Transmission and DistibutianBreetic systems lechnology Jostitute Electrical Transmission and Distribution Reference Book A DE ABB Power T&D Company Inc. DEDPreface to the fifth edition Fifiy-five years ago. the Central Station Engineering Group of Westinghouse Electric Company first published a book focused on the practical application of electrical engineering to the transportation and delivery of electric power. The Electrical Transmission and Distribution Reference Book proved to be a simple, practical, and useful reference book for electric utility engineers as well as electrical equipment designers. Three generations of power engineers have used what has become popularly known as the “T&D Book” both as a core technical reference and as a tutorial on the finer points of power delivery system design and operation, In the five and one half decades between its original publication and this latest edition, the T&D Book was revised and expanded three times. In many ways, the T&D book's growth and evolution mirrored that of the electric power industry itself. The original book focused almost exclusively on transmission systems, addressing the higher voltages and longer lines then becoming common, as well as the rapidly growing complexity of transmission systems, particularly due to interconnection of individual electric utility systems into large power grids, The long-term trend, however, was toward an increasing focus on distribution, that portion of the T&D. system nearest the customer. Subsequent revisions of the T&D book added sections on power distribution systems, primary and secondary network design, capacitor application, and voltage flicker. This latest revision continues the trend of increasing attention to the levels of the T&D system nearest the customer. Chapter 24, Characteristics of Distribution Loads, focuses on consumer load requirements and how they interact with distribution system economy and reliability. It presents detailed guidelines and design methods to identify the behavior of electric load on the distribution system, and to address it with respect to the “two Qs” ~ quantity and quality - that consumers of electric power have come to expect the T&D system will provide,A more recent and accelerating trend in the power industry is the growth through mergers and acquisitions of both power companies and equipment suppliers into international companies operating on a worldwide scale. Several enormous power companies operate large power grids on three or more continent More directly associated with this book, what was once the Central Station Engineering Group of Westinghouse Electric Company has been absorbed imto ABB ELECTRIC SYSTEMS TECHNOLOGY INSTITUTE, part of a global company with technical and business resources vastly beyond anything the original authors of the TRD book could have envisioned. ABB-ETI continues to maintain the traditional, practical focus of the T&D book's creators, but has added research and development activities focused on meeting the needs of the 21" century with new equipment, designs, and technology. This latest revision does more than just talk about new technologies. Electrical Transmission and Distribution Reference Book, fifth edition, is available in the traditional printed format as well as on computerized CD-ROM. The new format expands the book’s usefulness as a resource for modern power engineers The material presented here is the result of research, investigation and practical application by many engineers and scientists, including cooperative studies with electric utilities conductor and cable manufacturers, communications companies and industrial power users. It is not feasible to list here all of the names of the companies and individuals who have contributed to the body of knowledge covered in this book. — These acknowledgements are given in the individual chapters. The authors gratefully acknowledge the hearty cooperation of all those who worked to produce this book. in particular, we wish to thank Ms. Kathy Hendricks, who tirelessly assisted in the preparation, editing, and formatting of this fifth edition Enrique Santacana Vice-President and Director October 1, 1997CHAPTER | 10 Contents Original Awshor m and Revising Author General Considerations of Transmission CA. Powel @ CA, Powel Symmetrical Components LE, Hobson © D. L Whitehead Characteristics of Aerial Lines . Sherwin H. Wright and C. F Hall © D.#. Shankle and R. L. Tremaine Electrical Characteristics of Cables HN. Mulien de @J. 8. Williams Power Transformers and React JE, Hobson and R. L, Witcke R, L. Wh and 4S. Williams Machine Characteristics CF Wagner C.F Wagner Excitation Systems JE. Barkle, J Application of Capacitors to Power Systems . . ALA. Johnson Regulation and Losses of Transmission Lines G.D. McCann @ RF Lawrence Steady-State Performance of Systems Including Methods of Network Solution : . E. Harder © E.L. Hades Relay and Circuit Breaker Application EL Harder and J. C. Cunningham %© EL, Hariler and J. C. Cumingham Power-Line Carrier Application RC. Cheek nd Power-Systei Stability—Basic Application veces eee neeeeeee RD. Evans dnd HN. Maller Jn MJ. E Burke, 1 and R, L. remain jements of Theory Power-System Voltages and Currents During Abnormal Conditions RL Wirske W RL. Wiche page page page page + page page page + page + page + page page » page page page 64 96, 233 265 496CHAPTER 15 20 24 Original Author ® and Revising Author Wave Propagation on Transmission Lines C.F Wagner and G. D. MeCann ® CF Wagnee Lightning Phenomena . . CF Wagner and G. D. McCann WC Wagner and 4M. Claston Line Design Based on Direct Strokes. AC Monteith ® E, Le Harder and J. MM. Claston Insulation Coordination coves ALC Monteith and H.R Vaughn @ A.A Johnsen Grounding of Power-System Neutrals S.B. Griscon @ SB. Griscosn Distribution Systems Jobin S, Parsons and H. G. Barnett Joh 8. Parsons and H. G. Barnet Primary and Secondary Network Distribution Systems Joti 8. Parvons and HG: Barnet tah S. Parsous unud HG. Barnet Lamp Flicker on Power Systems SB, Griscom WS. B, Griscom Coordination of Power and Communication Systems RD, Evans @ RL Witske Characteristics of Distribution Loads HL. Wills Appendix Index page page page page page page page page page + page page page 610 643 666 689 719 741 784 809 838CHAPTER 1 GENERAL CONSIDERATIONS OF TRANSMISSI Original Author: G. A. Powel HROUGH discovery, iavention, and engineering ap- I plication, the engineer has made electricity of eon~ ‘inuslly greater use to mankind. ‘The invertion of the dynamo frit made engine power many times more fieetive in relieving the toil and increasing the opportu- nitics and comforts not only of industry but also of the home, Its scope, however, was limited to relatively short, distances from the power station beeause of the low voltage of the distribution citeuils. This limitation, for economic, reasons, kept the general use of electrieity confined to city, areas «hore a number of eustomers could be served from the same power station, ‘The next step in the development of the present-day eleetrie systems was the invention of the transtormer. ‘his invention was revolutionary in its elfect tn the electric industry because it made high voltage and Jong transmission distances possible, thus placing the engine power, through the medium of the alternating- curzent generator, at the doorstep of prietically everyone, ‘The first alternating current system in America using ransfortners was put in operation at Great Sarrington in Massachusetts in 1886, Mr. William Stanley, Westinge house electrical expert who was responsible for the installa- tion, ives an secount of the plant, part of which reads: “Before leaving Pittsturgh I designed several induction col, cor tiancformers as we now eall them, for paral conection ‘The original wae designed in the early aumracr of 1885 and wound for 5iX) volts primary and 1D) volts secondary emf. Several ather eos were constructed for eaperimental purses, “Abtho north end ofthe vllege oF Groat Facringtan eras aa old deserted rather mill which I lensed for a tiling suin and ervcted init 25 hp boiler and engine that T purchased for the purpose. ‘After wht scemed an interminable delay Tat last installed the Siemens altemator that Me. Westinghoise lise imported from London. Tt was wou Yo furnish 12 amperes of eutreat sth & rmasitaim of 600 valts. Tp the meantime {had starter the cox. struction of x number of traasformets in the laboratory and ene goged.a young man to canvass the tow of Groat Barston for light oustomers. We built in all at Great Bartington 28 teens: formers, 10 of which ere tent to Pittsburgh ta be used in ‘demonstration plant between the Union Switch and Signs! Com pony's factory” and East Liberty. “We installed in the town plant at Great Barringcon tno 50> light and four 25-ight transformers, the romainder being use in the taboratory for experimental work. ‘Tho transformers in the village lit 13 stores, 2 hotels, 2 doctons offices, one lacber sho, fand the telephone and pst offers. ‘The length af the Tine frm the lahorstary to the center of te town, sus abot 4000 feet." Our eontralstation industry torlay is, for all practical purposes, entirely alternating current. It, ean, therefore, be said to have grown from the sinall beginning at Great * About to ales Revisad 8, GC. A. Powel Barrington to its present size involving as it does a capital- ination in the privately-owned power companies of come 17 billion dollars th an ansial revenue of 4 illion dollars, ‘The growth since the beginning of this century in ine stalled generating eapacity of all eleetrie power plants, Fig. $—(a) Gaulard and Gibbs transformer for which George Weselnghouse fd secured all rights in the United States. First transformer designed by Wliliam Stanley. The proto: {ype of all transformers since built, ie definitely established the commercial feanibility of che altermating-currest syste a4 TAN. ‘contributing to the public supply has beem from about 14 rnillion kilowatts to 56 million kilowatts in 1948. Of this 56 million kilowatts the privately-owned utilities aecount— od for 41 million kilowatts and government-ovened utilities for 11 milion kilowatts divided equally botween the federal government anc local governments. Thus, $0 pereont of the generating eapacity of the countey is pri vately owned and 20 per cent government owned, With this 55 million kilowatts of generating capacity, 282 billion kilowatt-hours, divided 228 billion kilowati~ hours hy privately-owned generation and 5 billion public, ‘worn generated in 1948, ‘The average use of the installed capacity for the country as a whole was, therefore, 282 00 5120 hours, and 5130 country af 1 whole J ‘This capacity factor of 58.5 percent is generally con cereals being too high. It does not allow sufficient margin lu provide sdequiale spare eapaeity for maintenance and repairs. Fig. 2 illustrates how the spore snd reserve. eapiueity has shrunk in the past few years, A ratio of installed capacity to peak load of 1.15 to 1.20 is considered necosary to provide a safe margin for emergencies. Such capacity factor for the 8.5 pereent.General Considerations of Transmission Chapter 1 "The aversge cost of all electricity used for residential service has shown a steaay downward trend since 1925 from 7 eents per kilowatt-hour to 3 cents in 1048, “This is all the more remarkable ss sines 1989 all other items mak- ing up the cost-of-living index have shown increases range ing from 10 percent (for rents) to 121 pezcont (for food), the average increase of all items being 69 percent. The revenue from sales to residential customers accounts for ahout 36 percent of the total utility revenue; to large power customers about 29 percent to small light and power customers 27 percent, and to miscellaneous eus- tomers (railronds, street lighting, ete.) 8 1, Sources of Energy "The sourves of energy for large-scale generation of elec tricity are: 1, Steam, from (a) coal, (b) oil, or (e) natural gus 2, Water (hydro-electric) 3. Diesel power from oil Other possible sourees of energy are direct solar host, windpower, tidal power, shale oll, and atomic enongy, but ron of these aa yet has gone beyond the pilot-plant stage, for the reason that eaa! snd petroleum are still abtindantly available, But as fossil fuels become searcer and more expensive, there is every reason to believe that all of these, as vell as petroleum manufactured from vegetable matter, may become useful and economical supplementary sourves of energy. 2 a0 TTT] E EEL | 3 20 ++ + Z it | & Ll | 21 mai i HA t yy tl g i ttt z Ie xicowar- nouns iw a1LLioNs 3 SUTOTAL STALLED CaP MILLIONS OF RiLOWATTS 3 | /c:Sikoenen-comoewr pesos i tions 2 aA oPacowrrs MOGeeeeeen a a a a Co Fig. 2-Trend in production of electricity, tastalled eapactey, ‘and sum of peak demands, ‘5 margin in 1948 would have given a capacity-factor of, about 53 pervent, instead of 68.5 pereent. ‘Tavis 1—PaerERRED STANDARDS FoR, LARGE 8600-RPM 9-1 Steam TURED 60 ‘The estimated reserves of coal and lignite in the United States are about 3000 billion tons, This constitutes almost 99 percent of the mineral fuel energy reserves of the country; oil shale, petroleum and natural gos amounting to little more than 1 percent.» ‘Dy far the greater part of the electrio energy generated in this country is obtained from fuel, the 86 million kilo- PHASE 60-Cy Conpensing INE-GENERATORS TaeCoaledy TiydrogenCoaled Generar Generator] Rated for 05 Pig Hydrogen Pressure Turbinesgmnerator eating, kw T1500 | 75000 20000" 30000 doco sano ‘00 ‘Turbine expailty, kw 12030 33000 41000 3000 000) Goverator rating, iva 1820 3520047058 Toss 1oo382 ‘ower factor oss, 085 085 985 ass, Short-cieuit rato 08 os o8. os, 8 ‘Tarot pressare, pig 600 $0 [silorfsn\ star 120) 450 rato) ‘Throwde tempera, F = seo Yo00)."( 260) 00)" 380} 2000) toe) Rechoat temperature, F 1000 mnber of extraction openings og 3 3 5 5 Saturation temperatures at 1 5475. 1 1 15: pening st “turbine gear ‘nal 25S 235, 235 5 20 erator eating” with all ex 2 28288 Das, 2s 305 alo traction openings ia serv= se 380 380) 350 350 38 8T0) kee, F sal a0 0 10 40440 shane prose, inchee Hig abe 4s 13 43 115 Geer apy a 08 prorat and 15 peig hydrogen proseue, kva 27058 3058 BAIT sus 121764 Genet apy a 088 pone factor and 20 peig hydzogen pressure, va 132389 bine iit ti: Ue tor eames ey openea node ta ¢ eb ea acs woes Asnithe United States —a 2 enerator, 2000 kw, 2400 voles, Martiord Bleetrie Light Compan 1900, This turbine was about fou bul Hastiord, Connecticut, mes as large as any one before that time and cavsed auch comment the world vwatte of installed capacity being made up of approximately 38 million kilowatts of steam (urbines and one million kilowatts of diesel engines. Approximately 16 million kilo: ‘yattsof the installed eapacity are in hydvo-clectrie stations. Of the 282 billion kilowatt-hours generated by all means in 1948, roughly 200 billion eame from fuel; 76. percent from eoal, 14 pervent from natural gas, and 10 percent from oil. 2. Development of Steam Power ‘The modem steam-elecize station can be dated from the installation by the Hartford Flectrie Company in 1900 of 8 2000-Kw unit (Fig. 3) which at that time was a Tange machine, Progress in design and efficieney from then on has been eontinaous and rapid. Tn 1925 the publi utilities ‘consumed in their fuel-burning plants an average of 2 pounds of cosl (or coal equivalent) per kilowatt-hour, whereas today the corresponding figure is 1.3 pounds per Iilowatt-hour. ‘This average figure has not changed mate- tially in the last 10 years. It would appear that the coal consumption curve is approsching an asymptote and that 1 much better overall performance is not to be expected, even though the best base-load stations generate power for less then one pound of eval per kilowatt-hour. The vory high efficiency in the best basoload stations wed at a considerable increase in investment. Tt can not be economically carried over to the system as a whole for the reason that there must be some idle or partly idle capacity on the system to allow for peaks (seasonal and daily), cleaning, aljusiments, overhaul, and repaies. How hone can afford to spend for the improvement af stax tion efficiency above “normal” depends on the shape of the aystem load curve, the role of the station in that curve, and the oost of fuel Most of the credit for the improvement in steam con: sumption goo to the boiler and {urbine manufacturers who throuxh continuous betterment of designs and ma- terials have been able to mise steam prossuresand tempera tures. Between 1925 and 1942 the maximum throttle pres- sure was ruised from 1000 pst to 2100 pai and the average from 350 to 1000 psi. In the same period the throttle tema- perature was raised from 725 to 1000 dogrees I. and the General Considerations of Transmission. 3 : i te i Bf tech ne wt etd Fig. 4 Progress In turbine generator destan. average from 675 to 910 degrees, Genorator losses in the cantinne have been greatly reduced from about 6 percent in 1900 to2 pervent today, but these losses never di form, a large part of the total, and their influence on the overall, performance of the station has been minor. The increase in maximum size of 60-cyele, two-and four pole generating units over the years since 1900 is shown in Fig. 4. ‘The remarkuble inerease has been due to improved, materials and designs, particularly in large forgings, tm Line blading, and generator ventilation, In 15 the Ame! and the American Institute of Electrical Engineers adopted. standard ratings for tarbine-generatar units. ‘These were revised in November 1950 Lo include the 90.000 kw unit, and are Jisted in Table 1. ‘The snachines are designed to meet their rating with 0.5 psi hydrogen pressure, but ex perience has shown that between 0.5 and 16 pei the output, ‘of tho generator can be imereased one percent for each, pound increase in the gus prossuro without exeveding the temperatnre rise guarantee at atmospheric pressure. In many locations operation at more than 15 psi gas pressure4 General Considerations of Transmission may be diffcull because of codes regulating operation of “unfived pressure vessels” ut greater pressures, but serious consideration is being given to operation at 30 Ibs. Yor a hydrozen-air mixture to be explosive, the per contage-of hydrogen must lie between 5 and 75 percent. ‘The control equipment is designed to operate an alarm if the purity of the hydrogen drops below 99 percent. The density meter and alarm system is in principle a small constant-speed fan circulating a sample of the mixture. If the density varies, tho drop of pressure seross the fan, varios and rogisters on the meter. 43, Development of Water Power ‘The great transmission systems of this country reecived their impetus as a result of hydro-electric developments, Forty years ago conditions favored such developments, and in the early years of this eentury water-power plants costing $150 per kilowatt or less wore common. Steam, stations were relatively high in first cost and eoal consump tion per kilowatt hour was three times as much as to- day, and finally fuel oil was not readily available. As undeveloped water-power sites became economically less desirable, steam stations loss costly and their efficieney higher, and as oil fuel and natural gas became more gen- erally available through pipe lines, steam stations rapidly outgrow hydro-cleetrie stationa in number and eapacity. Today very few water-power sites ean be developed at such low east as to be competitive with steam stations in economie energy production, For this reason hydro electric developments of recent years have almost all been undertaken by Government agencies, which are in a position to include in the projects other considerations, such as, navigation, flood control, irrigation, conserva tion of resources, giving them great sociaf value. AAs the water-power developments within easy reach of the lod centers were utilized arid it became necessary t0 reach to greater distanees for water power, only large developments could be considered, and stations of less than 100.000 kw beearne the exception rather than the rule, ss witness Conovsingo with 252.000 kw, Diablo with 135.000 kw, Fifteen Mile Falls with 140.000 kw, Osage with 200000 kw, and many others. Tho developments of recent years undertaken by vatious governruettt ges cies have reached gigantie proportions, as for example Hoover Dam with 1000000 and Grand Coulee with 2.000 000 kw installed eapacity. A natural corollary to the inerease in station capacity hhas been a gradual increase in the size of the individual generator unite, the growth of which is shown in Fig, 5, culminating in the Grand Coulee generstors of 120 000 ‘kw at 120 rpm with an overall diameter of 45 feet, Most of the multi-purpose hydraulic developments call for large, slow-speed machines. For such eonditions yerti- eal units are used to obtain maximum energy from the water passing through the turbine. ‘The rotating parts are supporied hy a thrust bearing whieh is an integral part of the generator. ‘Two general types of generator design are used as dis- tinguished by the arrangement. of the guide and thrust, bearings. Where the axial length of the generator is short in relation to ite diameter, the “umbrella? design Chapter 1 is preferred, in which a single combination guide and thrust bearing is located below the rotur (Fig. 1, Chapter 6). Where the avial length of the machines is too great an additional guide bearing must be provided, In this case the combination thrust and guide beating is usually located above the rotor and the aulditional guide bearing Delow the rotor. ‘Tho advantages of tho umbrella design are (a) reduetion in overhoad room to assemble and dismantle the unit during esection and overhaul, and (b) simplicity of the single bearing from the standpoint of cooling: and mini- T T wl ey Pe ane ° Cte “THOUSANS OF KVA, 20) TH year a —Ts0 ig. 5—Trond in maximum waterwhoel generator ratings. mum amount of piping. The design slso lends itself readily to a totally-onclosed recireulating system of ventilation, which keeps dirt ont of the machine and faciti- tates the use of fire-estinguishing equipment. It also re- duces heat, and noise in the power house. 