board may be read and con- verted by a multiplying factor to the amperes which would flow in the actual system, Although the current. computed from the readings obtained from a d-c calculating board is due to the fault only and does not include the com- ponent of current due to londs on the system, the error due to the omission of load currents is not great. ‘The total current in any part of the system, during a fault is, of course, the sum of the components due to the loads and due to the fault. ‘The load component is small, however, compared to the fault current, and the load current and fault cur have a large difference in phase. Tt ean easily be seen of a small phasor and a large phasor having a large very nearly equal to the magnitude of the large total eurrent in a branch during a fault is very nea to the component due to the fault. current aloy An ae calculating board is much more accu de board because magnitude and phase adj D y\ lore, the & @ 5 “REGISTERED 9) “VERSION ADDS NO WATERMARK &, & fe < Crint-aiy180 ELEMENTS OF FOWER SYSTEM ANALYSIS every individual emf in the system and because all the circuit parameters can be set up on the board. 8.10 A-C Calculating Boards.? ‘The first modem type of a-e caleu- lating board was built jointly by the General Eleetrie Company and the Massachusetts Institute of Technology in 1929, This board oper- ‘es at 60 cycles, Since ite installation, many boards have been built.* Most of the ac boards operate at either 40 or 480 cycles, although some have been built for other frequencies, two of them for operation xt 10,000 eycles. Higher frequencies allow the use of components of smaller size. All moder a-e calculating boards operate in a similar manner and are composed of similar types of units. The biugest variations between boards result from different. designs of individual units and different metering arrangements. The board installed at Schenectady by the General Electric Company in 1049 is shown in Fig. 8.21 and will be described. ‘The network units composing the various circuit, elements of this board have adjusting dials ealibrated in per unit. Each unit terminates in a pair of cords with plugs. Two units are conneeted in series by plugging a cord from each into horizontally adjacent receptacles on a lange pane! board. The generator units have independent phase-angle and magnitude adjustments, ‘They ean be connected at any point in the system. A voltmeter, watimeter, and varmeter are built into each generator unit ‘The series impedances of nes and associnted transformers are repre- sented by standard line units composed of resistance and inductive reactance. Capacitors are provided to represent shunt capacitance, and a nominal-r cireuit is made by connecting one capacitor at each end of a standard line unit, Unite to represent loads have their own Alternating-current ealeulating boards are called a-e network analyzers by the General Electric Company and ae netiork ealeulntora by the Westinghouse Electric Corporation, "Por a list of 40 ae caleulating hoards, 30 of them in the United States, se S. B. Crary, 1, W, Gross, and C, F, Wagner, “Progress and Future Trends in Electric Transmission,” Trans, ALBE, vol. 71, Part I, p. 968, 105% See for instance H, P, Kuehni and Tt. G. Lorraine, “A New A-C Network Ana- lyzer,” Trans. ATEE, vol. 57, pp. 67-73, 1958; and W. A. Morgan, F. S. Rothe, and J.J. Winsness, “An Improved A-C Network Analyzer,” Trans. ABE gol. ‘391-895, 1949, For a list of the frequencies and component parte of a Doards see E, T. B. Gross, “Symposium on Network Anal Remarks,” Proc. Am. Power Conference, vol. 14, pp. 381-888, ‘Technology, Chieago, 1052, See alzo P.O. Bobo, “ Handlin con an AcC Network Caleulator,” Blec. Zng., vol. 6 pp. E,W. Kimbark, J. H. Starr, and J. E. Van Nes, Network Analyzer,” Trans, ATEE, vol. 71, Part f, pp. 1 "A New Principle Is Employed for 60-Cyele A-C Networl vol. 71, Part 1, pp. 18-22, 1952 FS REGISTERED 9) VERSION ADDS NO WA ena & J we rint-driNgREPRESENTATION OF POWER SYSTEMS ro individual voltmeters and are connected to the system through auto- transformers in order to keep a constant voltage on the unit regardless of the system voltage. ‘This arrangement of the load units is helpful in adjusting the loads to precaleulated values. Additional units repre- sent synchronous condensers, autotransfarmers, and mutual reactance