4. Combination of Water and Steam Power ‘There are very few locations today where an important market can be supplied entitely from water power be- ‘cause of seasonal variations in river flow, but in most cases a saving will be realized from combining water power and steum. The saving results from the combina~ tion of low operating cost of water power plante with love investment, cost of steam stations. Moreover, hydro electric units in themselves have certain valuable ad- Vantages when used in combination with steam units. ‘They start more quickly than steam-driven units, pro- viding a bigh degree of standby readiness in emergency.Chapter 1 ‘They are sell adapted to maintenance of frequency, and also to providing watiless energy at times of low water los. And finally, hydre-pondlage ean he drawn upon to relieve steam plants of short-time peaks to save banking. extra boilers To what extent a water-power site can be developed ceonomically involves a thorough investigation of indi- vidual eases. An economic balance must be struck be- tween the steam and water power to give masioum economy. One might install enough gencrating capacity, to take care of the maximum flow of the river during a. short period. ‘The cost por kilowatt installed would be low but the use made of the equipment (capacity factor) ‘would also be love, Or one might put in only enougls gon- erating capacity to use tho minimum river flow. In this ease the eost of the development per kilowatt Installed, would be high, but the eapacity factor would be high 20 7 | 100 as soo 3 rE oH CENTS FER MLOWATT=WOUR a 8 | ‘outa oe 1 T | ‘CAPACITY FAGTOR-PERGENT 2 oe Fig. (Cost of eneray at various capacity factors of steam, lana hydro-electric plants. also. Obviously between these two extremes Ties an. optimum value, The ratio of installed water-power eae pacity to the peak load of the system that gives the mini= ‘mum annual cost of power supply has been refereed to as the “economic hydro ratio,” and it can be deter- mined without great difficulty for any partieulae get of eon ditions In a paper® presented before the American Society of Mechanieal Engineers, Irwin and Justin disoused in an, Interesting and graphieal manner the importance of ineremental costs on the economies of any proposed de- velopment. Fig. 6, taken from iheir papor, shows in Curve C the eapital east per kilowatt of installation for various eapacity factors. ‘The costs were segregated tn ftems that would be the same regardless of installation, Gand, ater rights, dams) ond those that vary with the amount of installation (power house, machinery, trans General Considerations of Transmission 5 tission). ‘The latter group in this particular study was about $70 per kilowatt. Curve A gives the (otal cost of energy per kilowatt hour for 4 mosiern steam plant costing $05 per kilowatt with fixed charges at 12 pereont and coal at $4 a Lon. Cree Ts gives the total enst of energy from the waters power phint having the eapital eost indieated in Curve C. ‘To obtain such a curve it is necessary to determine Ue amount of energy available at the varions eapaeity Fue tors, the assamption being mado that all hydro capacity, installed is fem capacity’, that is, that the system loud cean absorh all of the energy generated ‘Curve B shows the typically high cost of hydro-electric energy as compared with steam st high eapacity factors ‘and its low eost ab Tow capacity factors, 5. Transmission iability In a hydro-leetrie development. the transmission. be- comes a large factor of expense and in comparing such developments with equivalent steam plants, itis necessary 40 include the transmission as a charge against the hydro electric plant. Figures of cost published on the Hoover Dam-Lus Angeles 287-ky line indivate that this tras mission cost over 800 a kilowatt, and other Jines eon templated will probably show higher costs Under certain conditions it may be more eostly to trans- mit sleetrical energy over wires than to transport the equivalent fuel to the steam station. Tt has been shown? hat the cost of electrie transmission for optimum foal and voliages ean be expressed as a linear funtion of power and distance, as follows: 1 loud factor 61 Fig. 4A typical Anaconda Hollow Copper Conductor. panded” ACSR conductor. dyetor is known as “expanded” ACSR and is shown in Fig. 3. In Fig. 4 is shown a representative Anaconda Hollow Copper Conductor, It consists of a twisted copper “I”Fig. 5A typleal General Cable Type HH. Big. 6A typical Copperweld conds Copperweld-Copper conductors Fig. 7Typtc (1) Upper photograplitype ¥ (0) Lower photograph: Type F eam 98 a eore about which strands of copper wire are wound. The *T” beam is twisted in a dievetion opposite 0 that of the inner Inyer of strands, Another foria af hollaw copper © Fig. 5, Known as the General Cable per conductor, it ie made up of segmental sections of eop- per mortived ‘nto exch ether ta form a self-supporting hollow eylindor. Hollow eqpper conductors rest neactor is shown in HI hollow cop- ictorsof lange diameter for a piven arose scetinn of copper Corona tosses are therefore smaller. ‘This vonstraction also produees 9 reduction in skin offeet 6 well as inductance us compared with stranded cord A discussion of large diameter conductors and heir characteristics is given in, reference 1 Cppereld eonductors consist of different: mumbers of copper-coated steel strands, a typical canilactor beige illus trated in Fig. 6. Strength is provide by the core of stool and protection by the outer coating of copper. When bigh current-carrying capacities arc desiwed ae well as high tensile stronath, copper strands are used with Copperneld strands to farin Copperwoll-C'opper eouduet- Characteristics of Aerial Lines 33 ors as shown in Fig. 7, Different relationships between ing capeity, outside diameter, and tensile stngih ean be obtained Ly varying the number and s ‘of the Coppervveld and copper strands, Il, ELECTRICAL CHARACTERISTICS OF ‘AERIAL CONDUCTORS ‘The following diseussion is primarily eonecrned with the evelopment of electrical characteristics and constants of aerial eondhictors, particularly those required for analysis, of powersystem problems, The constants developed are particularly useful in the appliestion of the principles of hytunielsies! components to the solution of power-svatem, problems involving positive, negative, and zero-sequenee impedances of transmission’ and distribution lines. ‘The Dasie quantities needed are the positive, negative-, and zero-se\juonve resistances, induetive reactances and shunt capacitive reuctances of the various types of conductors and sce general equations showing how these quantities are used. 1, Positive- and Negative-Sequence Resistance ‘The resistance of an eerial conductor is affected by the three factors: vemperature, Hrequeney, current density. Practical formulas and methods will nov be given to tale nto aeonnt those factors, Temperature Effect on Resistance—The resistance fof expper aml aluminum eonduetors varies almost di- rectly wilh teraperatire, While this variation isnot strictly Tinwar for an extiemely wide range of temperatures, for practical purposes i can be considered Finear over the range of temperatures normally encountered ‘When the ce resistance of a conductor at a given tem= pporatuze is kuown and it is desined to find the dee resistaneo at some other temperature, the following general formula, ay bo used. Re M+ Ra" Mu ® where Rade resistance at any temperature ts dogree ©. Rade resistunce 9 iny other temperatare t, dagnee C. Af=a constant for any ane type of conductor material, inferred absolute sero lemperabure =©ZH1.5 forannealed 100 percent conductivity eapper, = 211.5 for hard drawa 97.8 percent conductivity pe. 228.1 for aluminum. ‘The ahove formula is useful for evaluating changes in Le resistance only, and cannot be used to give ae resist ance variations unless skin effect ean be neglected, For small conductor sizes the frequency has a negligible effect, esistanee in the dhe to 60-eyele range, ‘This is generally for conduetor sizes up to 2/0, he vuriutions of resistance with temperature are usually unimportant because the actual ambient temperature is indofinive as well as variable along a transmission line. Au illustration of percentage change in resistance is when tem peratare from winter to summer over a range of O dese C 10 40 degrees C (32 degrees F to 104 degrees F) in which ease copper resistane increases 17 pe ea.a Skin Effect in Straight Round Wires— The resist- ance of non-magnetie conductors varies not only with ten perature but also with froqueney. ‘This isdue toskin effect. Skin effect Je due to the current flowing noarer the outer surface of the conductor as a result of nonuniform flax, distribution in the eonductor. This increases the resistance of the conduetor by reducing the effective erass s the conduetor through which the current flows, ‘Phe conductor tables give the resistance at commerei Frequencies of 25, 50, and GO cycles the following formula should be used. r= Krug obms per mile where r ather frequencies 2 ¢ ae resistance at the desired frequency {eyeles por second). g reietaneo at any’ known temperature, = value given in Table 5 Tn Table f, K is given as s function of X, where X=.003508 [ah Jafrequency in eyoles per second. be pertneability=1.0 for non-magnetic mater da vesisiance of the conductor in chs per mile. (3) Table 5 (skin effect ible) is carried in the Bureau of Standards Bulletin No. 169 on pages 220-8, to values of X=100. To favilitate interpolation over a small range of the table, it is aecurate as well as convenient to plot a curve of the values of Avs. values of X. Combined Skin Effect and Temperature Effect on Resistance of Straight Round Wires—When both tomperature and ekin effect are eansidered in determining conductor resistance, the following procedure is followed. First calculate the d-e resistance at the new temperature using Eq.) ‘Then substitute this now value of dee re- sistance and the desired frequeney in the equation defining , Having ealeulated X, determine K from Table 5. ‘Then using Eq. (2), ealeulate the new a-c rosistanee re, using the hnew dee resistance for rz, and the value of KC obtained from Table 5. Effect of Current on Re: magnetic conductors varies with eustont magnitude as well as with the factors that affect non-magnetic conductors (temperature and frequency), ‘Current magnitude determines the flax and therefore the iron or magnetie losses inside magnetic conductors, presence of this additional factor complicates the determi- nation of resistance of magnetic conductors as well’'as any tabulation of such data. For these reasons the effect of current magnitude will not be analyzed in detail, How- ever, Fig. 8 gives the resistance of stuel conductors a a function of eurrent, and the tables on magnetie conductors such ax Copperweld-copper, Copporweld, and ACSR. cot: ductors inelude resistance tabulations at cio current carry ing levels to show this effect. ‘These tabulated resistances, are generally values obtained by tests Zero-Sequence Resistance—The zerorsecgience 1 sistance of aerial conductors is discussed in detail in the section on zero-sequence resistance and inductive reaet= ance given later in the chapter since the resistance and i Characteristics of Aerial Lines Chapter 3 tice REESE, Bo Zp Rene enSeeyety peat erp Inpre Ee wees Be a Fig, § Electrical Characteristics of St fol Ground Wites* ductive reactance presented to gero-sequence currents is, influenced by the distribution of the 2ere-sequence current, fn the earth return path, 2. Positive- and Negative: actance ‘equence Inductive Re- ‘To develop the positive- and negative-eequence induc tive reactance of three-phase aerial Lins it is iret necessary to develop few concepts that greatly simplify the problem. First, the total induetive reqetance of a conductor varry- ing curront vill be considered as the sum of two com ponent= “This ee as ben talon fom Symmetrical Cony by C.F, Wagner and TCD, 133. ent (a book) ne, Mera Hill Bou Compeny,Chapter 3 (2) The inductive reactance due to the flux within a rains of fonw foot from the conductor center, including the fox Inside the conductor @) Thr inductive reactance due to the fus ext rast of one foot and out to sone finite distance ‘This concapt was first given in Wagner and Evans book on Symmetrical Components? and was suggested hy W. A. Levis. T-can be shown most easily hy considering a two-0on- ductor single-phase circuit with the current flowing out in ‘oneconductorand returningin theother. [n Fig. 9such acir~ cuit isshown with only the lux produced by conductor | for simplicity. Conductor 2 also producce similar lines of flax ‘The classic inductance formula fora single round straight wire in the two-conduetor single phase circuit is: toa Kin De L=4+2 In abhenries per em. per conductor, 5 : P P tor. (4) where he=permeahility of conductor material, r=radius of conductor, Dy=distance between conductor 1 and conductor 2. Daz und r must be expressed in the same units for the above ‘equation to be valid. For practical purposes one foot is ‘used as the unit of length since most slistances between, aerial conductors are in fost. In cable eireuita, however, the distance between conductors is less than one foot a ‘he inch is a snore common unit (Gee Chap. 4). rom drain frag term wh 2 reprosents the flux and associated inductance between, circles of radius « and radius @ surrounding # eonduetoe carrying current. (See Fig. 10), Rewriting iq. (4) keeping in mind the significance of the sel te a2 abhenries per em. per con- . (5) \ \\\ \ \ conoseroa | \ Fig. I-A two conductor single phase circult (inductance) Characteristics of Aerial Lines 35 ig. 10-—Inductance due to lux between radius aand radivs b (Qin § abhenries per em.) here [= inductance due to the flux inside the conductor. ln! = inductance due to the flux outside the conductor 19 4 radius of one foot. 2in? = inductance due to the Hux external to 9 one foot radius out to Di fect where Dy is the distance between conductor 1 and conductor 2. From Fig. 9 it can be seen that itis unnecessary to in- clude the fix beyond the return conductor 2 because thia flux does not link any net current and therefore does not, affect the inductance of conductor 1, Grouping the terms in Eq. (3) we have: Dr abhensies per em. per eon T ductor, o 1B nla Ldueto L due flux out to fhux toaone external ft, rade to a 1 us ft rae dius out toDuit Examining the trm in the frst bracket, itis evident that this expression i the sum of the fx both inside the conductor (#) acd that external ko the conductor out to a radius of one foot (2int), Furthermore this expression contains terins that ae strictly a funetion of the conductor characteristics of permeability and radius, ‘The term in the second bracket. of Ey. (0) is an expres sion for inductance due to flux external to a radius of one foot and out to a distance of Dis, which, in the twore ctor ease, is the distance between conduotor 1 snd con- ductor 2. ‘Thie term ie not dependent upon the conductor characteristies and is dependent only upon conductor spacing Equation (6) ean be written again as follows:36 Characteristics of Aerial Lines 1 gy,Du abhenries per om, por (7) ain? 1 conductor. GMR in the first term is the conductor *feometrio roean radius", Tt ean be defined asthe radius of a tubular ¢on- dyetor with am infinitesimally thin wall that has the same external fox ost to a radius of one foot asthe internal and external fu of solid conductor 1, out to a radius of ane foot. In other words, GME is 'a mathematieal rudius assigned a sold conductor (or other configuration such 25 siranded eonstustors), whieh deseries in one term the inductance of the conductor due to both its internal fax (2) adhe ester os ost atin (2) GMR therefore makes it possible to replace the two terms 5 u erm (aint (42!) with om omg) expressed in feet. Converting Eq, (7) to practical units of induetive reactance, L=2in, whi entirely de- 1 De awe 1 ohms per conductor per mile ® hore fefrequeney in eps GaN Seoul? gems mean rat in fs, ‘Da mitance between conductor 1 and 2a fee AF ee th term baled and the nso tera roihen t= 4.424 obma per conductor per .aile (@) F vog voit 27015 logs gig 027045 lon where 2,=induetive reactance due to both the internal flux and that external to conductor I toa radius of one foot ductive reactance due to the flux surrounding, condnetor 1 from radius of ome foot out to a. radius of Da feet For the two-conductor, single-phase circuit, then, the total inductive reactance is (+24) ohtns per mile of eireuit (19) since the cireuit has two conductors, or both 9 “go” and “return” conduetor. sometimes a tabulated or experimental reactance with 1 foot spacing is known, and from this it is desired to eal- culate the conductor GMR. By derivation from, Eq, (8) V Tteparing Tess ap When reactanee is known not Lo # one-foot radius but out to the conductor surface, it is ealled Lhe “internal react~ ance.” ‘The formula for ealeulating the GMR from the internal reactance” is omR= aay feet j al Revetance™ (Wi oya 02794 ‘The values of GMR at 60 cycles and 2, at 25, 50, and 60 eycles for each type of conductor are given in the tables of electrical characteristics of conductors, ‘They are given Antilogs oTa8 te 0.778 ioe stranded manduetors and A.C Coogheting steel strands some laser 0.8260, 2Bstwe age oso, Bethree aver 9.8100, Single lager LOS 0364-0708 Point within stele ta eve ° Point oats ence to cle intacce to center of cele ‘eclangularvection of sides a asd OxsK(at8) crncuLaR TUBE 100 0801 F E 088 pt +~ 80 oF 0#1/2 OUTSIDE DIEMETER Fig. 11-Geometric Mean Radil and Distances, PHASE © Fig. 12—A Three-conductor three-phase clrcuit (symmetrical ‘pacing).Chapter 3 in these tables because they are @ function of conductor characteristics of radius and permeability. Values of za for various spacings are given in separate tubles in this Chapter for 25, 50, and G0 ayeles. This factor is de- pendent on distance between conductors only, and is not associated with the conductor characteristics in any way. Tn addition to the GM given in the conductor char- acteristics tables, it is sometimes necessary to determine this quantity for other conductor configurations. Figure is given for convenience in determining such values of GMR. ‘This table is taken from tho Wagner and Evans book Symmetrical Components, page 138. Tlaving developed 2, and 2 in terms of a tro-conductor, single-phase circuit, these quantities can be used to do- termine the positives and negativescquence inductive reactance of a threo-conductor, three-phase circuit Figure 12 shows a throo-conductor, three-phase circuit carrying phsse currents Ta, Iv, Lo produced by line to ground voltages F,, Ky, and E,. First, consider the case where the three conductors are symmetrically spaced in a triangular configuration so that no transpositions are re~ quired to maintain equal voltsge drops in eaeh phase along the line, Assume that the three-phase voltages Fy, Ey Hs are balanced (equal in magnitucle and 120° apart) so that they may be either positive or negative-sequence volt- agea, Also assume the currents [,, I, I. are also balanced, so that Jtf+/.a0. Therefore no return current floves in the earth, which practically eliminates mutual effects between the conduetors and earth, and the currents Za, Ts, Ze can be considered a8 positive- or negative-sequence currants. In the following solution, positive- or negatives sequence voltages B,, By, Be, are applied to the conductors ‘and corresponding positive- or negative-sequience currents are assumed to flow producing voltage drops in each con- ductor, The voltage drop per phase, divided by the eur- rent per phase results in the positive- or negative-sequence inductive reactance per phase for the three-phase cireuit.. "To simplify the problem further, consider only one current flowing at a time, With all three currents flowing simul- taneously, the resultant effect is the sum of the effects pro duced by each current flowing alone. ‘Taking phiase a, the voltage drop is: Ryn Bel Latee tb Titan tb Totag where (aa) nductive reactance of conduetor a. ‘4=toutual inductive reactance between con ductor a, and conductor b 24=mutual inductive reactance between con- ductor « and conductor. In terms of % and se, inductive reactance spacing factor, i) there only I, is lowing and returning by a remote path e feet away, assumed to be the point k. Considering only [flowing in conductor ® and returning by the same remote path f feet away fa acs) aay as) where za) is the inductive reactance associated with the Bux produced by 7; that links conductor @ out to the return path f foot away. Characteristics of Aerial Lines at Finally, considering only F, flowing in conductor ¢ and returning by the same remote path g feet away eae (ao) whore 24; is the inductive reactance ascovinted with the flux produced by Z, that links eonductor a out to the cetura, path g feet away ‘With all three currents Z,, In, I. flowing simultaneously, we have in terms of 2, and 2y factors Ey By act 2450)) +(e) Zaha) Fel tately~ Beton) an Expanding and regrouping the terms we have: Bg B= Ly Tistesen~ Teton FUerain + eran Fetsctsh as) Since I= 1, weitven fa—Tv, the terms in the bracket may be Tylraes) = 4.04) + Fleas 000s) D, Using the definition of a, 0.2704 1 tog thivexpression Using t i 0.2704 Flog thisenp Foy dom’ ) 10204 ) dais, dot, and dyay to the remote can be written I, ‘osrasl 1 (02704 tow Assuming the distane Goa) ang 0 fas ne des approach uoity. Since the log of unity iy sero, the te torms in the bracket are zero, and Eq. (18) reduces to path approach infinity, then the ratios Lats fetuses Fetny a9) (20) en where “inductive reactance for conductor a due to the flux ‘ut Lo one foot inductive reactance corresponding to the flux ex ternal to a one-foot radius from conductor @ out to the center of conductor 4 or conductor ¢ sinee the spacing belween conductors is symmetrical. ‘Therefore, the positive- or negative-sequence inductive reactance per phaso for a three-phase circuit with Tateral spacing is the same as for one conductor of a single- phase circuit as previously derives. Values of 2, for vari= ‘ous conductors are given in the tables of electrical char- acteristics of conductors later in the chapter, and th values of wa are given in the tables of inductive reactance pacing factors for various conductor spacings When the conductors are nnaymmetvieally spaced, the voltage drop for cach conductor is different, assuming the carrents to be equal and balanced. Also, due to the unsym- metrical conductor spacing, the magnetic field external to the conductors is not zero, thereby causing induce volt= ages in adjacent clectrical eireuits, particularly telephone cireuits, that, may result in telephone interferen ‘To reduce this effect to a minimum, the conductors are ‘transposed so that each conductor occupies successively the88 Characteristics of Aerial Lines a ete 21a SECTION ty ky te secTIon ‘ng SECTION t t Fig. 1A Three-conductor three-phase clceuit (unsymmet~ ‘eal spacing). ‘same positions as the other twoconductorsin'two successive Iineseetions. For throe sueh transposed fine seetions, called “barrel of transposition’, the total voltage drop for each, eonduetor is the same, and any electrical eirenit parallel to the three transposed sections has a net voltage of very low ‘magnitude induced in it due to normal line currents Tn the following derivation use is mado of the goneral ‘equations developed for the ease of symmetrically spaced conduetore, First, the inductive reactance voltage drop of phase @ in each of the three line sections is obtained Adding these together snd dividing by three gives the ‘average inductive reactance voltage drop for a line section, Referring to Fig, 13 and using Bq. (19) for the first line section whore 1, is flowing in conductor 1, Ey B = Lea leragny Lac In the second line section where /, is lowing in conductor 2, B= Bel = Dory Iota ~Lotsan, In the third line section where 1. is fowing in conductor 3, BYES Ter zea Detain ‘Taking the average voltage drop per line section, we have (BABA) + (By By) (Bo Ee") i Dividing by 1,, we have the positive- or negative-sequence inductive reactance per phase sae (tebe) ubts per phase per mile where (sus zam cua) lune per pao er mile (22) Chapter 3 Exprossed in genoral terms, AN 0.270145) og dan +196 ayo dos) oro 24=4 0.27045 Tog didnt ig 08 dnd f tog Vids sy 0.27045 log Vesta ort, 450.2704, log GMD where GMD (geometrical mean distance) = diag, and is mathematically defined as the nth root of an n-fold product. For a three-phase circuit where the conduetors are not ‘sytametrically spaced, we therafore have an expression for the positive. or negative-sequence inductive reactance, which is similar to the symmetrically spaced ease except za js the inductive-reactance spacing factor for the GMD {goometric mean distance) of the three conduietor separa tions. For zy, then, in the ease of unsyinmetrical eonduet- or spacing, we ean take the average of the three inductive- reaetunee spacing Factors a= H(taan+tuan+teen) ohn or we can ealeulate the GMD of the three spacings OMD=Wdadastn Took 3) and use the inductive-reactance spacing factor for this distance. This latter procedure is perhaps the easier of the two methods, ‘zis taken from the tables of electrical characteristies of ‘conductors presented later in the chapter, and zg is taken por hos per phase per mile 1300000 Jsoa00 5000 joeea00 7+ EOUNALEMT CONDUCTOR SPACING=FEET $7 ye wo Fig, 14 Quick reference curves for 60-cycle inductive seact= ancy eee phate inc pee phat) ung hard ran copper ‘conductors. For total reactance oChapter 3 a0 26.800 Z74795,000 382,500 396000 REACTANCE X* Xyh%y IM OMMS PER MILE Sas Te EQUNALENT CONOUCTOR SPACNG— FEET 2 40 Fig, 15—Quick reference curves for 60-cyéle Inductive react- lance of three-phase lines (per phase) using ACSR conductors. For total reactance of single- ‘multiply these values by two, Seo Eqs. (10) and (21). COPPERWELD COPPER. eerage Caren) REACTANGE 3+ Xg2%q IN OANS PER MILE oar 2 3 45 7 0 20 EQUIVALENT CONDUCTOR SPROING—FEET 30 40 Fig. 16—Ouick reference curves for 60-cycle inductive react fanco of three-phase lines (per phase) using Copperweld- Copper conductors. For total reactance of single-phase lines ‘uluiply these values by two. See Eqs. (10) and (21). Characteristics of Aerial Lines 39 ‘COPPERWELD (AVERAGE CURRENTS) uw | 5/16 @ 3 NO.IO- Be aanosL 1/2 308. REAGTANCE X #X,+Xq IN OHMS PER MILE 2 Osh 23 as 7 16 26 EQUIVALENT GONDUGTOR SPACING-FEET Fig. 17 Ouick reference curves for 6B-cycle Inductive reacts lance of three-phase lines (per phase) using Copperweld con- ‘Guctors. For total reacrance of single-phase lines multiply these valuce by twe. Soo kgs, (10) and (21). 30 40 from the tables of induetive-renetancs spacing factors. Geometric mean distance (GMD) fg sometimes referred to 1as “equivalent conductor spacing.” For quick reference the eurves of Figs. (14), (19), (16), and (17) have been plotted giving the reactance’ (raz) for different con- ductor sizes and “equivalent conductor spacings.” ‘Since most tree-phase lines or circuits do not have com ductors symmetrically spaced, the above formula for pos! tive- or negative-soquence inductive reactance is generally used, ‘This formula, however, assumes that the circuit is transposed. When a single-cireuit line or double-cireuit line is not ‘transposed, either the dissymmetry is to be ignored in the calculations, in which ease the general symmezrieal com ponents methods ean be used, or dissymmetry is to be eon- sidered, Unus preventing the use of general symmetrical components methods, In considering this dissymmetry, unequal currents and voltages are calculated for the three phases even when terminal conditions are balanced. In ‘most eases of dissymmetry itis most practical to treat the cireuit 43 transposed and use the equstioas for x, and 2 derived for an uneymuetrically-spaced transposed eizeuit Some ervor results from this method bul in general it is stnall ax compared with the laborious calculations that ‘must be made when the method of symmetrical compo- nents eannot be used.40 Characteristics of Aerial Lines Positive- and Negative-Sequence Reactance of Parallel Circuits—When two parallel three-phase its are close together, particularly on the same tower, the elfect of mutual inductance between the two circuits is not entiroly eliminated by transpositions. By referring to Fig. 18 showing two transposed cireuits on a single tower, the positive- or negative-sequence reactance of the paral: leled cireuit senor! [age ati pane O23 ls Ga ee gs ‘ohms per phase per mile (24) in which the distances are those between conductors in the first section of transposition, ‘The first term in the above equation is the positive: or negative-sequence reactanee for the combined eixeuits. The second term represents he correction factor due to the 0 oF 20 ov 0 ov Fast section oF %0 oe 20 oe ‘econo section 1 of ‘0 of 20 of ‘Ro seorion Fig. 18—Two paraflet chree-phase clrcults on a single tower showing transpositions, mutual reactance between the toro circuits and may reduce the reactance three to five percent, “The formula assumes ‘transposition of the conductor as shown in Fig. 18, ‘The formula also assumes symmetry abont the vertical axis but not necessarily about the horizontal axis. ‘As contrasted with the usual conduetor arrangement as shown in Fig, 18, the arrangement of conductors shown in Fig. 19 might be used. However, this arrangement of cori- Chapter 8 20 20 ov v- ‘materially Increnses the Inductance per plese. ductors results in five to seven percent greator induetive reaetanes than the usual arrangement of conductors. ‘This hhas been demoustrated in seversl references.* 3. Zero-Sequence Resistance and Inductive Re- actance ‘The development, of zero-serquence resistance and in ductive reactance of aerial lines will be considered simul taneously as they are related quantities. 20r0-50- ‘quence currents for three-phase systems are in phase and equal in magnitude, they flow out through the phage con ductors and return by a neutral path consisting of the earth alone, neutral conductor alone, overhead ground wires, or any combination of these, Since the return path often consists of the earth alone, or the earth in parallel with some other path such as overhead ground wires, it is necessiry to use & method that takes into account the re sistivity of the earth as well as the eurrent distribution in the earth. Sinee both the zero-sequence resistance and inductive-reactanee of three-phase circuits are affected by these tavo Factors, their development is considered jointly. ‘As with the positive- and negative-sequenee induetive reactanes, first consider a single-phase circutt consisting of a single conductor grounded at its far end with the carth acting. a8 a return eonduetor to complete the cireuit, “This permits the development of some useful concepts for eal- culating the zero-sequence resistance and inductive re- actance of three-phase cireuits. Figure 20 shows a single-phase cireuit consisting of a single outgoing conductor o, grounded at its far end with the return path for the eurrent consisting of the earth, A. second conductor, b, is shown to illustrate the mutual effects produced by current flowing in the single-phase eir- uit. ‘The zero-sequence resistance and inductive reaetanes of this circuit are dependent upon the resistivity of the ‘earth and the distribution of the current returning in the earth, This problem has been analyzed by Rudenherg, Mayr, Fig. 20—A single conductor single phase ciecutt with earth,Chapter 8 ‘and Pollaezek in Europe, and Carson and Campbell in this country. The more commonly used method is that of Carson, who, like Polluezek, considered the return current to return through the earth, whieh was assumed to have uniform resistivity and to be of infinite extent, ‘The solution of the problem is in two parts: (1) the de- termination of the self impedance 2, of conductor a with ‘earth return (the voltage between « and earth for unit eur- rent in conductor a), and (2), the mutual impedance Zen between conductors and b with common earth retumn, (the voltage betsseen b and earth for unit current in a ond earth return). As a result of Carson’s formulas, and using average heights of conductors above ground, the following funda ‘mental simplified equations may be written: B sero 0.00188f 470.004657/ | wv ear 5 Jon ON hms per mile 5) Zen 0.00159-4 0.004657 loge ohms per mile 5) where resistance of conductor @ per mile, ‘F=frequeney in eps. earth resistivity in ohms per meter cube, eomotrie mean radius of conductor in fect. listance between conductors a and 6 in feet. ‘A useful physical concept for analyzing earth-return circuits is that of concentrating the current returning through the earth in a fictitious conductor at some con siderable depth below the outgoing conductor a, This equivalent depth of the Getitious return eonduetor is rep- ented as Dy ‘or the single-conduetor, single-phase circuit with earth return now considered as a single-phase, two-wire circuit, the seli-inductive reactance is given by the previously de vied 027014 toe P28 oa nae phase red j0. Mg ome Gite ‘See Ea. (8)) for a single-phase, >, twomire ert o se see whore Da in re it, oF OOOH Tope ¢2> whore D, ‘substituted for Dy, the distance between conductor a and the fictitious return conductor in the earth. This expres- sion is similar to the inductive-ronetanee as given in Carson's simplified equation for self impedance. Equting the logarithraie expressions of the two equations, i B Dy 21007 fp: OO657 logue se = 70.004657 flog” GMR- or bow et St ‘This defines D,, equivalont dopth of return, and shows that itis a function of earth resistivity, », and ftequency, f. Also an inspection of Carson's simplified equations show that the self and mutual impedances contain a resistance component 0.00159, which is a function of frequency. @ Characteristics of Aerial Lines a Rewriting Carson's equations in terms of equivalent depth of retuen, De, 24> ro $0.001504+40.004857f loge ohma per mile. _D, GMR @s) =0.00159/+470.004652flogu”* ohms per mile. (29) "hs These equations can be applied to multiple-conductor cirouits if r., the GMI and ds refer to the conductors as & group. Subsequently the GMI of a group of conductors are derived for use in the above equations. ‘To convert the above equations to zero-sequenee quan- tities the following considerations must be made. Con- sidering three conductors for a three-phuse system, unit, ‘ro-sequence current consists of one arnpere in each phase conductor and three amperes in the earth return eircuit ‘To use Faqs. (28) and (29), replace the three conductors by 1 single equivalent conductor in which three amperes flow for every ampere of zero-sequence current, Therefore the corresponding zero-sequenee self and mutual impedances per phase are three times the values given in Carson's simplified equations, Calling the zero sequence imped- ances £9 and zm, We have: r<+0.00477f 4 j0.01397f logis aan coms per phase per mile. (60) 20 0.00477}+70 01807) logs Pt ohms per phase per mile en where J frequency in eps Fe=tesistance of a conductor equivalent to the three conductors in parallel, rz therefore equals the resistance of one conductor for a three-phase circuit. GMR=geometric moun radius for the group of phase conductors. ‘This ie different than the GMR for ‘single conductor and is derived subsequently 85 CMR aa=distance from the equivalent conductor to a parallel conductor, oF some other equivalent Conductor if the mutual impedance between to parallel three-phase cireuits is being considered For the case of a single overhead ground wire, Eq. (80) ‘vee the zero-sequenee self impedance. Equation (31) ives the zero-soquence mutual impedance between two overload ground wires Zero-sequence self impedance af two ground wires with earth return. Using Eq. (30) the zero-sequence self impedance of two ‘ground wires with earth return can be derived. D, 20= Br.+0.00477j+0.01897f logio ‘ohms per phase per mile (30) where r= resistance of a single conductor equivalent to ‘the two ground wires in parallel. (re therefore the resistance of one of EMR where te ‘the two ground wires).42 GMR-=geometrie mean radius for the two ground wires. (GMR therefore becomes VGIIRY stuart M(GMBY Gas) where day is the distance between the two conductors sandy) Substituting 5 for r, and /(GMRY Ga) for GMR in Ea. (G0), the zero-sequence self impedance of two ground, swires with earth return beeomes D. MR)(da.) “ Bry ' oF 00047774 70.01807/ le Fes cobms per mile per phase. Zero-sequence self impedance of m grownd wires with earth return Again using Eq. (80), the zero-sequence self impedance ‘of » ground wires with earth return ean be developed. 20 Bre+0.00477/-+70.01897f lop ‘ohms per mile per phase. ry oy Je conductor equivalent to nm ground wires in parallel, then ry=7* Since ris the resistance of a where 7 is the resistance of one of the » ground wires, in ohms per phase per mile GMR is the geometric mean radius of the n ground wires fas 8 group, which may be written as follows in terms ofall, possible distances, GMR= VIGMEY sessaur Cindinm Tine) (ree ieey Ue) deg ten Fenn) (€or Aina) Ft. ‘This expression can also be written in terms of all possible pairs of distances as follome. GM CR ga pena fea (3) ‘The equation for zero-sequence self impedanee of n ground ires with earth return can therefore be obtained by sube for r, and Eq. (33) for GMR in Eq. (30), stituting Self impedance of parallel conductors with earth return In the preceeding discussion the self and mutual im- pedanees betwoen single eylindrieal eonduetors with earth return were derived from which the aero-sequence velf and, mutual reactanees wore obtained, ‘These expressions wore expanded to include the ease of multiple overhead ground wires, which are not transposed. ‘The more common ease is that of three-phase conductors in a three-phase cireuit which can be considered to be in parallel when sero-se- {quence currents are considered. Aleo the three conductors in a three-phase circuit, are generally transposed. ‘This, factor was not considered in the preceeding cases for niul= tiple overhead ground wires. In order to derive the zero-sequence self impedance of three-phase circuits it is first necessary to derive the self impedance of three-phase circuits taking into account Characteristics of Aerial Lines Chapter 8 ‘transpositions, ‘The exprossion for self impedance is then converted to zero-sequence self impedance in analogous to the ease of single conductors Ww return, ‘Consider three phase conduetors a, , and c as shown in Fig, 21, With the conductors ursusposed the current L 3 ‘ ee 4 1 . ‘tbe = 4 =. - Fig, 21—Self impedance of parallol conductors with earth divides equally between the conduetore so that for a total ‘current of unity, the current in each conductor is one third. The voltage drop in conductor @ for the position in- dicated in Fig. 21 is sg 2 4 fae faye For conductor b: in which 2, 22, and za are the self impedances of the three conductors with ground return and Zap tho) and te are the mutual impedances between the conductors, ‘Since conductor a takes euch of the three conductor positions successively for a transposed line, the average drop per conductor is fltetee teed Beat eee Substituting the values of elf and mutual impedances given by Eqs. (28) and (29) in this capes, all a[rss0onn ‘ona loa +2 lope P42 loge P42 la 7 ‘ohms per mile, cath "$-+-0.001509/ +40.001687/ logs ‘ohms per mile.Chapter 8 "The ninth root in the denominator of the logarithmic term, is the GMK of the eireuit and is equal to an infinitely thin tube which would have the same inductance as the three- conductor system with earth return shown in Fig. 21. feet. ‘Therefore GME sna. Y (GM gasses OMD pany, feet. (35) Substituting GMReeas from equation (83) in equation 84), a= e0.001807 + 0.004657f logu ohms per mile, Te am (36) Tn equations (81) and (8), 1 the rsistance por mile of tne phase condctor Zerorsrquence self impedance af tree parallel conductora wilh earth veturn Bquation (30) gives the self impedance of theoe parallel ‘conductors with earth return and was derived for a total ceurront of unity divided equally among the three con- Guctors. Since aero-saquence current consists of unit eur= rent in each conductor or (ota of three times unit current for the group of three conductors, the vollage drop for aero-cequonen cureents is three times as great. Therefore Eq, (36) must be multiplied by three to obtain the 2ero- sequence self impedance of three parallel conductors with earth retura. ‘Therefore, sym rcb 0.00877 10.1807) 6 eee D. ohms per phase per mile (GMD Fincwacen (37) where YGMR wamolGMDYFaganan 3 the GMRyron de rived in equation (35) or V(GMR) senor dan“dov'dect Zero-sequence mutual impedance between two eirewits with earth return. Using « similar method of derivation the zero-sequence mutual impedsnee between 2 three-phase cireuits common earth return is found to be D. 20a) =0.00877f-+J0.01997f loge eMp ohms per phase per mile (38) where