Push-button switches at the master-instrament and control panel connect any unit in the circuit, to the master-instrament busses. Master instruments have light-beam pointers with a short time response, Read- ings are made on S-in. scales at eye level in front of the operator. An ‘821 Viow of the front and one side of the General Hloctrie Network Analyzer No.2, installed at Schenectady, N.Y. (General Blecirie Company.) ammeter, voltmeter, and wattmeter-varmeter read magnitude and phase ‘angle as well as real and quadrature components of current, voltage, and power. The ammeter reads directly in per unit on any of six current scales, and the wattmeter-varmeter reads either in per unit or in meg watts and megavars on a 20, 50, oF 100-megavolt-amp base, ‘The real and reactive power readings are of the proper sign looking away fromthe bus being metered. ‘An impedance diagram with impedances specified ingger unit is followed in setting up the board. Operating obtained by adjustment of the magnitude and phas voltages of the generators, Nominal voltage of § and nominal current is 50 ma. ‘Table 8.1 gives FS neGIS TERED O ON VERSION ADDS NO of each type and their use, range, and size of ste WATERMARK &, & * < Crint-ari182 ‘Tams 81 Buacests oF Gesenae E Tstalled at ferme AC nectacly in 1949 ELEMENTS OF POWER SYSTEM ANALYSIS Awauraun No.2 Rating Nam- ; © | meee Nun Flements Ranger | Stepe en Poca | volts [amperes ‘om V0-2.5 [Continuous] 2.5 | 10.0 @ 02180" | Continuous 12 | Synehronons impranees | 22 0-0.11 oor | 1.25 cores 3% XO-111 001 1H} | Line impedances Roos oor | 125 | 50 Geren tak) | 00.81 o-001 6¢ | Load unite Boot (Continuous) 125 erica or poral) | X0-16.05 | 0.08 100 | Copasitors oa oo) 135 “Gurceptanee) 4 | Large capacitors 030.0 1.0 1.25, (euseeptanee) 15 | Mutual transformers 1:1 ratio 1,25 | 5.0 Tn per anit unless otherwise noted 4} Of which 30 can be made into» lines hy eonneoting a eapacitor at each end, ‘Hcluling continuously variable autotransformers for load adjustment, ‘The a-e calculating boards built by the Westinghouse Electrie Corpors~ tion operate at 440 cycles, and impedances are marked in ohms. Modern Westinghouse boards contain load units connected as x circuits to repre- sent the nominal-r circuits of transmission lines. Only one setting need be made to adjust the capacitors at both ends of the x line, The board built for Commonwealth Edison Company in 1951 has, in addition to other line units and capacitors, 24 lines, two of which 9 Fig. 8.22. Many modern boards are equipped with a recordin plastie recording surface consisting of a lurge n spots, The one-line diagram is placed on the small lamp bulb corresponding, to each metercg spot at the point on the diagram represented DN ings are taken for a particular unit on the bog 4STERED ot of light shows VERSION ADDS NO WATERMARK &, & Op; $ Crint-drigREPRESENTATION OF POWER SYSTEMS tea through the paper at the point on the diagram at which data for that unit is to be recorded. So many spots are available for lamps that the lamp connected to any unit may be put in almost any position under the diagram, A calculating board built. by the Detroit Edison Company for its own use has a fixed, permanently connected section representing the bulk 822 View of two wline units of a Westinghouse Network Calculator, Resist= tnd reactance values marked in ohms are equal to per-cent impedance, since 100 ohms is the base impedance. Toxgle switehes marked in per cent susceptonce insert the indicated values in each shunt arm simultancously. (Westinghouse Bletri¢ Corporation.) power system of the company. Completely variable units are also avail- able on this board.1® ‘The calculating board is a great time saver compared with algebraic methods of solution of power networks. It is particularly advantageous in the study of the effects of changes in a system. Once a board igget up it takes but a few moments to determine the cireuit loadi voltages which occur upon the temporary loss of part of t conditions of contemplated future expansion, with the af FSi Tight and [s>oREGISTERED VERSION ADDS NO WATERMARK &, & ™ Aids in the operation of ealculating boars are desert “Auxiliary Equipment Facilitates Network Caleulator Opes Power, vol. 24, pp. 50-54, October, 1946, "See E. A. Baldini and A. P. Fugill, “A Power Sys Computer,” Trane, ATEE, vol. 71, Part 111, pp, 291-204 * < Crint-drig