GMD is the geometric mean distance between the 2 three-phaso circuits or the ninth root of the product of, the nine possible distances between conductors in one group and conductors in the other group. Note the simi- Inrity between Bq. (88) and Bg. (31) Characteristics of Aerial Lines 43 Zero-sequence self impedance of to identical parallel cir cuits with earth retura For the spociat ease where the two parallel three-phase Cireuite are identical, following the same method of dorivation D, ats j0.01307/ loge 205 40000177 40.0199 lowe ees GD) 9) in which GMR is the geometric mean radius of one set of conductors, (Y(GMR cago EMD)aperine )» 81d GMD is the geometric mean distance between the two sets of conductors or the ninth root of the product of the nine possible distances between conduetors in one cireuit wad conductors in the other circuit "This equation is the same as $(2+22qn) Where 2s is the zero-sequence sell impedance of one cireuit by equa tion (87) and 2a; is the sero-sequence mutual impedance between {wo circuits as given by Eq. (88). For non identical cireuits it is better to compute the mutual and. self impedance for the individual cireuits, and using H(2e+eum) compute the zero-sequence self impedance. ‘ohms per phace per mile ero-soquence mutual Impedance betieen one circuit (with ‘arth return) and ni ground wires (with earth return) Figure 22 shows a three-phase circuit with n ground 24 ¢% TTT TET TTTTTTTTTOTTTOOTT TOTO Fig, 22-A three-conductor threo-phase circuit (with earth return) and m ground wires (with earth return) ‘wires. Equation (81) gives the zero sequence mutual im pedance between two conductors: 26 =0 00477f-+70.01897 loge P= cohims per phase per mile a) where da, ie the distance betwaon the two conductors. This, equation ean be applied to two groups of conductors if dy, is replaced by the GMD or geometrie mean distance between the two groups. Ip Fig, 22, if dhe ground wires are considered as one group of eonduetors, and the phase con- ductors a, b,¢, are considered as the second group of con- ductors, then the GMD between the tivo groups is Cnggtdeg—Tagnligeegy feet ‘Substituting this quantity for day in Eq. (31) results in an, ‘equation for the zero-sequence mutual impedance between, ‘one circuit and m ground wires. This ca i “no:44 Characteristics of Aerial Lines oom = O.00477F O1397f logue x = POST OB 8 i do a (40) Zeronsequence impedance of one circuit with n grown wires (ant earth) return Referring to Fig. 20 the zero-sequence self impedance of @ single conductor, and the zero-sequence mutual im- pedance between a single conductor and another single conductor with the same earth return path was derived. ‘Thece values are given in Kige. (30) and (21). Aa stated bo- fore, these equations ean be applied to multi-conductor cirouits by substituting the cireuit GMR for the conductor GMR in Eq. (30) and the GMD between the two circuits or dy, in Eq. (31). First, consider the single-conduetor, single-phase circuit, with eurth return and one ground wire with earth retnen, Referring to Fig. 20 conductor a is considered as the single conductor of the single-phase circuit and conductor b will, be used as the ground wire. Writing the equations for Band By, we have: Ey= Let loin a) Biya Lean Tite a2 Jf we assume conductor # as a ground wire, then By since both ends of this conductor are connected to ground, ‘Therefore solving Fg. (12) for Z, ond substituting thie value of fy in Eq. (41), To obtain 2,, divide £, by I, and the result is lms per phase per milo aa) ‘The rero-sequonce impedance of « single-conductor, single- phase circuit with one ground wire (and earth) return is, therefore defined by Bq. (43) when zero-sequenee self im= pedances of single-conductor, single-phase circuits are sub- stituted for zq and ayy and the zerosequence snutual im pedance between the {wo conductors is substituted for 2, Equation (48) can be expanded Lo give the zero-sequence impedance of a three-phase circuit with m ground wires, (and earth) return, Fm t0en = an ‘Where 29=zero-sequence impedance of one eizeuit with 1 ground wires (and earth) return, zero-sequence self impedance of the three hase circuit. fero-sequence self impedance of n groand Zeno“ zero-cequence mutual impedance betwoen the phase conductors as one group of conductors ‘and the ground wire(s) as the other conductor group, ‘Equation (44) results im the equivalent circuit of Fig. 23, for determining the zero-sequence impedance of one eireuit ‘with n ground wives (and earth) return, Bae te Chapter 3 General Method for Zero-Sequence Calculations —The preceding seetions have derived the zero-sequence self and mutual impedances for the more common cireult arrangements both with and without ground wires. For more complex eireuit and ground wire arrangements a 2atgr Zo) Zotar Paae) Fig. 28—Equivalent circuit for zero-sequence impedance of ‘one circult (with earth return} and n around wires (with earth return). general method must be used to obtain the zero sequenes impedance of a particular cireuit in such arrangements. ‘The general method consists of writing the voltage drop for eaeh conductor or each group of conductors in terms of zero-sequonce self and mutual impedanees with all com ductors or groups of conductors present. Ground wire conductors or groups of conductors have their voltage Arops equal to zero, Solving these simultanoous equations for 7, of the desired ciouit gives the zero-sequence im ppedance of that eireuit in the presenee ofall tho other zero- sequence circuits, This general method ie shown in detail in Chap. 2, Part X, Zero-Sequence Reactances, Two circuits, one with two overhoud ground wires and one with asingle aver- head ground wire are used to show the details of this more enorsl method. Practical Calculation of Zero-Sequence Imped- ance of Aerial Lines—In tho preceling diseussion number of equstions have been derived for zero-sequence self and mutual impedances of transmission Tinos taking into account overhead ground wires. ‘These equations can be further simplified to make tse of the already familiar quantities, 2s, and x4, To do this two additional quan tities, re and 7, are necessary that result from the se of the earth as o return path for zero-sequence currents. ‘They are denved froin Carson's formulas and ean be defined as, iollows: ¥.=0.00877f ohms per phase per mile, = 0.006086/ loge 4.6055 1087 ohms per phase per (46) tis now possible to write the previously derived equations for zero-sequenee sif and mutual impedances in terms of Tay Tata Yo, and se. The quantities ra ta % are given in the tables of Electrical Characteristies of Conductors and Induetive Reactance Spacing Factors. ‘The quantities r» and ty are given in Table 7 as functions of earth resistivity, p, in meter ohms for 25, 50, and 60 eyelea per second. ‘The following derived equations are those most commonly used in the analysis of power aystem proby rileChapter 8 Zero-sequence impedance—one circuit (with earth return) ‘hut without growl wires sorb O.OTTS D, FROIN FST assaalCMD ovr ‘ohms per phase per mile” 7 ayer $10,000 Tog 4.6850 10°F s t HOG LOR ES TRcioee = 20027945 1oRe CMD apetod sonnet ret ilzetza—Bze) ohms por phase per mile (47) where a= Mtauntbtxon e500) and £44) =24 from Table 6 for spacing @ to 6, ete. Metuat 2ero-soquence impotance between to cireuits (with earth return) but without grovnd wires Dy pee fm) LOOT +30 01597 lon Sry phase per mile. (38) Poyay = ret 0.006985/ loge 4 665x105 ~j0.006885F login GMD* om Fets(e—32s) ohms per phase per mile (48) where tis § rainy thn tte ttsews EEO Fa Qer tay t Lacey FTA;009) Zero-scquence self impedance—one ground wire (with earth return) big = Sr + 0.00877F+ j0.01897/ logie ohms per phase per mile, De 0) sun dret Tet 0 00088oge 650% 10° 1 0.006985) OB EST ee 2a = Brat rerbjler+824) ohms per phase per mile. (49) Bera sequence self impedance—twro ground wires (with earth ne su 80001777 47001807) le * GM racer hms per phase per mile. 2 Pn prot LODGES on 4. 0050> 10% 580 gh 08D og Se +> GM ie? test we tons Bt retilect pee) hms per phase per mile 0) where auezq from Table 0 for spacing between ground vies dy Characteristics of Aerial Lines 45 Zero-sequence self impedance—n grout wires (with earth return) Dy GMR FO00477/+.90.01897f losin ‘obra per phase por mile 0) dohrns per phase per mile (GMB create === yom deaatlere Bn Soaron’. a1 rubitetJ02791)E loge 4.005% 10°% tretieet/ MOTO loge 4.005% 10% 1 GHR ee a i === da ates ~ Bozo gon | aatoa-dan) lad deen) asia eran) Boston wen ow toatl dee) tented -dewe-s) |* wrote titet te 80D) ‘ohms per mile per phase i) D betwvoen all possible pairs of ground wires). Zero-sequcnce mutual impedance letwren one cirouit (with ‘earth return) and n ground seires (with earth return) 2 = 0.008775 +4 40.01307f logw x De Vibardoarhas ‘ols per phase per mile theta (40) st /0.008085f logis 4. 5855 x10"% = 70.006985f loge (Vilvatondear ~~ aes Tot tem Fron ‘herondeen)” ‘grq) ohims per phase per mile (52) whore 215 (ney Heaney te) not unen t240 F Eee) Zero-sequence impedance—One circuit with n grownd wires (and earth return) phase eizeuit, jero-sequence self impedance of m ground wires. zere-gequence mutual impedance betsveen the three-phase cireuit as one group of conductors and the ground wire(s) as the other eon duetor group.46 Positive-, Negative-, and Zero-sequence Shunt Capacitive Reactance ‘The capacitance of transmission lines is generally a negligible factor at the lower voltages under normal oper= sting conditions. Howover, it becomes an appreciable effect for higher voltage lines and must be taken into con- sideration when determining efficiency, power factor, regulation, and voltage distribution under normal oper~ sting conditions Use of eapacitance in determining the performance of long high voltage lines is covered in detail in Chap. 9, “Regulation and Losses of Transmission Lines.” Capacitance effects of transmission lines are also useful in studying such problems as inductive interference, ight- ning performance of lines, corona, and transients on power systems such as those that occur during faults For these reasons formulas are given for the positive, nogative-, and goro-eequence shunt eapacitive reactance for the more common transmission line configurations. The case of a two-conductor, single-phase circuit is considered ‘to show some of the fundamentals used to obtain these formulas, For a more dotailed analysis of the eapacitance problem # number of references are available. © In deviving eapacitance formulas the rlistribntion of & charge, g, on the conductor surface is assumed to be uni= form. This is true beeause the spacing between conductors in the usual transmission circuit is large and therefore the charges on surrounding conductors produce negligible dis- tortion in the charge distribution on a particular eon- ductor. Also, in the ease of a single isolated charsed eon uotor, the voltage between any two points of distances 2 and y meters radially from the conductor ean be defined as, the work done in moving # unit lange of one coulomb from point P: to point P, through the electri field produced by the charge on the conductor. (Seo Fig. 24.) ‘This is given © cconpucTon 2 Fig. 4—A two conductor single phase cleeult feapacitance). Boe De ‘eonpucroR 1 by (63) where ¢ is the conductor charge in eoulombs per meter. _ By use of this equation and the principle of superps tion, the capacitances of aystems of parallel conductors cas be determined Applying Eq. (58) and the principle of superposition to the two-conduetor, single-phase vireuit of Fig, 24 assuming ‘conductor I alone to have a charge gs, the voltage between ‘conductors 1 and 2 is, Va= 18X10" qn 2 volts, Voy 18X10" glu volts = 64) Characteristics of Aerial Lines Chapter 3 ‘This equation shows the work done in moving « unit charge from conductor 2 a distance Diz meters to the sur- fuee of conductor 1 through the electric field produced by q. Now assuming only conduetor 2, having a charge gs the voltage betwoen conduetors 1 and 2 is Yuan 1810" a In 7 volte (55) ‘This equation shows the work done in moving a unit charge from the outer radius of eonduetor 2 to conduetor 1 a distance Diz meters away Unrough the eleetrie field pro- duced by ge With hoth charges and gs present, hy the principle of superposition the voltage Vis is the sum of the voltages resulting from q: and g, existing one at a time. Therefore Vis the sum of Kgs. (54) and (55) when both charges @ and g, are present, Dasa , ssin(ain Peta eZ) oite Also if the charges on the two conductors are equal and their sum is 2er0, ate Substituting —g. for gs in equation (56) or 4 Vig=30% 10% In Pe volts ay ‘The capacitance between conductors 1 and 2s the ratio of {he change tothe voltage oF — 1 torads per meter, (58) cio Be ‘The capacitance to neutral is twice that given in Eq. (58) because the voltage to neutral is half of Vv 1 c= —— tua per mate, 18X10" In . ‘The shunt-capacitive reactance to neutral (or per con~ ; Asa eh or re rte a0 gy 2 mgs pe od pe ale wo ‘hn gt wien 008 ee 400082 egos pe ondtr pe mile 6 where Dn and rare in fect and f is cycles per second. Bq, (61) may be written Za 21 +24 mogohms per eonduetoe per mile, (63) ‘The derivation of shunt-caparitive renetanee Foemul brings about terms quite anslagous to those derived for inductive reactance, and asin the case of inductive react- fanco, those ters ean be resolved into components a8 shown in Eg. (62). The term 24’ accounts forthe elocteo- static flux within’ 8 one foot radius and is the termChapter 3 0539 gy! By, (3). Te eft a he oe i ctor outside radius only. The term za’ accounts for the leetrie flux between a one foot radius and the distance Dato the other conductor an isthe term 0.0683 logy Pay bn, (0D). Nee shat wale indtnerectance wher he condor gems man rs (CR) i Sathish set oan De ut et fea ord teat an Se an Hideo henge om a Soave ware bot tig stove rnd hegye zy is given in the tables of Electrical Characteristics of Reaetance Spacing Factor, and r¢ is given in Table 9, ZetScnt Set cape ln ater making use of the terms 2, 2° and x{. They are sum- ‘hvu ne fon tot logis 2k megohins per toile per @3) Shunt-Capacitive Reactanee, 2, of Throo-Phase Cireuile (Conduetors a, b, &) () Positive (and negative) sequence a{naf=2/-+24 mogohmsporconduetor permile. (64) 24 —Hioun ofall three x's for distancos between all possibile pairs). = Helen 2lact-2ded). Seo Table (8) (0s) (©) Zero-Sequence 2, of one citeut (and earth). ‘Bley =c-+5i—2r¢ megolims per conductor per alle (60) aevalue given in Eq. (66). Table (9) gives xf. (©) Zero-Sequence x, of one ground wire (and earth), Hig =Bzd a2 Megohms per conductor por 7) (@) Zero-Sequence x, of two ground wires (and earth). Phos aed telay 2 mile, (68) 2d™= 2d ue) = 24 for distance between ground wires. (©) Zero Sequence x, of n ground wires (and earth). 30 een per condor pe a) mile Characteristics of Aerial Lines a whore d= 2 Goan ofall x forall possible distances Koticen all posible pairs of ground wires) or z/= 1. (eum of ll 2's forall possible distances betiveen all ground wires) (f) Zero-Sequence x. between one circuit, (and earth) and n ground wites (and eaeth) ‘x/—QaJ mogohims per conductor per mile. (70) Fielitrdoertelcey tL oes “Fede tem). (@) Zero Sequence 2 of Boag ne cireuil with n ground wires :megobros per eonduetor por mile. (71) Shunt Capacitive Reactance, 26 of Single-Phase Circuile (Conductors « and 6) (h) 2. of single-phase cireuit of two identical conductors 2'=2(e{-+a4) megohms per mile of eireuit. (72) x{=ard for spacing botween conductors. (D 2 of single-phaso circuit of two non-identical eon- ductors a and b. 2) = atu) ttle) +2xd megohms per mile of eireuit. (73) (2. of one conductor and earth. pinned megohtns per mile, a) Tn using the equations it should be remembered that the shunt capacitive reactance in mogohms for more than one ile decreases because the caparitanes incresses, For more than one mile of line, therefore, the shunt-capacitive re- actance as given by the above equations should be divided by the number of miles of Tine, 5. Conductor Temperature Rise and Current- Carrying Capacity In distribution- and transmission-line design the tem- perature rise of conductors above ambient while carrying current is important. While power loss, voltage regula- tion, stability and other factors may determine the elioice ‘of a conductor for a given line, it ie sometimes necessary t0 consider the maximum continuous current carrying capa city of a conductor, ‘The maximum continuous eurrent rating is necessary because it is determined by the max- ‘umn operating temperature of the eonduetor. ‘This tem- perature affects the sag between towers ot poles and do- termines the loss of conductor tensile strength due to annealing. For short tie lines or lines that must carry excessive loatls under emergency eonditions, the maximum continuous current-carrying capacity may’ be important in solecting the proper conductor. The following discussion presents the Schurig and Frick* formulas for caloulating the approximate eurrent-carrying ‘capacity of conduetors under known conditions of ambient ‘temperature, wind velocity, and limiting temperature rise. ‘The basis of this method is that the heat developed in ‘the conductor by J*2 loss is dissipated (1) by convection48. in the sureou objects: PR 2 air, and (2) radiation to surrounding ‘This ean be expressed as follows: WoAW,)A watts @) ‘where I=ccnduetor eursent in amperes conductor resbtance per fast atts per square ine cliseipated by convection Wy walls per square inch dlinsipated by raliaton, conduttor surface area in square inches por foot af length ‘The watts por equate inch dissipated by conveetion, We, cean be determined from the follossing equation: Mt watts per square inch (7) where p=pressure in atmospheres (p= 1.0 for atmo- -spherie pressure), = velocity in feet per aovond. ‘Ty (degrees Kelvin) average of absokute temper- atures of conductor antl ain outside diameter of conductor in inches, (degrees C) temperature rise, At: ‘This formula is an approsimation applicable to eon uector diameters ranging Irom 0.3 inch to 5 inches or more when the velocity of air ts higher thun free convection sir currents (0.2 -0.5 sce) The watts por sqrare inch diipated by radiation, Wy, can be determined from the following equation -si{ (2) -(iPa)'] alts per sauiare inch where E clative emissivity of conductor surface (P=10 for “black bony,” oF 0.5 for average dined copper) grees Kelvin) absolute temperature of eon- ductor, (legs Kelvin) rounding. By calculating (1.41), A, and R, it is then possible todotermine I from 4, (75).. The value of f° to use is the sce resistance at the vondkctor temperature (ambient temperature plus temperature rise) taking into aecount kin effect as discussed provieusly’ in the section on pos itive: and negutive-soquence resistances, ‘This method is, in general, applicable to both copper ‘and aluminum conductors, ‘Fesis have shovwn that alim= num conrluctors dissipate heat st abont the same rate as copper conductors of the same outside diameter when the temperature rise is the same. Where test data is available ‘on conduetors, it should be used, ‘The above general mothed ean be used when test data is not available, of to chook: text results, ‘The effect of the sun upon conductor temporature rise is generally neglected, being some 3° vo 8°C. This small eect sles important nde condition of high tempera: ture rise above ambient ‘The tables of Hleetsieal Characteristics of Conductors inelude tabulations of the approsimate maximum eurrent- Te psolute temperature of sur. Characteristics of Aerial Lines Chapter 8 a e200 ‘06 APPROXIMATE. CONOUCTO ach Bo 3 ar AMBIENT TEMPERATURE °C 3 ee Fig. 23—Copper conductor curres pores VS. Ambient Tempe ying capacity in Am= "Copper Conductors s¢ ince "78°C, wind velocity at 2 fps.). sod g APPROXIMATE CONOUCTOR, CURRENT. 0: cc +9 aol oF 1 2 Gao AMBIENT TEMPERATURE °C se Fig, 26—Aluminum conductor current carrying capacity in Amperes VS. Ambient Temperature in °C. (Aluminum Con ductors at 78°C, wind velocity at 2 Fp).Chapter 3 TABLE 1—Cuaracterisrics oF COPPER Conpuctors, Hand Drawn, 97.3 Percent Coxpuctivity Characteristics of Aerial Lines 49 ® aa & ef 2a a le Bis Sea Bens aay Sp oe Bi rae [atk S288 4488 388 BS Bess S325 pe) ee Set lsc for ie oe ti a deagaee sahara: bam ax aa be ea tp ey ebm hae iaieps mana nan kee STEMS ae po har Ret ing capacity based on 50°C rise above an ambient of 25°C, (73°C toual conductor temperature), tarnished sur- face (=0.5), and an air velocity of 2 feet per second. These conditions were used after diseussion and agreement with the conductor manufacturers, ‘These thermal limitations are based on continuous loavling of the conductors. ‘The technical literature shows little variation {rom these conditions as line design limits.’ The ambiont air tem- perature is generally assumed to be 25°C: to 40°C whereas the temperature rise is assumed to be 10°C to 60°C. ‘This ives a conductor total temperature range of 35°C to 100°C. For design purposes copper or ACSR conductor total Cemperature is usually assumed to be TH°C as use of this value has given good ennduetor performance from an annealing standpoint, the limit being about 100°C where ‘annealing of copper and aluminum begins. ‘Using Schurig and Frick’s formulas, Fig. 26 and Fig. 26 Ihave been calculated to show how eurrent-earrying eapa- city of eopper and alumina eonduetors varies with sm bient temperature assuming a conductor temperature of 75.C and wind velocity of 2 foet per second. These values tare conservative and ean be used as 2 guide in normal line design, For those lines where @ higher conductor tem- perature may be obtained that approaches 100°C, the con- ductor manufacturer should be eonaulted for test data or father more acentate information as to conduetor tempera ‘ture limitations, Sueh data on copper eonduecors has been, presented rather thoroughly in the technical literabure? JIL TABLES OF CONDUCTOR CHARACTERISTICS. ‘The following tables contain data on copper, ACSR, hollow eopper, Copperweld-copper, and Copporweld con ductors, which along with the previously derived equations, ‘permit the derermination of positive-, negative, and zero fequence impedances of eonduetors for use in the solution of pawor-system problems, Also tabulated are such eon ductor charseveristies as size, weight, and current-earrying ceapacity as limited by heating ‘The conductor data (ry, 24, 1.) along with inductive and. shunt-capacitive reactance spacing factors (xa, 4?) and zero-sequenee resistance, inductive and shunt-capacitive reactance fnotors (Fm 26) 2!) permit easy substitution in the previously derived equations for determining the sym- ‘metrieat component sequence impedances of aerial circuits, ‘Tho eross-seetional inserte in the tabley ary for eage in60 Characteristics of Aerial Lines Chapter 8 & ‘TaBLe 2-A—CHARACTERISTICS OF ALUMINUM CABLE STEEL REINFORCED sao mon to een % sige |_ seen oentn atest | abgiaree eee] save, | Gye we | aE San fae [offal oe | wer | wert alate & ESRB a3 abana 2 welonlubest phemganed ya ou |e Behe a 8328 & abe iB 12 EB as pelte g : ae sa pie sou) 2 a0 | ort oes 230) ae z om | | BB Be ig g GB | EE Re Bucs a oe Hig we | we a oi vet be Baas el gE Wega 67a abate aes Tala fp aei 351 ana be Be 6 oie ee eee eee ee eRe RE Eee GRRE 2 18] ad SEPRR PERRO TE TREE pibeebepel ¢ |) ahs tah ta babe i ES 1B) ea PERS DEES Pe rea AE ve ere, t ite ret Me i i Bt a e ‘TABLE 2-B—Cuanacrenistics OF “EXPANDED” ALUMINUM CABLE STEEL REINFORCED (Glam Company ether) Pees Cater eu nen ease ag Heig «51 & Characteristics of Aerial Lines ‘Tape 9-A—CHaRactertstics oF ANACONDA HOLLOW COPPER Conpucrons Chapter 8 of [ERE EERE SENSE SEE] gy pal REE it NE ee Esccesse| g | “Gib! stetote: cacasate_asa: * i af |SSESEEES ARREARS ereaeata| B | 2 HL HEAD UEloiieg gacaazes geaaaza sue) f Hy | of (avaees EEE ¥ sete] 8 Susinede G4282602 Eeaaeede 8tg| 7 ol |G jeueaasa eect piceean aoeeeees| © G02 SG0RERE HES0G252 | : face se: and f2snres! Fl! : E Sees seen) © HAHBIEH HS? ie el | Ge g saaneney ggunaragli S eq ELNISSRH Sicseag i i Ha i] “HL | f/Eise Gubtades Beet thee Z i) Ga [eseazaae viceszas eeeasse aranezsli 3 3 /SEEE EEEREES SEEEEEEE 8 fa oe he be EuDESE ee fey |sessezes asaeaaas seataeea|? 5 = ia Bi [asensace Bssresn smreconll @ E ie BSR RARGEEDE SEORECEEE i 4 u ce = qeaggzey eeetesve sagaeeeelt ae i Hi LE 2 GRRISST SESAGEL: EIU: 3 fu SgF GERELEE GELIUEST TERIEEES TEE| BE | siaig [psaazse cuevenee eee sal [aa enn sae ala Hie (BREESE caesaes acses'®? S505qg.|6 | aijed [ABH ddeacseg euenorse S24:3582 244] (le a yes wawiiaz wavass[, © | Jgsiiy [=e SESE BREESE ele a BG PEG GS jg [pen ganntene geeeeees suensia: 595]62 Characteristics of Aerial Lines Chapter 3 $5 Tus ¢4-Cusmernmes or GoerERWLD-COPrER Conares Nomi ‘ck Wire ce ies es ESE Sah Saabs ke [RE ee sae] ge [pane sole ae bank h ee i _ i ESE a be y Qegeueg a SEINE 8 SOO ER EELS BSP aE H See E ee Beseek s: u BEES ceReRe YEE ERE g BER ES Re HERIER elec! hehe ERE be ka ke Bt mo 5 a hm fra = Fieettetdve tine ett ol eve. Pa ye finding the appropriate table for a particular conductor. Aluminum Cable Stoel Reinforecd, represents stranded. For these figures open circles, solid circles, and ervss paper, hatched cireles represent copper, steel, and aluminum con ‘The authors wish to ueknowledge the cooperation of Ue ductors respectively. ‘The double eross hatched area in the eonduetor manufacturers in supplying the information for insert for Table 2-1, Characteristics of “EXPANDED” compiling these tables.Chapter 8 Characteristics of Aerial Lines 58, ‘TABLE 4-B—CHaractenistics or COPPERWELD Conpuctors & | te fe A ett al Sees | omigetiene | cattle, | Me le ae = el Ey Sram CSE | ont oe lets Salata ate ue hea her ied Me 3 ise] 8 hae bse ke pes iB iS] BOP ER eas PRES BS aby Se ® PERSE Re 12) ES @ RBG ES Baa es i eer alas AB pa UP ico ing ic of cr. "Poon maple eh ay eens oe ON en Se a et ng pty 6 Cea” ‘Taste 5—Ssiw Errzcr Tani K x x | | x 00 100000 10 zo 1.07616 30 1 31800 o 1 00 rn a sn 35102 02 oon01 n | 22 32 Test os 1.00004 13 onsro 2a 33 £41000 a4 ooo rr o1oee 24 Bt Yas570 0s | Leone us ass 2S 35 rave oo 1.0967 10 + omen 26 | 36 1 20 oF Fooi24 ut ‘1205 2t a7 Loaner os 192i 18 5200 28 DSS) Von oo Foss v3 00 23 | HeaChapter 3 Characteristics of Aerial Lines TABLE 6—INDUCTIVE Reactance SPACING Factor (24) Orns pen CoNDUCTOR FBR MILE. 3 g | 8 2, 2% a iii q 2 é 8 ao a s f Se ts e A BS oS 8 5 & BB OLS 3 B al? a 2g $f a 2 Els ae BE Lt 38 25/8 ay a ac | pea] " sae 2 RE | fed * geo 2bS sek a 2 5 6 § z secs i & a B sesecese| 8 —— 5285 | |Fececece _ SESE = i Sosnte hen ennasn shakasnaat Denes » oEEEEEE Sag see dN ouencca Rneaeaaqaneess: econ easnaima seaaandataane® 5 See Newnenae grunge: #RRRRARRRRARRSERTSE! crewsCharacteristics of Aerial Lines 58 ‘Chapter 3 ‘Taste 8~SuUNT CAPACITIVE ReactaNce Sracine Factor (xi) MEcouMs PER Conpuctor Par MILs sits, FUNDAMENTAL EQUATIONS msde 2s, es Sinn ft. slat 25 eyeles af at 50 eyclos “Soeur ee3288E SEGCINE = (SEER ‘Tas 9—Zeno-SuqueNce SHUNT CAPACITIVE REACTANGE FACTOR 22022882 Ese paccesse. gocessos [-[ e2bEeaaE, ~ | BeSSe Hee Megohma per Conductor per Mile FREQUENCY, He geg8G2 | 6 | igeseesee| = (oBGSaucedeaendeeesee este oe Sceszes|5| | _f [eonnreenessznszushanaatanamninanhesessmeesvgsysse qa sTEI EE SEDESEESESTE Fg SSES IEEE SESSA ANGSaURMEanE: SRARERERRA onanwoerseanszenshenandnanghanasnasnsheRgsdas9ee3 Jixanssneze g|Seestecessey dpgpenseceys J ipanseeeenza56 Characteristics of Aerial Lines IV CORONA ‘With the inereased use of high-voltage transmission lines and the probability of going to etill higher operating volt- ages, the common aspects of corona (radio influence and corona lose) have become more important in the design of transmission lines. Tn the early days of high-voltage transmission, corona was something whieh had to be avoided, largely because of the energy loss associated with it. Tn recent years the RU (radio influence) aspect of corona has become more im- portant, In areas where RI must be considered, this factor ‘might establish the limit of acceptable corona performance. Under conditions where abnormally high voltages are present, corona can affect ayetem behavior. It ean reduce the overvoltage on long open-cireuited lines, Tt will atten uate lightning voltage surges (gee See. 29 Chap. 15) and witching surges."7 By increasing the electrostatic cou- pling between the shield wire and phase conductors, co- ona at times of lightning strokes to towers or shield wires reduces the voltage across the supporting string of insu- lators and thus, in turn, reduces the probability of flast ‘over snd improves system periormance. On high-voltage lines grounded through a ground-fault neutralizer, the in- pphuuse eurrent due to corona loss ean prevent extinction of the are during a line to ground fault, 6. Factors Affecting Corona At a given voltage, corona is determined by conductor diameter, line configuration, type of conductor, condition of its surface, and weather. Rain is by far the most im portant aspect of weather in increasing corona. Hoaritost and fog have resulted in high values of corona loss on ex- perimental test lines, However, it is believed that these high losses were caused by sublimation of condensation of ‘water vapor, Which are conditions not likely to oeeur on an operating line because the conductor temperature would normally be above ambient. For this reason, meastre- ments of lose made under conditions of fog and hourfrost ‘might be unreliable unless the conductors were at operat ing temperatures. Falling snow generally causes only & moderate inerease in corone, Also, elative humidity, tem perature, atmospheric pressure, and the earth’s electrie field ean’affeet corona, but their effect is minor compared to that ofrain. There are apparently other unknown factors found under desert conditions which can increase eoron."? ‘The effect of atmospheric pressure and temperature is generally considered to modify the eritical disruptive volt= fge of w conductor direty, or as the 9s power ofthe ait deity factor, 8 which ie piven by: wr a 0rF “ here mbarometricpremre in inches of mereury P= temperature in degrees Fahrenheit, ‘The temperature to be used in the above equation is gen- erally considered to bo the conductor temperature. Under Chapter 8 ‘Tauus 10—Stavoaro Barowornic ESAS ‘ais | een] Mis | Roa sion | 4.000 BG | ia vs o | am | Sih wt | Bi | tr mo | B | fee fo | Ss | mw standard conditions (29.92 in. of Hg. and 77°F) the air density factor equals 1,00. ‘The air density factor should be considered in the design of transmission lines to be built, in areas of high altitude or extreme temperatures. “Table 10 gives barometric pressures as a function of altitude, Corona in {air westher is nogligible or moderato up to a voltage near the disruptive voltage for a particular con- ductor. Above this voltage corona effects increase very rapidly. The calculated disruptive voltage is an indicator of eorona performance. A high value of eritieal disruptive voltage is nol the ouly criterion of satisfactory corona, performance. Consideration should also be given to the seositivity of the conductor to foul weather. Corona in ‘creases somewhat more rapidly on smooth conductors than, it-does on stranded conductors. ‘Thus the relative corona, characteristics of these two types of conductors might intorchange between fair and foul weather. The equation for critical disruptive voltage Begalr may Dt Bu evitionldiaraptive voltage in kv to natal duocriticl gradient in kw per centimeter. (Ref. 10 and 16 use gr=211 Ker/om ema, Recent work indicates value given ia ‘See, 10is more aeourate,) meas of conductor in contimneters he distance in ceutimetirs ttween conductors for single phase, or the equivalent phase spacing, for three-phase voltages, ‘me nurface factor (common valves, 0.84 for stranded, 0.92 for Segmental conductors) snair donsity factor Gy ‘The more elocely the surface of u conductor approaches 1 emoath cylinder, the higher the eritieal disruptive volt- ‘age assuming constant diameter. For oqual diameters, a stranded conductor is usually satisfactory for 80 to 85 per- cont of the voltage af a smooth conduetor. Any distortion of the surface of a conduetor such as raised strands, die ‘burrs, and scratches will increase corona. Care in handling, conduictors should be exorcised, and imperfections in the surface should be corrected, if it is desired to obtain the best eorona performance from a conductor, Die burrs and die grease on a nev conductor, particularly the segmental type, ean appreciably inerease corona effects when it is first placed in seeviee. ‘This condition improves with time, taking some six months to become stable. Strigel" concluded that the material from which a con ductor is made has no effect on its corona performance, In,Chapter 3 4 rit -@ ao) “ “© | geo] — 1 é / Ry } | " T s|—- t — : a a KV industrial areas, foreign material deposited on the con ductor can, in some cases, seriously reduce the eorons per- formance. '(Reference 28 gives some measurements made in an industrial area.) Corona is an extremely variable phenomenon. On a con- ductor energized at a voltage slightly above ils fair weather corona-starting voltage, variations up to 10 to 1 in eorona, loss and radio-influenee factor have beon recorded during fair weather. ‘The presence of rain produces corona loss on, conductor at voltages us low as 65 percent of the voltage at which the sume loss is observed during fait-weather, ‘Thus it is not practical to design a high-voltage line such, that it will never be in corona. ‘This also precludes ox- pressing a ratio between fair~ and foul-weather corona, sinee the former might be negligibly small If @ conductor is de-energized for more than about a day, corona is temporarily inezeascd. ‘This effect is moderate ‘compared to that of rain, It can be mitigated by re-ener= sing line during far weather where sucha choices pos 7. Corona Loss Extensive work by a large number of investigators has been done in determining corona loss on conductors oper- ‘ated at various voltages. This work has lead to the devel- Characteriotice of Aerial Lines ot Curve I-14 io, HH copper. 3°0388, Ref. 19. Corona loss test ‘ade in dasert at lotion where abnormally igh corona Tos Dbserved on the Hoover-Los Angeles 287 5-ky line, which is strand Irth this conductor. Meanurement mad in thre phase tst line. ‘This partionlar eurve is plotted for 30.88 to show operating cor ddion in desert, All other curves are for 51.00. ‘Curve 2—Same as ourve 1, exeept converted to 5=100 Gurve Lot in. HH copper. Ref 12. Corona Tost test made in California. Comparison with curve 2 shows effect of desert condi tons Measurements made on throe-phsco tet line, U-foot fat spacing, 10-foot sag, A-oot ground clearance, 700 fect long. Curve Ln, TU. Ref. 19. Measurements made on tbreeshace Tost Jing, 22fo0t flat spacing, 18foot sug, 8D Toot clearance to ground, 700 feet long, Curve 61.65 in smooth, Ref, 12, ‘Phe conductor had a poor sur face. Measurements made on thre-phate tt Hing, 000% spe: ing, I-oot sg, 30foot ground clearance, 700 feet Tong (Curve 61.85 in smooth sluminura. ‘Ref, 27. Reference curve ab- tainod by converting por phase measurertent co loss on throw phase Tine. Dimensions of fine not given Curve T—104 in. smooth evlinder. Ref. 23, Ta reference this eon ductor is referred to a8 having an infinite number of strands, ‘Plotted curve obtained by conversion of perphise measurementa lo thrae phase values using an entimatod wedi for charging hve, to give lass on a line having 45fo01 fat eontiguration. (Curve $1.90 in. emaoth sania. Raf. 28, Teaference curve gives threephace los, but Hine dimensions are not give ‘Queve 0-157 in. smooth Ref. 23. ‘Tht eonduetor was smooth and ‘lean. Referenen curve gives perphase values. Plotted curve it for 45-10% fat epacing. Fig. 27—Pair-Weather Corona-Loss Curves for Smooth Con- ore; Air Denalty Factor, 51. opment of three formulas generally used country (Reference 18 gives a large number of formulas). ‘The Carroll-Rockwell and the Peterson formulas are con- sidered the most. accurate especially in the important low loss region (below 5 kr per three-phase mile), The Peter- son formula, when judieiously used, has proved to be a re- liable indicator of corona performance (sce See. 9) for transmission voltages in use up to this Lime. Recent. work ‘on corona Joss has been directed toward the extra-high- voltage range and indicates that more recent information should be used for these voltages. Fair-weather corona-loss measurements made by a-num- bor of difforent investigators are shown in Figg. 27, 28, and 20. All curves are plotted in terms of kilowatts per three- phase mile, The data presented in these curves has been correcteil for air density factor, 8, by multiplying the test voltage by 1/823. Some error might have been introdueed jn these curves because in most cases it was necessary to convert the original data from per-phase measurements, ‘The conversions were made on the basis of voltage gradient, at the surface of each conductor. The curves should be used as an indicator of expected performance during fair ‘weather. For a particular design, reference should be made to the original publications, and a conversion made for the design under consideration. ‘The relation between fair-58 — Lo] 35} | | -® « Cc . ge 3 § \ 4 8 305 — 0 Characteristics of Aerial Lines Chapter 3 Curve 1-14 jo. ACER, Ref, 12, Conductor was washed with gasc- Tine then 4oup and water, Test configuration: three-phase Tine, B0-foot at spasing, T6footeag, foot ground clearance, 700 feet lone: Curve 21.0, ACER. Ref. 11. Conductor weathered by exposure toair without contiguous energization. ‘est configuration: three hace line, 20-ft fat spacing, 700 eer long Curve 31125 in, hollow copper. Ref. 14. Washed in seme manner ‘as for eutve I. Test configuration: three-phase line, 22-fo0t Bat Spacing Gurve 41.40, hollow eopper Ref. 14. Washed jn sume manner for curve 1. Test conBiguration: three-phase line, 0-foot fat Spacing, 6-fot sag, Hoot ground elearanee, 700 fet long. ‘curve $2.00 in, bolow aluminate, Re 14. Washed in same man ‘ner as for curve 1. Test configuration: three-phase fine, B-foot fat spacing, J8foot sag, 30-foot ground clearance, 700 feet long Curve 6-100 in, stlaluminur. Bef, 22, Raferenco curve isaver- fue fair-weasher corvin hse obtained by converting per phave renurerpents {three-phase values, fora ine 22.9 foot at spac: ing, 328 foct high, This conductor used on 220-ky lies in Sweden ‘whieh have above dimensi Curve 7128 in, slechaluminum, Ref. 22 App. A. Plotted curve ‘obtained by estimating average of a uunber of Fut-weather per phase curves given in reference and converting to threo-phase loss for line having 32-fo0 fat spacing, Hoot average height. Curve 8-104. stovlsluminim, 24strand, Ref-23. Plated curve ‘obtained by conversion of per-phase measurements to three-phase ‘values, using an estimated value for charging kva, to givo los on a Tine having 46-toot flat eoofgaration Cure 9-01 in. Hollow Copper. Ref. 1 rest configuration: Ubre-phase line, 20foot fat spacing, long. Conductor washed. feet ‘Fig, 28 —Pair-Weather Corona-Lows Curves for Stranded Con- ductors; Air Density Factor, 3= fand fonl-weather corona loss and the variation which ean be expected during fair weather is shown in Fig. 30 for ‘one conductor. Corona loss on a satisfactory line is primarily caused by tain. This is shown by the fairy high degree of correlation between total rainfall and integrated corona loss whicl has been noted. 2% ‘The corona loss at certain points on a transmission line can reach bigh values during bad storm conditions. However, such conditione are not likely to oc- cur simultaneously ail along « line. Borgquist and Vre= thom expect only a variation from 1.6 to 16 kor per mile, swith an average value of 6.5 kw per mile, on their 380-ky lines now under construction in Sweden, ‘The measured loss on their experimental line varied from 1.6 to 81 kw por mile, ‘The ealeulated fair-weather corona loss common in the U.S.A. is generally less than one lew per mile, based on, caleulations using Reference 16, Where radie-influence must be considered, the annual corona loss will not be of much economic importance, and the maximum loss will, not constitute a serious load, Corona loss is characterized on linear coordinates by a rather gradual increase in loss with increased voltage up to the so-called “knee” and above this voltage, a very rapid inerease in lose, ‘The knee of the fair-weather loss curve is generally near the eritical disruptive voltage, A transmis- sion line should be operated st a voltage well below the voltage at, which the loss begins to increase rapidly under fair-wenther eonditions. Operation at or above this point cean result in uneconomical corona loss. A very careful an- alysis, weighing the annual energy cost and possibly the maxiraum demand against reduced capitalized line cast, ‘must be made if operation at « voltage near or above the knee of the fai-weather loss eurve is contemplated. Corona loss on @ conductor is a function of the voltage ‘gradient at its surface. Thus the effect of reduced con- ductor spacing and lowored height is to increase the corona loss as a function of the increased gradient. On transmis- sion lines using a flat conductor configuration, the gradient ‘at the surface of the middle phase conductor is higher than ‘on the outer eonducior. This results in corona being move prevalent oa the middle conductor. 8, Radio Influence (RI) Radio influence is probably the factor limiting the choiee of a satisfactory conductor for a given voltage. ‘The RE performance of transmission lines has not heen as thor oughly investigated as corona loss. Recent publications, {cee references) present most of the information available. RI plotted against voltage on linear graph paper is eharactoriaed by 4 gradual increase in RI up toa viChapter 8 — 35] 30] KW/S-PHASE MILE Fig. 29—Pair-Weather Gorona-Loss Curves for Two-, Three tage slightly below the minimum voltage at which meas- urable corona loss is detected. Above this voltage, the inexease in the RT is very rapid. ‘The rate of inerease in RT is influenced by conductor surface and diameter, being higher for smooth conductors and large-diameter com ductors. Above a certain voltage, the magnitude of the RI field begins to level off. For practical conductors, the leveling off value is muck too high to be acceptable, and where RT is a factor, lines must be designed to operate be- low the voltage st which the rapid inereaze starte during fair weather. Figures 32 and 33 are characteristic RU curves, ‘The relation between fair- and foul-weather corona per- formance is shown in Fig. 32. An evaluation of RI in the desiga of a high-voltage line must consider not only its magnitude, but its effect on the various communication services which require protection. Amplitude-modulated broadcasting and power-line carrier fre the most common services encountered but other serv- ices such as aviation, marine, ship-to-share SOS calls, po- lice and » number of government services might also have to bo considered, In determining the RI performance of a proposed line, the magnitude of the RI factors for the entire frequency Characteristics of Aerial Lines 59 (Curve 1—4/0.985/15.7* (Suooth) Ref. 25. bot given, but assumed 1L10, which is average value for Germany. Reference curve ob- fined by converting tingle phate maasurements to Uoe-phase values on the basis of surface gradient, Dimensions of line used i raking conversion are not given. Curve 2—4/0827/15.7" (stranded. sluminum-stedl), Ref. 25, 3= 1.002, See dixcusion of Carve 1 Curve $3/0.985/11.8" (Smooth). Ref, 26. $=1.092. Reference ‘corve give singl-phste mesturements versu line torground volt fer, but ie not clear whether actual test voltage or equivalent ‘voltage at line height ie given. Latter was ured in euling the com ‘version to three-phase. If thie is wrong, curve is approximately TS percent low in voltage, Converted t0 fat configuration of 45 feet. (Curve 4—2/1.00/17.7" (Stranded slumioum-stel). §~1.01. Ref.12, ‘App. A. Reference carve gives persphase measurements versus {adieot. Converted to three-phase corona lose on ine of 2.54008 ‘average height, 3.4fo0: fat configuration. Curve 6 2/1.35/11.7" (Stranded auininam ste! & not given, prob ‘bly close to unity. Rel. 12. Reference curve, which gives three [hase corona los, was converted from per-phase meanirerenta Dimensions 42.5 feet average height, 89 feet flat configuration ‘This conductor was celeoted for use on the Swish 380 ky eystom. Criginal suthor probably selected a wotve fair-wouther condition ‘han the write did in ploting carve 4, which eould acsount for tha closeness, Curve 62/108/28.7° (Stranded shaman stel). fant given, Ret 13, Plotied curve is average of tivo single-phase fsir-meather trves, converted to three-phase loa Tor 46-fo0t Hat configurstion. See Curve 7. (Curve 7—2/1.04/157" (Stranded alaminarnated). 8 vot given, Ret 1% Plotted curve ie average of Uo single-phase fair-woather carves converted to three-phase lass for 45 foot fat configuration Data for curves 6 and 7 ware taken st came time inorder to ahow ‘ect of sub-eonduetor separation. *Bunute-conductor designation—number of aub-covductors/out- fe diameter of each subseondvctor in inehes/separstion between ‘adjacent sub-eonductors in inches, and Four-conductor Bundles; Alr Density Factor, 31.00, range of communication services likely to be encountered, should be known. Am evaluation of these Factors in terms of their effect on various communication services must take into consideration many things. These are available signal intensities along the line, satisfactory signal-to-noise ax tos, effect of woathor on the RI factors and on the im- portance of particular communication services, number and type of receivers in vicinity of the line, proximity of particular receivers, transfer of RI to lower-voltage cit- cuits, the general importanee of particular communication services, and means for improvement of reception at indi- vidual receiver locations” For extrachigh-voltage and double-circuit high-voltage lines the tolerable limits of RI might be higher because the number of rmeeivers affected, the coupling to lower voltage circuits, and the eoupling to roceiver antennas is reduced. Also fewer lines are required for the same power handling ability, and wider right-of- ways are used which tend to reduce the RI problem Although RI increases very rapidly with increased gradi- ent at the surface of a conductor, theoretical considerations of the radiation characteristies of « transmission line as spacing is reduced, indicate that the RI froma transmission line will not be seriously affected by reduced spacing.35] so} 4 Kn/SPHASE MILE I ovencasT, woarFRost,——] Fos I RAW Ran an o 100355 e600 in ig. 30—Corona Loss on 1.09 Inch Stranded Aluminum-Steel \ductor Under Diflerent Weather Conditions. "This con- in use on the Swedish 220-ky system. Note yariatio {in fair-wenther corona love and the relation between fair- and foul-weather corona lose, Plotted curves obtained by convert Ing per-phase measurements to three-phase values for line hhaving 32-foor Mat spacing, 80-Po0t average height. No cor- section made for air deneity factor. Ref. 22, App. A. ‘The conductor configuration, the number of clreuits, and the presonce of ground wires affect the radiation from the Tine with » given RI voltage on the conductors. Very little js known about the radiation eharacteristies of transrais- sion lines and caution should he exercised in applying dats not taker on a line configuration closely approximating the design under consideration, ‘The RI field from « transmission line varies somewhat as the inverse of the radio frequeney measured. ‘Thus serv- ices in the higher-frequeney bands, (television, frequency modulated broadcasting, microwave relay, radar, cte,) are less upt to bo affected. Dieeetional antennas which are gencrally used at these frequencies, on the average, it crease the signal-to-noise ratio, The lower signal strength, and wider bundvidths generally found in the high-fre- ‘queney bands can alter this picture somewhat, Frequene modulated broadcast is inkereatly less sensitive to RI ecause of its type af modulation, Characteristics of Aeriat Lines Chapter 8 Standard radio-noise meters##*# can measure the aver- ‘age, quasi-peak, and peak values of the RI field, The aver- ‘age value is the amplitude of the RI field averaged continu- ‘ously over 34.second, For quasi-peak measurements, cir- ‘cuit, having a short time constant (0,001-0.01 see.) for charging and a long time constant (0.8 (0.0.6 sec.) for dis- charging ia used, with the reault that the meter indication is near the peak value of the RI field. Aural tests of radio, reception indicate that quasi-peak readings interpreted in terms of broadcast-station field strengths represent more accurately the “nuisance” value of the RL field. The peak, value is the maximum instantaneous value during a given period. The type of measurements made must be known Before evaluating published RT information of misending ceonelusions ean be drawn, ‘The lateral attenuation of RI from a transmission line dopends on the line dimensions and is independent of volt= age. At distances betsceen 40 and 150 feet from the outer ‘eonduetor, the attenuation at 1000 ke varies from 0.1 to 0.3, db per foot, with the lower values applying generally to high-voltage lines. ‘Typical lateral attenuation curves are shown in Fig. 34. Lateral attenuation is affected by local conditions. Because of the rapid attenuation of RI later ally from a line, a change of a few hundred feet in the loea- tion of a right-of-way can materially aid in protecting a ‘communication service. 9, Selection of Conductor In the sclestion of @ satisfactory conductor from the standpoint of its corona performance for voltages up to 230 kv, operating experience and current practice ure the ‘best guicle, Experience in this country indicates that the corona performunce of a transmission line will be satisfac tory when a line is designed so that the fair-weather corona loss according to Peterson's forroula, is less than. fone kw per three-phase mile. Unsatisfactory corona per- formance in areas where RI must he considered has been reported for lines on which the caleulated corona loss is in excess of this value, or even less in the ease of mecliuen high voltage lines. Figure 31 is based on Peterson’s formula indicates satisfactory conductors which can be used on high-voltage lines, For medium high-voltage lines (138 kv) considerably more margin below the one kw eurve is neces- ssury becuse of the incressed probability uf exposure of re- ecivers to RI from tho line, and a design approaching 0.1 Jew should be used 10, Bundle Conductors A “bundle conduetor” is a conductor made up of two or more “sub-conductors”, and ia used as one phase condue~ tor, Rundle conductors ate also called duplex, triples, ete, conductors, referring to the number of sub-conduetors and are sometimes referred to as grouped or multiple con- ductors. Considerable work on bundle conductors has been done by the engineers of SiemensSchuckertwerke™ who concluded that bundle conductors were not economni- cal nt 220 ky, but for rated voltages of 400 kv or more, are the best solution for overkead transmission, Rusok and Rathsman® state that the inerease in transmitting eapac- ity justifies economically the use of two-conduetor bun- des on 220-ky Lines.Tae vote TTT ie ve teats ais ts recurs te srenos— Characteristics of Aerial Lines 61 ea zo 2 stnanos. ———~uoen or staanos futsie LAYER Fig. 81—Quick-Eetlmating Corona-Loss Curves. Curves based on Peterson's formula with a few check points from the Carrol and Rockwell paper for comparison. ‘The advantages of bundle conductors are higher diseup- tive voltage with conductors of reasonable dimensions, re- duced surge impedance and consequent higher power capa- bilities, and less rapid inerease of corona loss and RI with increased voltage."*"4* These advantages must be weighed against inereased circuit cost, increased, charging kva if it cannot be utilized, and uch other considerntions as the large amount of power which would be carried by one cir. cuit. It's possible with a two-conduetor bundle composed ‘of conductors of practical size to obtain electrical charac teristics, excepting corona, equivalent to a single conductor up to eight inches in diameter ‘Theoretically there isan optimum sub-conductor separa- tion for bundle conductors that will give minimum crest gradient on the surface of a sub-eonductor and hence high- est disruplive voltage. For a two-conductor bundle, the separation is not very critical, and it is advantageous to use a larger separation than the optimum which balances the reduced corona performance and slightly increased eir- cuit cost against the advantage of reduced reactance. Assuming isolated conductors which are far apart eom- pared to their diameter and have @ voltage applied be- ‘tween them, the gradient at the surface of one conduetor is given by: 79) ‘where the symbols have the same mesning as used in Eq. (792). ‘This equation is the same as equation (79a), except that surface factor, m, and air density factor, 6, have been ‘omitted. These factors should be added to Eys. 80 and 81 for practical calculations. For a two-conduetor bundle, the ‘equation for maximum gradient at the surface of « sub- ‘conductor is: qunt(tt2r/8) loge 2. (80) Poe where: Sssseparation between sub-conductors in centimeters. Because of the effect of the sub-eonductors on each other, the gradient at the surface of a wub-conductor is not uniform. It varies in a cosinusoidal manner from s maxi- nim at a point on the outside surface on the line-of-een- ters, toa minimum st the corresponding point on the inside surface. This ellect modifies the corona performance of a bundle conductor such that its corona starting, point cor responds to the voltage that would be expected from calcu ations, but the rate of increase of eorona with increased voltage is less than for a single conductor. This effect can bbe seen by comparing curve 6 of Fig, 28 with curve 2 of Fig. 29. Cahen and Pelissier™ concluded that the corona performance of « two-conductor bundle is more accurately indieated by the mean between the average62 Characteristics of Aerial Lines Chapter 3 — 000 roo eco Te 9 00 1 ooo Ee 2000} rE { 'g8| : BE we loo 0 e3| 800 po feo 4 = t—-—- iS s| & 3 Be| Eo Sooo I Fe 400 ao 8 i ido 3 i) | : 7 | | i 2 | [i 100} i a to \ a 4 re 100" ‘200 Fig, 39—Fae-Wenther Radio:Touence Piel fom a Tram i Talsin Line av Function ot ettage: Menutcrents se ds | Spposie midspan on the Se-br Covington-Grand Cou 5 Soweto | BMoNox ur che eames Romer Adminttectons RU vetoes KILOVOLTS: Fig. 22—Radio influence and corona foss measurements made ‘08 ant experimental tert line, Ref. 26 and maximum gradient at the surface of 2 sub-conductor, ‘whichis given by: e(ltr/S) D. 1) Bro Vis If it is desired to determine the approximate disruptive ving ends, gn ats( 14222) ne pe ct ler can ete erga pte aed for ¢ in ky rms, This value neglects air density Factor les in nant tow 050 os Mice onl os man ete olson SO etn sg Uocoutos ae or under eonsideration in Sweden, France, and Germany Carve [Average lateral ettemuation for «mimber of tansnision Tines from 199-to 450kv,, OX ACT sxe plotted values which apply to thi curve only, Toit frequency 1000 ks. ef. 21. Curve 2—Lateral Atienuation from the 220-ky Bgutoo-Chefngy line i Prance. Lin has erilaterel spacig, but diinensions not given. Divtance measured from middle phase. Test frequency—-S08 ke. Rel 26, ‘Carve Lateral Attenuation fom 20-kv Midway-Columbia Line ‘of the Bouneville Power Administration. Conductor height 47.5, fee, teat frequency 850 ke. Ref. 42, 1,108 inch AGSR conductor, 27-foot flat spac ‘ing, foot helght, cst frequency —800 ke. PERCENT OF RI 35 o 725 3 HORIZONTAL DISTANCE FROM OUTSIDE CONDUCTOR-FEET Fig, 4—Lateral Attenuation of Radio Inftuenco in Vicinity of High-Voltage Transmission Lines.‘Chapter 8 REFERENCES Line Conduotors—Tidd 500-kv Tost Linos, by B. L. Peterson, D.M, Simmons, L, F. Hickauell, M.E. Noyes ATEE Paper 4728 ‘Symonerical Componente, (a boald), by C.F. Wagner and R. D, Evans, MoGraw-IUll Book Company, 163, . Rodcing Taductaaco on Adjacent Tranamission Cireuts, by HB. Dwight, Blearical World, Jan. 12, 1924, 89. |. Eletie Power Tranemivsion (a book), by L. F. Woodruff. Joha ‘Wiley snd Sans, Ino, 1988. 5, Hlucticel Transmission of Power and Signals (abock), by Eaward wo. ML 2. 18 a as. aL, 2. m. W. Kimnberk John Wiley and Sons, Ino, 1940. Heating and Curront Carrying Capacity of Bare Conductors fot Outdoor Service, by O. f. Sehurg and CW. Fuck, Gener Electric Review Value 38, Number 2, March 1890, 142 Hy-Therm Copper-—An Improved Ovorkead-Linc Conductor, by LF Hickernell, A, A Jones, C.J, Snyder. ATEE Paper 40-3. 5. Elarieod Characteristica af Trovamsasion Circuit, (m book), by W. Nesbit, Wostinghouse Technical Night School Press, 126. . Resiatence and Reactance of Commercial Steel Conductors, by Prof. HB. Dwight, Bledrie Journal, January 1910, page 2 Dislctre Phenomena in High Voltage Bngineeriag (Book) F- W. Peek, Jr. MeGraw-Flil Book Co, Ine. New York, 1920 Corona Loss Measurements on « 220-KY 60-Cyelo Three’ Phase Experimental Line, J. 8. Caroll, L. H. Browa, D. P. Dinapol, ALLER, Tronsactine Wo. 8, 1081, pages $5, (Corona Losses from Conductors LA-ineh Diameter, J.. Carel, B. Cozens, T.-M. Blakeslea, ALB. Traneections Vol. 58, 1864, pages 1727-88 Corsa Lowes at 230 KV with One Conductor Grounded, J. Carll, D, M. Simmons, AJLE.B. Transactions Vol. $4, 1995, agen 846-7, Enspitical Method of Caleulating Corons Toss from High-Yalt- ‘age Transmision Lines, J. Carrol, M. M, Rockwell, AWE. ‘Tronazcions Vol 56, 1037, page 558. Corona Lost Meaatrements for the Design of Transmission ince to Oporata at Voltages between 220-KV and 330-KV. J 8. Grol, B. Corsens, ALB.E. Trancactions Vol. 52, 1888, pages 85-62, 3. Development of Corone Lacs Formula (iseusion of reference 15), W.S. Peterson, 4 J FE. Transactions Vol 52, pages 62-2. 7. Now Techniques on the Anacarm—Eletzio-Analog Compater, L- Harder, JT, Carleton, ALE Techniaal Paper 0-85 Bin neues Verluntgesete dor Wecbeclpannangakovona, H. Prins, Wise, Veroff, Sioment Schuekertwerke A.G.—Vol. XIX, July 2, 1940 Desert Measurements of Corone Loss on Conductors far Opert- tion above 280 KV, W.8, Peterson, B, Coston 8, Carol, a2 Convention Pasadena, Calif, June 12-16, 1950, letsio Power at Extra High Voltegos, Philp Spoon, A. C. Montuith, ALILB, Transactions, Val. 06, EMT ages 1S7I-7, dis. 1582 Progress Report on 600-KY Trot Project of the American Go snd Eleetie Company-—Corona, Radio Infence, and Other Factors. Philip Spom, A. C. Monteith, se prevented ATBE Convention, Pecadona, Calif. Sane 12-16, 1950. ‘The Swish SS0LIV System, W. Borgguist, &. Vrethon, see alo Appadix, A. B, Heaning 8. Skagerind, CIGRE paper 412, 1038 senion, June 24 to July 8, Confarenee Internationale det Grands Reseaux Bieotriques a Haute Tension. nflance, er MHet de Couronne, du Diarotte ot du Profi dos Cables des Liguos Anrininosa Tr Hasty Tens, P-Caben, Polis, Resa Generale de U Blatt, Val. 58, pugns 279M, ‘Leinplol de Conductnare 20 Faiveeaux pour Lv Armernt dus Lignes a Tris Hote Tension, F, Caben, R Pelssier Bull. Soe Fronguive dee Blecriciens, Gt Sarieg, Vol. VILE, Ne. 79, 148, Characteristics of Aerial Lines 63 28. Rachershes Experimentales sur le Comportement das Candice tours des Lignes « 400 KV, T. Caheu, Bt Pelissies, Ball. So. Foangtiee des Blevrcions, 6th Series, Vol. IX No, 9, Doe. 1900, Mecenisme de 1Eet de Couronne wa les Tignes de Transport Energie en Courant Alteraif, R. Peliier, D. Reneudin Bull Soe, Frangois ds Elecreions, 6th Seis, Vol. 9, Feb. 1080, Bundelleitungee, W. ¥. Mangoldt, P. Basernann, A. Buerkli, G. Markt, P. 1. Kromer, Siemens Schuekeriwevse, A. Q. pam ‘hlet, Berlin-Sierensstns, 1082 400-KV ‘Tranamission Linea with Special Reference to Multile (Conductor Lines (Bundelletangen), British Intelligence Objec- tives Sub-committea, Final Report No. 188, Tem No. 3, 8.0, Code—No, $1-8275-88, Technical Information and Docuents ‘Unit 40, Cadogan Square, London 8.W-1 Eogana. Dreletomferausberteagung mit Bundeleliern, G. Markt, B. Mencele, Bletiratcinik und Maschinenbo, 1952, page 298. Die Wieehafiche Bemessung von BundelliterLeitungoa. Elektrotechnik wnd Maschinen, 1885, page 410, {500-KV Experimental Station at Chevilly: Use of Bundle Con- ductors; Corona Effects; Clearances, P. Alive, B. Cahen, Con. Tot. des Grands Re. Electr, a Haute Tension (CIGRE), 198, peper No. 10. Relative Surface Voltage Gradients of Grouped Conductors, M. Temothok, 4.J.B.8. Transastions Vol. 67, Part I, pages 1585-0 Discusion of Referce 82 by C.F, Wagner, ALE. Transuce ona Vel. 61, Part I, page 1290, ‘Three-Phase’ Mnitple-Condhictor Cireait, E, Clarke, A.DE.E. ‘Transactions, Vol 5, 1932, page 808, Appendix C by 8. Crazy. Methods of Meacuring Radio Noiee 1040-4 report of tho Joins Coordination Comittee on Radio Reception of EET, NEMA, sand RMA Proposte American Standard Specifeation for a Rado Noise ‘Meser—0.015 t0 25 mepwiycle. Oct 189 (Published for one year tial use). ‘Television Interference Seldom Comes from Power Systems, F L. Groene, Bletrcal Wort, an. 16, 1960, pages 85-9. ect of Radin Frequencian of a Power Systein in Radi-Re ceiving Systems, C. V. Agger, W. E. Paks, W. A. Stoke, ALIE.B. Transactions, Vol 62, 34, pages 108-72. ‘Measurements Pertaining ta the Conrdination of Hadi Recep= tion with Power Apparatus and Systems, C. M. Foust, C. W. Frick, A.1.E.B. Tromsations Vol 62, 1043, pages 284-01, die, 458 asia Intererenes Suppresion ia Canada, HO. Nertimsa, AIEE paper No. 47-140, “Result of Tests Carried out atthe 500-kv Experimental Station ‘of Chovlly (France), Especially on Corona Behavior of Bundle Conductors, P. Cabin, -LE.B, Trantactions, 1988, Vol. 67, Part IT, pages 1118-25 Radio-Noise Tflucnce of 250-KV Lines, H,L. Rordeu, A.LB.E. ‘Traneacion, Vol. 65, 1947, pages 677-8: die, 682 io In uence from High Valtage Corona, G. Tt Slemoo, ATES anor No. 8-60, ‘Coniparatia Invetigation of D.C. and A.C-Corona om ‘To- Gondustor ‘Transmisson Lines (in German), Ie Strige, Wise chose Verectenilangan Aus Den Siomone-Workon, Vo. 18, Patt 2, 1938, pages 68-01, ‘Tho Swish 380 KV System, A, Rawok, Bo G, Rathsman, Ble ricl Brgincering, Doo. 149, pages 1025-0. Series Capacitor tad Double Condctons inthe Swedish Trans nission System, A use, Bo G, Ratheman, Electrical Pnginser- ng, Jaa. 1050, pages $87 Eiflect of Eardhing on Carona Lasss, Conductor Diameter And Length of Tasulatoe Strings, The Brown Boveri Reciaw, Vol. ZEXXV Nos, 7/3, July/August, 1048, pages 102-201 ‘The Tronsmisson of Blecrie Power (a book), by W. A. Lewis, (918 Lithopriie! Edition) Uinoia intents of ‘Poshoology. 26, 28 29, 30. 3 38, sr. 2 8. a 45, 46.CHAPTER 4 ELECTRICAL CHARACTERISTICS OF CABLES Original Author: H. N. Muller, Jr. |ABLES are classified according to their insulation as ‘paper, varnishes-earabrie, rubber, ar asbestos, each of theso materials having unique characteristics which render it suitable for particular applications, Be- ceause cables for powor transmission and distribution are composed of so many different types of insulation, con ductors, and sheathing materials, the diseussion here mast be limited to those eable designs most commonly used. Reasonable estimates of electrical characteristis for eables not listed ean be obtained in most cases by reading from the table for # cable having similar physical dimensions. Paper ean be wound onto a conductor in succossive lay- ers to achieve u required dielectric strength, and this is the insulation generally used for cables operating at 10 000 volts and higher. Paper insulation is impregnated in dif- ferent ways, and secordingly eables 60 insulated ean be sub-divided into solid, oililled, or yuefiled types. Solid paper-insulated cables’ are built up of Inyors of paper tape wound onto the conductor and impregnated with @ viscous oil, over which is applied a tight-fitting, extruded lead sheath. Multi-conductor solid cables are also available, but the material shown here covers only single- and three-conductor types. Three-conductor cables are of either belted or shielded construetion, ‘The belted assembly consists of the three separately insulated con- ductors eablod together nnd wrapped with another layer of Iimprognated paper, or belt, before the sheath is applied. In the shielded construction each conductor is individually ingulated and covered with a thin metallic non-magnetic shielding tape; the three conductors are then cabled to- ether, wrapped srith a metallic binder tape, and sheathed with lead, ‘The purpose of the metallic shielding tape around each insulated conductor is to eontral the electro- static stzess, reduce corona formation, aund decrease the thermal resistance. To minimize circulating eurrent under normal operating conditions and thus limit the power loss, shielding tape only three mils in thickness is used. Solid single-conductor cables are standard for all voltages from, 1 to 69 kv; solid three-conduetor eables are standard from 1 to46 kv. Sample sections of paper-insulated singlo-cou- ductor, three-conduetor bolted, and three-conduetor shielded. ceables are shown in Fig. 1(a}, (b), and (c} respectively. Oilsfilled paper-insulated eahles are available in single- or three-conductor designs. Single-conductor oil-filled cable consists of a concentric stranded conduetor built round an. ‘open holical spring core, whieh serves as. channel for the flow of low-viscosity il. ‘This cable is insulated and sheathed in the same manner as solid cables, as 2 compar ison of Figs. 1(a) and 1(@) indicates, ‘Three-conductor oil- filled cables are all of the shielded design, and have three 5 Revised by: 1.8. Williams () Single-eonductoreoid, ompact-round conductor. () ‘Thieoconductor belted, compactaector co (©) Three-conductor chet, compuct-sector conductors, Seoenias (@) Single-conductor oiled, hollow-strandnd sondutor, (@) Theor conductor ailfiled, eompact-eetor conductons, Fig. 1 Paper-insulated cablee. il channels composed of helical springs that extend through the cable in spaces montoally oecupied by Aller torial. This construction is shown in Fig. 1c). Oil- filled eablos are relatively new and their application has become widespread in a comparatively short time, The oil used is only slightly more viscous than transformer oil, and.Chapter 4 remains fluid at all operating temperatures, ‘The oil in the cable and its connected reservoirs is maintained under moderate pressure so that during load eyeles oil may ow between the eable and the reservoirs to prevent the devel- opment of voids or excessive pressure in the cable. The prevention of void formation in paper insulation permits the use of greatly reduced insulation thiekness for @ given, operating voltage. Another advantage af oil-filled cables is, thet oil will soep out through any’ eraek or opening, which develops in the sheath, thereby preventing te entrance of water at the defeetive point. This action prevents the eenrtence of a Fault caused by moisture in the insulation, and since operating reeords show thal this cause accounts fora significant percontage of all high-voltage cable fuulls, it is indeed a real arlvantage. Single-conductor oil-filled cables are used far voltages ranging from 69 10 230 ky’; the cususl range for three-conductor oil-filled eables is from 23 t0 69 kv Gas-filled eables of the low-pressure type have recently become standard up to 48 ky. ‘The single-eonductor type employs construction generally similar to that of solid cables, except that longitistinal utes or other channels ace provicled at the inner surface of the sheath to conduet nitrogen along the cable. ‘The three-conduetor design em ploys channels in the filler spaces among the conductors, much like those provided in. oil-filled three-conductor cables, ‘The gus is normally maintaincd bobween 10 and 18 pounds per square inch gauge pressure, and serves to fil all cable voids and exelude moisture at faulty points in the sheath or joints, Fig. 2- High-pressure pipe-type oll-flled cable. High-pressure cables, of either the oil- or gas-filled vari- ety, aro being used widely for tho higher range of voltages. ‘The physical and electrical characteristies are fairly well known, but their specifications ar not yet standardized, ‘The usual application ealls for pressuse of about 200 pounds per square inch, contained by a sicel pipe into which three single-conductor cables are pulled. ‘The ianmodiate prose ence of the iron pipe makes difficult the calculations of circuit impedance, particularly the zero-sequenee quanti= ties. Most high-prosstre eables are designed 80 that the oil or gas filler comes into direct. comtact with the conductor insulation; in oil-filled pipe-type cables @ temporary lead sheath ean be stripped from the eable ast is pulled into the steel pipe; in ges-flled pipe-type cables the leed shewth sure rounding eweh eonduetor remains in place, with nitrogen introdueee hoth inside and outside the sheath so that no Aiffcrential pressure develops across the sheath, Examples of oit- and gasfiled pipe-type cables are shown in Kigs. 2 and 3. Blectrieal Characteristics of Cables 6 oti of Genet Cte Carton 3 Gross-section of high-pressure pipe-type gas-filled ie. Oileilled ppe-type cable may have similar crose= esetion. Compression cable is another high-pressure pipe-type cable in which oil or nitrogen gas at high pressure is intro duced within a steel pipe containing lead-sheathed solide ype single-eonductor exbles; no high-pressure oil or gas is introduced directly inside the lead sheaths, but voids with- in the solid-type insulation are prevented by pressure exerted externally on the sheaths. This construction is sketched in Fit 4 ‘During recont Years there hus been a trend toward th modification of eable conductor to reduce cost and im- prove operating eharaeteristies, partienlarly in rulti-con- Guctor eables, Referring to Fig. 5, the first departure from. concentrie round conductors was the adoption of sector shaped conductors in three-conductor eables. More ro cently a crusted stranding dat results in x compacted sector has been developed and has found widespread use for conductor sizes of 1/0 A.W.G. and larger. Tta use in smaller evuductors is not practical. ‘The principal advan tages of such a conductor are: reduced overall diameter for ‘a given copper eross-seetion; elimination of space between ‘the conductor end the insulation, which results in higher ‘Fig. 4Grose-aectional sketch of compression cable,66 ay “Flt as refrain he Oko lint Orie Coton Fig, 5—Cable conductors, (a) Standard concentse stranded, () Compact rand (0) Non-compact sector, (A) Compact sector, (6) Annilar stranded (rope core), (Segmental (Rope strand. (h) Hollow core. electrical breakdown; low a- resistance due to minimizing of proximity effect; retention of the close stranding during bending; and for olid cables, elimination of many lon situdinal channels along which imprognating compound ean migrate. While most single-conductor eables aro of the Electrical Characteristies of Cables Chapter 4 concentrie.strand type, they may also be corapact-round, annulur-stranded, segmental, or hollow-core, I. ELECTRICAL CHARACTERISTICS ‘The electrical characteristics of cables have been die cussed comprehensively in a series of articles? upon which much of the material presented here has been based. ‘This chapter is primarily concerned with the determination of the electrical constants most, commonly needed for power- system ealeulations, purticular emphasis being placed on quantities necessary for the application of symmetrical components." A general rule is that regardless of the plexity of mutual inductive relations between earaponent parts of individual phases, the method of eymmetrical components can be applied rigorously whenever there is symmotry among phases. All the three-conductor cables inherently satisfy this condition by the nature of their con struction;singlo-conductor cablesmay ormay not, although ‘usually the error is small in calculating short-circuit cur- s. Unsymietrical spacing and change in permeability resulting from different phase eurrents when eertain meth- ods of eliminating sheath currents are used, may produce dissymmetry. ‘Those physical characteristics that are of general inter cst in eleetrieal application problems have been ineluded along with electrical charaeteristies in the tables of this cation, All linear dimensions of radins, diameter, separation, or distance to equivalent earth return are expressed in inches jn the equations in this chapter, This is unlike overhead ‘transmission line theory where dimensions are in feet; the use of inches when dealing with cable construction seems appropriate. Many equations contain a factor for fre- ‘quency, f, which is the eireuit operating frequency in eycles per second 1. Geometry of Cables ‘The space relationship among sheaths and conductors in a cable eirenit is » major factor in determining reactance, capacitance, charging current, insulation resistance, dieloo- ‘rie loss, and thermal resistance. ‘The symbols used in this chapter for various cable dimensions, both for single-con- ductor and three-conductor types, are given in Figs. 6 and 7. Several factors have come into universal use for defining the eross-seetion geometry of a cable circuit, and some of ‘these aro covered in the following paragraphs. Geometric Mean Radius (GMR)—This factor is a property usually applied to the conductor alone, and de- pends on the material and stranding ueed in its construe tion, One component of conductor reactance? is normally calculated by evalusting the integrated flux-linkages both inside and outside the eonduetor within sn overall twelve- radius. Considering a solid conductor, some of the ux nes lie within the conduetor and eontribute to total Aux nkages even though they link only a portion of the total conductor current; if a tubular conductor having an inf- nitely thin wall were substituted for the solid eonduetor, 3 flux would necessaeily all be external to the tube. A theo- retical tubular conductor, in order to be inductively equiv= alent to a solid conductor, must have a smaaller radius 20Chapter 4 MEAD SHEATH [tbe inst asion CONDUCTOR INSULATION, svc < (o) SINGLE-CONDUCTOR CABLES CQUILATERALLY SeaceD. THREE-CONDUCTOR CABLE, PHASE ¢ SNGLE-cONDUCTOR CABLES LUNSYWMETRICALLY SPACED, BUT PERFECTLY TRANSPOSED Fig. 6-Geometry of cables. that the fux-tinkages present inside thesolid eonduetor but absent within the tube will be replared by additional link ‘ages between the tube surface and the limiting eylinder of ‘twelve-inch radius, A solid eopper eonduetor of radius d/2 ean be replaced by # theoretieal tubular eonduetor whose radius is 0.779 4/2. This equivalent radiasiscalled the geo metric mean radius of the actual conductor, denoted herein by GM. where the subscript denotes reference to oaly a single actual conductor. ‘This quantity ean be used in re- actance calculations without further reference to the shape ‘or make-up af the conductor. ‘The factor by whieh actual radius must be multiplied to obtain GMVse varies with sector DEPTH — 4 Fig. 7—Typlcal sector shape of conductor used In three-con- ‘Suctor cables, stranding or hallow-eore construction as shown in Chap. 3, Fig. 11. Sometimes in calculations involving zero-se- quence reactances, simplification may result if the three conductors comprising a three-phase circuit are considered as 8 group and converted to a single equivalent conductor, ‘This requires the use of a new GMR, demoted here as Electrical Characteristics of Cables er GMRe, which applies to the group as though it were one ‘complex conductor. This procedure is illustrated later in Eq, (18). ‘Geometric Mean Distance (GMD) Spacings among ‘conductors, or between eonduetors and sheaths, are impor= tant in determining total cireuit reactance. ‘The total flux Linkages surrounding a conductor ean be divided into two components, one extending inward from 2 cylinder of 12- Jhvadiusas discussed in the preceding paragraph, and the other extending ontward from this eylinder to the enrrent return path beyond which there are no net flux-linkages? The flulinkages per vnit, conductor current between the 12inch eylinder and the return path are & function of the separation between the conduetor and its return, ‘The re- turn path cnn in many eases be a parallel group of wires that a goomeirie moan of all the separations between the conductor and each of its returns must be used in ealeul- tions. Geometric mean distanee, therefore, isa term that ‘ean lie used in the expression (or external fuxlinkages, ot only in the simple ease af wo adjacent conductors where it evjual to the distance between conductor centers, but also in the moro eomnplex ease where Lwo cireuits each composed (of several conductors are separated by an equivalent GMD. ‘The positive. or nogative-sequence reactance of a three ‘phase cireuit depends on separation among phase condue- tora. Ifthe conductors are equilaterally spaced the distance from one conductor center to another is equal to the GMD. ‘among conductors for that circuit. Using the terminology in Fig. 6, enn. for an oquilateral eireuit ‘The subscript here denotes that (is GMD applies to sepa- rations among three conduetors. TF the conductors are arranged other than equilaterally, but transposed along. their length to produce a balanced circuit, the equivalent separation may be calculated by deriving a geometric mean, istance from the eube root of three distance products" (eee Chap. 3) GMD2.= VSa Sie Se a) ‘The component of circuit reactance caused by flux outside fa tivelve iuch radius is widely identified as “reactance spasing factor” (x4) and ean bo caleulated directly from, the GMD: I jg, MD i 240.2704 Flog I2* ohne per phase per mile, (2) When the equivalent separation is less than twelve inches, aus eu occur in cable cireuits, the reaetanee spacing factor fs negative 20 a8 to subtract from the carmpenent of con- ductor reactance due to fhus out to a bvalve-inch radius “The zero-sequence reactance of « three-phase circuit may. depend on spucing among conductors and sheath ts well 8 aniong conductors. A distance that represents the equiv= alent spacing between a conchctor ora group of conductors sn tho enclosing shenth can be expressed ae GMD. Also, the equivalent separation between eable eonduetors and the sheath of a nearby cable, oF the equivalent separation between two nearby sheaths, ean be expressed as & GMD. Because these and other versions af geometzie meats dis- tance may bo used successively in a single problem, care68 ust be taken to identify and distinguish among them dur- ing ealculations, Geometric Factor—The relation in space between the cylinders formed by sheath internal surface and conductor external surface in a single-conductor lead-sheathed cable can be exprossed as a “geometric factor.” This factor is applicable to the calculation of such vable characveristies 5 capacitance, charging current, dielectric loss, leakage current, and heat transfer, because these charseteristies, Gepend on a field or flow pattern between conduetor and. sheath. ‘The mathematical expression for geometric factor Gin a single conductor cable is @ tsido diameter of conductor, Geometric factors for single-conductor cables can be read. from Fig. 8. Geometric factors for three-phase shielded cables having round conductors are identieal, except for heat flow esleulations, to those for single-eonduetor cables. The shielding layer establishes an equipotential surface surrounding each concuctor just as a lead sheath does for single-conductor cables, ‘The heat conductivity of the ‘three-mil shielding tape is not high enough to prevent a temperature differential from developing around the shield ‘circumference during operation: this poses & more complex problem than can be solved by the simple geometric factors Electrical Characteristics of Cables Chapter 4 Because of the various possible combinations of con- ductors and sheaths that can be taken in a Uiree-conductor belted eable, several geometrie factors are required for com- plete definition. Two of these factors, the ones applicable to positive- and to zero-sequence electrical calculations, are shown in Fig. 9, 2. Positive- and Negative-Sequence Resistance Skin Effect—Iv is well known that the resistance of a. conductor to alternating current is anger than its resistance to direct current. ‘The direct-current resistance in eablos can be taken as the resistance of solid rod of the same length and crost-section, but increased two pereent to take into account the effeet of spiraling of the strands that com= pose the conductor. When alternating current flows in the ‘concluctor there is an unequal disteibution of current, with the outer filaments of the conductor carrying more current than the filaments eloser to the center. This results in & higher resistance to alternating current than to direct cur- rent, and is commonly called skin effect. The ratio of the ‘two resistances is known as the skin-effeot ratio, In small conductors this ratio is entirely negligible, but for lerger conductors it becomes quite appreciable, and must be eon- sidered when figuring the 60-cycle resistances of large cou- "Tanta 1—Davenstons asm 60-CyctR Sens-Rerace Ranta oF ‘STRANDED CoPPER CONDUCTORS AT B5°C. cman | Tener Distro Arve ven her, Round | Btrandad Conductor, inches ® Sie | ConoentrieStranded > 0.50 0% (Cresiar Be Mil) | Diameter Outer po | Outer | inches Diam P49 Diam, Ratio . { Bio | 0s —- 200m | O57 2 { 01000 | 0.639 © 400.000 | 0.728 Boa sooo | ost 8 fsmoo0 | 0.803 300000 | 3.031 z, 1.000.000 | 1.182 bos 1.500.000 1412 4 2000000 | a1 3 2500.00 | 1.895 8000000 | img u od | ductors. Some skin-offect ratios are tabulated in Table 1 i (SHIELDED) a nario F Fig. &Geometric factor for alngie-conductor cables, of ‘threo-conductor shielded cables having round conductors, NOTE: This is approximately conect for shielded sectorconductor cable if eurvo is entered with the dimensins of a round-esndcton fable having identical conductor am and fasulaion thickness, “This geometric factor is not applicable for heatlow calculations ia shielded cablea. See Sees. 6 and 6. for stranded and representative hollow conductors. Proximity Effect—The alternating magnetic flux in conductor caused by the current flowing in a neighboring ‘conductor gives rise to circulating currents, whieh eaue an apparent increase in the resistance of a conductor. This phenomenon is called proximity effect. The increase in resistance is negligible except in very large conductor. Proximity effect ean, however, become important under certain conditions of eable installation. When eables are laid parallel to metal beams, walls, ete., sis frequently the cease in buildings or ships, proximity effect increases the apparent, impedance of these cables appreciably. Booth, Tlutchings and Whitebead® have made extensivChapter 4 Electrical Characteristics of Cables 69 6, FoR Tanee-cowoucTOR BELTED caBLes oe Fon Yomeeconwcton BELTED CABLES “GEOMETRIC FACTORS (6,06) FOR THREE CONDUCTOR BELTED CABLES iu Rarer || ~ ~ | : | pa i tare Fig. 9 Geometric factor for three-conductor belted cables having round or sector conductors. NOTE: For cable having scctor coaductors enter tho curve with the dim#nsions ofa round-conductor cable having identical conductor ‘ra snd intulation thicknowes. Multply the resultant goometre factor by the sector ecrretion facior given ebove. G. is eaoulsted for three-phase operation; Gy is ealeulatd for single-phase operation, with thes conductors. paralleled and se ‘urn in wbeath. See Bees. 5 and 6)0 the impedance and current-carrying capacity of eablos, ag they aro affected by proximity to fiat plates of conducting aand magnetic material. Figures 11 and 12, taken from this work, illustrate forcefully that prosimity effect can be significantly lange. Although these tests were performed at 50 eyoles itis believed that the results serve Lo indicate effects that would be experienced at 60 cyeles. ‘The result in an actual installation of cables close to metal surfaces are influenced so greatly by the material involved, and by the l T T & GEOMETRIC FAGTOR,G Electrical Characteristic of Cables Chapter 4 0 — Foo nen spans srvwery ears Fol gcimges amines : = ENS SSSR 2 oi | ys oo T : DISTANCE FROM STEEL. PLATE NCHES crease in cable resistance and reactance caused ty to tect plate for ‘sheath wre i terials will contribute lowest losses. Some practical pos- sibilities, drawn from experience in the design of switeh- sear, transformers, and generators, aro listed hore: a. The magnetie plate can be shielded by an assembly of Iaminated punchings, placed between the cables and the plate, £0 that flux is diverted from the plate and into the laminations. The laminations normally have ow edily-current losses and they must be designed so that Mux density is not excessive b. The magnetic plate ean be shielded with « sheet of conducting material, such as copper or aluminum, placed so that the magnetic field acts to build up G04 008 OIE aE 20 ratio F Fig. 10—Geometric factor for three-conductor shielded cables hhaving sector conductors, in terme of ina ‘and mean periphery P. tion thickness T structural shape of the eurface, that calculation and pre- diction is difficult. ‘The additional losses eaused by placing a metal plate or other structural shape close to @ cable circuit arise from both hysteresis and eddy-current effects within the plate, Hysteresis losses are large ifthe fluxclensity within the plate i high throughout « large proportion of the plate volume, A material having high permeability and very high re Sstivity would promote hysteresis loss, because lux devel- ‘oped by cable eurrents could concentrate within the low- reluetanee plate, and because the action of eddy-current to counteract the incident flux would be comparatively small in a high-tesistance material. Eddy-current losses depend on the magnetic field strength at the plate, and also upon the resistance of the paths available for within the plate. Because the factors that affect hysteresis loss and those that affect eddy-current lass are interdependent, itisseldom ‘easy to theorize on which material or combination of ma spacmc BETWEEN suEaTis (3a. ot 908} 7: ars a ! : Bs T ° 0s}: i ah INCREMENT OF RESISTANCE AND REACTANGE-OHMS PER MLE DISTANCE FROM STEEL PLATE INCHES Fig. 12—Increase tn cable res by prozimity to steelChapter 4 the con- counteracting cireulating currents within ducting sheet: these currents considerably reduce the ‘magnetic field strength at the plate. ‘The conducting sheet must have sufficient cross-sectional area to ae- commodate the currents developed. ‘e. The magnetic material ean be interleaved with con- ducting bars that are bonded at the ends so that cir- culating currents develop to counteract the incident magnetic field as in (b). 4. The magnetio plate can be replaced, either entirely or partially, by © non-magnetic stesl. Non-magnetic steel has low permeability and high resistivity when compared with conventional steel plate: these char acteristies do not act in all respects to reduce losses, Dut the net effect is often a loss reduction, Non-mag- netic stec! is of parti fit when the structure or entirely surrounds ‘The effect of parallel metal on reactance is much larger than on resistance as Figs. 11 and 12 indicate. These figures also show that the magnitude of the increase in im- peilance is independent of conductor size. Actually, when large eables approach very elaze to sec) the resistance in- crements become higher and the reactance inerements be- come somewhat lower. ‘The eurves of Figs. I and 12 are bused on tests perforiued al approsimmately wo-thinds of rmaximanm current density for each eablo used. The incre ‘ments in resistance and reactance do not, however, change sreatly with current density; the variation is only about T percent per 100 amperes. In three-phase systems the nnildle cable of the three is influenced less than the outer ies by the presence of the paralicl stoel. ‘This variation ‘again is less than variations in materials and has not heen accounted for in Figs. 11 and 12. ‘These curves cover only a few specific eases, and give merely an indication of the im portance and magnitude of prosimity effect. More detailed information ean be found in the reference listed. Proximity effect algo haa an important bearing on the current-carrying capacity of cables when installed neat steel plates or structures. ‘This subject is diseussed in the section on current-carrying capacity. ‘Sheath Currents in Cables—Alternsting curzent in the conductors of singlo-conductor eables inden aleernat ing voltages in the sheaths. When the sheaths are contin uous and bonded together at their encs so that sheath eure rentsmay low longitudinally, additional FR losses develop inthe sheath. ‘The common way to represent these lases i by incroasing the resistance of the conductor invelved. For single-conductor eables operating in threphase avstems, this increment in resistance can be ealeulated by the fol lowing equstion, the derivation of which is given in refee- ences:!* yigohms per phase per mile. w Here 2m is the mutual reactance between conductors and sheath in ohms per phase per mile, and r, is the resistance of the cheath in ohms per phace per mile. ‘These twa quan- tities ean be determined from the following equations: 2794 5, tog 2 ot Zh ile. (6) Ag 8 5, Om por phase por mile. (6) Electrical Characteristics of Cables m1 and 0.200 : a= 0200 ts por phase peril for lead shesth, Get nesay me Pe Pie Pee @ in which pacing between conductor centers in inches, ‘outer radius of lead sheath jn inchs nner raclins of lead sheath in inches. ‘Thus dhe total resistance (r4) to positive. or negative. sequenee ourrent flow in single-conductor eables, including, the effect of sheath currents, is johns per phase per mile. (7) where rs it the alternating-current resistance of the con- ductor alone including skin effect at the operating fre- xeney. Bq. (7) applies rigorously only when the eables are in an equilateral triangular configuration. For other ar- rangements the geometrie mean distance among three conductors, GMDj, can be used instead of S with resulta sufficiently accurate for most practieal purposes. ‘The sheath loce in a three-conductor cable is usually negligible exeept for vory large eablos and then it is im- portant only when making quite accurate calculations. In these largest cables the sheath losses are about 3 to 5 per= cent of the conductor loxs, and are of rolatively little importanee in most practical caleulations. When desired the sheath foss in three-conduetor eables can be cxleulated from the equivalent resistance, 44160(8) ratotri? 10-+ ohms per phase per mile. (8) where 1, is sheath resistance from Eq. (0). Teand r, are sheath radil defined for Bq. (5). 1 . S.=14(¢4+27), and is the distance between vs ‘conductor center and sheath (9) center for three-conductor tables made up of round con- ductors, d=conduetor diameter. ‘P=eonduetor insulation thickness. For sector-shaped conductors sn approximate figure ean be had by using Ea. (8), except that d should be 82 to 86 percent of the diameter of a round conduetor having the Same cross-sectional area Bxample I--ind the resistance at 60 cycles of 2750 000 cireular-mil, three-conductor belted cable having 158 mail conductor insulation and 133 mil lead sheath, “The overall “inmeter ofthe eable is 2.88 inches and the conductors are sector shaped. From conductor tablos (aoe Table 10) the diameter of sn equivalent round eonduetor is 0.098 inches. From Eq. (9), = J4(0.998(0.81) +210.150)] =0.664 inches. Since the overall diameter is 2.833 inches, rye 1.417 inches2 Electrical Characteristics of Cables and r= 1.284 inches, From Eq. (8), 000 B71) 0.135) =0.557 ohms per phase per mile, Substituting in Eq. (8), 44160(0.664)2 o.n57¢ =0100479 ohms per phase per mile. From Table it is found that y., the conduetor resistance, including skin effect is 0.091 ohms per phase per mile. ‘The ‘otal positive- and negative-sequence resistance is then, 14=0.091-+.005=0.096 ohms per phase per mile. Sheath currents obviously have little effect on the total alternating-current resistance of this cable. ‘Theoretically some allowance should be made for the losses that occur in the metallie tape on the individual con- ductors of shielded cable, but actual measurements indicate that for all practical purposes these losses are negligible ‘with present designs and ean be ignored in mast cases. ‘Phe resistance to positive- and negative-sequencs in shielded cable can be calculated os though tho shields were not rosent. Three Conductors in Steel Pipe—Typical values for positive: and negative-sequence resistance of Inrge pipo- type cables have been established by test, and an empi ical ealeulating method has been proposed by Wiserman® that checks the tests quite closely. Because the caleula- tions are complex, only an estimating curve is presented 10-4 Sa 7 24 8H 22 y" ©. qd i | OY AF Bao (oe 20 uly gi _ a 6 Ble ga - 1 Sia g ae Soe she 5 : * pal — — eg ei | | |] 8 oh ee ‘CONDUCTOR SIZE, MoM Fig. 13_Positive-sequence resistance of high-voltage cables {in eteel pipe (estimating curve). Chapter 4 here. The ratio of actual resistance as installed to the d-e resistance of the conductor itself based on data obtained in laboratory tests is shown in Fig. 13. ‘The increased resist ance is due to conductor skin effect, conductor proximity effect in the presence of stoel pipe, and to I°R loss in the pipe itself. In preparing Fig. 13, the pipe size assumed for ‘each cable size was such that 60 percent of the internal pipe cross-sectional area would have been unoccupied by eable material: choosing nearest standard pipe size as 2 prac- tical expedient does not affect the result appreciably. The conductor configuration for these tests was « triangular grouping, with the group lying at: the bottom of the pipe. I, instead, the eonduetors were to be laid in an approx imately flat cradled arrangement, some change in resist~ ‘ance would be expected. Actual tests on the flat arrange- ment produced variable results as conductor size was changed, come tosta giving higher locees and some lower than the triangular. “If a maximum value is desired, on estimated inereaso of 15 pervent. above the resistance for triangular configuration can be used. Field tests have been :muacle on loy-voltage circuits by Brieger and these results are shown ia Table 2. 3. Positive: and Negative-Sequence Reactances Single-Conductor Cables—The reactance of single- conductor lead-sheathed cables to positive- and negative- sequence currents can be ealculated from the following. equation, which takes into account the effect of sheath currents, amet Soy, GMD __ tat sem es20.2704 66 1986 GM, ate ohms per phase per mile or (10) aeamancts, ‘ohms per phase per mile, (11) ‘The conductor component of reactance is f 12 #.=0.2704 § lowe ante aa geometric mean radius of one conductor. ‘The separation component of reactance is GMDs, 2 24-0270 E tone aay where GMD,.~geometrie mean distanoo among three conductors (see Eq. 1). ‘The component tobe subtract because of th effect of sheath curents is composed of terms defined. by. Bas Gand 6 ‘Three-Conductor Gables—Decause aeglgible seath curcentofeste are present in three conductor non-shiekded tables, the reactance to postive: and nogativoamuense currents can be calculated quite simply as: GMDs, GME, chs per phase per ile aeae0870 pn (14) (15) ~+4 ohms por phase per mileChapter 4 Electrical Characteristies of Cables 3 ‘TABLE 2—IMPEDANCE OF THREE-PHASE 120/208 Vour CaBLE CrncurTs IN FIBRE AND IN IRON ConbUtTs.t Positive and Negative-Seavence Impedance, Ohms per Phase per Mile at 60 Cycles, Phase Conductor | Duct Materat ‘Gable Sheath Resistance Resetance Bee Conductor Assembly (4 tach) (Phaan Conductors) | _ (Ohm et 26°C.) (Ohms) 0 MCR acaba bee cd Cperphas)| Ten Cabia Gabiok (000 AWG Vacabled* (2 per phase) (Gaia ‘ero-Sequence Impedance, Ohms Por Phase Por Mile at 60 Cyetes, Piase Gandwtr | Neutral Cmducor | cContctr ) Dace mesa | OES | Resse | Renton & a Saonuy | cma | fim fomectioey| “ea ‘too MC “OOD AWG | Uneabled* ‘Fibro erin) | (oman bay a ‘0 MON ce (conductor, bare) abled? 000 AWG Gable ( 0000 AWG 0000 AWG Uneabled* ‘00MCM ‘Cabled* ‘Tron jeometrie mean distance among three conductors, and the remaining values are as defined in Eys. (12) snd (13). For sector-shaped conductors no accurate data on change in reactance because of conductor shape is available, but Dr. Simmons can be quoted as authority fur the stavernent that the reactance is from five to ten percent less than for round conductors of the same area and insulation thick- oss. inh ene AT. C-ATO, ith te tr neal ennui hav cot For shielded three-conduetor cables the reactance to Positive- and negative-sequence currents ean be ealeulated 4s though the shields were not present, making it similar to belted three-conductor cable. This is true because the effect on reactance of the circulating currents in the shielding tapes has been ealculated by the method used for determining sheath effects in single-conductor eables and ‘proves to be negligible. Three Conductors in Steel Pipe—Couductor skin effect and proximity effeetsinfiuence the apparent reactance ‘of high-voltage cables in steel pipe. Beeauso the detailed,4 caleulation of these factors is complex, a curve is supplied in Fig, 14 that sorves for estimating reactance within about five pereent accuracy. ‘The eurvo is drawn for triangular conductor grouping, with the group lying at the bottom of the pipe. IF the grouping iz instead a flat eradled arrange- ment, with the conductors Iying side by-side at the bottom of the pipe, tho eurve results should bo inereased by 15 percent, Aealeulating method that secounts in detail for 2s s00 & 80 g 1200 g25|— = 00 2 ie $20! esc = e500 Zo 5 ty 3 go} Z 4 & 7 conatorne strane | z Sowenrat. sreenoeo | Bos} 1 |} 3 8 A L 51s ——"15 35 50 CENTER TO CENTER SPACING BETWEEN cABLES, McHES Fig. 14—Positive-sequence reactance of high-voltage cables {in stool pipe (estimating curve) the variable factors in this problem has been presented by Del Mar’. Table 2 emtains information uselul in esti mating the impedance of low-veltage (120/208 volt) cables in iron conduit, 4, Zero-Sequence Resistance and Reactance ‘Whon zero-sequenee current flows slong the phase con- ductors of a threesphase cable circuit, it must return in aths, oF in'the parallel combi- nation of both ground and sheaths? As zcro-sequence ont ent flows through each eonduetor it encounters the sc resistance of that conductor, and asit returns in the ground ‘or sheaths it encounters the resistance of those paths, The zero-sequenee current flowing in any one phase encounters also the reactance arising from conductor self-induetane from mutial inductance to the other tivo phase conductors, from mutual inductance to the ground sind sheath return paths, and from self-inductance of the return paths. Each Electrical Characteriatice of Cables Chapter 4 of these inductive effects cannot always be identified in- dividually from the equations to be used for reactance cal- culations because the theory of earth return circuits’, and the use of one CMR to represent « paralleled conductor group, present in combined form same of the fundamental effects contributing to total xero-sequenee reactance. ‘The resistance and reaetanee effects are interrelated so closely that they are best dealt with simultaneously. Cable sheaths are frequently bonded and grounded at several points, which allows much of the zero-sequence return current to flow in the sheath, On the other hand, ‘when any of the various devices used to limit sheath current are employed, much or all of the return current flows in the earth, ‘The methed of honding and grounding, therefore, ‘haa an effoct upon the zoro-sequence itapedaneo of cables. ‘An actual eable installation should approach one of these three theoretical conditions: 1 Retarn current in sheath and ground in parallel. 2 All return curzent in sheath, none in ground. 3 All return current in ground, none in sheath. Three-Conductor Cables—Actusl and equivalent ci cuits for a single-circuit three-conductor exble having solidly bonded and grounded sheath are shown in Fig. 15 (a) and (e). ‘The impedanee of the group of three paralleled cconiluctors, considering the presence of the earth return hut ignoring for the moment the presence of the sheath, is zziven in lags. (16) or (17) in terms of impedanee to zero- ago f De sen tere 08382 55 l080 GTS ‘ohms per phase per mile a6) semrebretitiett204) ‘ohms per phase per mile an ‘Tans 9—Equvauex7 Deprutop Exes Revenw (D,), No EAR TMPEDANCE (F. AND 2), AT 60 CYCLES Egaivalent Equivalent | Equivalent Earth Depth af Earth ‘Earth Resistivity] Earth Rotor, D, | Resistanco | Renctance meter ohe|———————| a inches | foot _ohmspermik)obmspermie) 1 | aaecio" mo) Ose 208 3 | tase | 6a | 0256 234 10 | 0G | 580) 0286 2at 50 | 20x10" | 2000) 0256 276 yaa | Ramio | Dann) 2a 2k so | 74x10" | 6200 | 0.86 a8, 1000 | locsio# | $300) 0.86 331 5000 | 240.10" | 29000 | 0280 2.00 090 | 6x10 | aen00 | 0.256 378 where: resace resistance of one conductor, ohins per nile raswace resistance of earth return (See Table 8), ‘olims per mile Dy=distance to equivalent earth return path, (See Table 3), inches.Chapter 4 Electrical Characteristies of Cables 5 GMR,,=geometrie mean radius of the conducting path made up of the three actual conductors taken 8 « group, inches. = /(GMR,J(S)! for round conductors. (18) GMR,.=geometric mean radius of an individual conductor, inches. = reaetanee of an individual phase conductor at twelve inch spacing, olims per mile. aroun det or rita erm en = 0682 Jay 2 ts peri, lero neo ae ried scrote Ztos (2422), tn ite ae “ F GMD..=ageometric mean distance among conductor oS TEE rund contin tes conductor cables, ‘The impedance cf the sheath, considering the presence of ‘the earth return path but ignoring for the moment, the presence of the conductor group, is given in terms of im- pedance to zerosequence currents: Gag tag Tag) +t 4g)20 i = carn 2D, 2.-Sretroti08362 2 logs 22% ohins per (rime saLeconDveTON GABL¢3) pphase por mil 0 oo or seaSnecbret (Beets) ohms per phase per mile, ep = heres (wsmnesenrs nites conoveron) [p, sheath resistance, obms per mile. | 0.200 . j = eGo ReO yy orto sheath r=inside radius of sheath, inches. reoutside radius of sheath, inches. y=reactance of sheath, ohms por 2 il ep ohms per mile, (22) ~ £ 0.2704 FF lore rouravenr oncor ‘The mutual impedance between conductors and sheath, weoances exrtsato eno SEOUENCE TERMS) Goniering tho prance of che earth retura path which i o common to both sheath and conductors, in zero-sequence terms is th (ay-te) a L hogy 22 at ee senretie J tape 20 1 ‘ohms per phase per mile 3) (22m | o +n zq2=r-ty(8ie+2.) ohms per phase per mile, (24) Im ‘The equivalent circuit in Fig. 15(d) is a conversion from con just above it and combines the mutual impedance — common meres element. From this ete hen —h, EARTH both ground and sheath return paths exist, total zero sequence impedance is: opirieo couivavent cincUrT (IMPEDANCES EXPRESSED IN ZERO" SEQUENCE TERMS) @ Fig. 15—Actual and equivalent sero-sequence circuits for ‘hree-conductor and single-conductor lead-sheathed cables. =2—% obms per phase per mile, (25)6 Electrical Characteristics of Cables Tf current returns in the sheath only, with none in the ground tam (atm) + (2~En) Sate le (26) = ret 3r4+j0.8382 phase por mile 7) = rot Bret j(e,—2re— 3x4) ohms per phase per a (28) If eurrent returns in ground only with none in the sheath, fas would be the case with non-sheathed cables or with Insulating sleeves at closely spaced intervals, the zero- sequence impedance becomes: 290 (ta) Hn ‘ohms per phase per mile 29) ‘The zero-sequence impedance of shielded eables ean be caleulated as though the shielding tapes were not presen hhecanse the impelance is alfeeted only slightly by eine culating currents in the shields ‘The equivalent geometric mean radius (GMRx) for threesonttucior cables having sector conductors i dificult to calculate accurately. ‘The method used to ealelate values of GMRie for the tables of characteristics is of practical acouracy, but is not considered Lo be appropriate for explanation here. As an alternate basis for estimations it appears that the GMRx, for three seotor-conductors ia roushly 90 percent of tho GMRs- for three rotnd conduc~ tors having the same copper ares and the same insulation thickness. Example 2—Vind the zoro-sequence impedance of a threeconductor belted exble, No. 2 A.W.G. conductor (7 strands) with conductor diamoter of 0.202 inches. Con diyctor insulation thickness is 156 mils, belt insulation i 78 mile, lead sheath thickness is 109' mils, and overall cable diameter is 1.732 inches. Assume D,=2800 feet and resistance of one eonducior=0.987 ohms per mile at 60 cycles. Distance betwean conductor ccnters is ‘S~0,202+2%0.156~0.004 inches GMR of one conductor is (ee Chap. 3, Fig. 1 GMR,.=0.726x0.146~0.106 GM of thrvo conductors is OMRsy= Y(.106)(0.5047*=0.338 inches ‘The conductor component of impedance is (re= 0.987, 1,=0.280) 280012 987 +0.286-+0 8382 logy SOI 27-4j4.18=4.87 ohms per mile ‘This would represent total zero-sequenee eireuit impedance if all current returned in the ground, and none in the sheath, For the sheath component of impedance: e020 *=(1.623)(0.109) = 3X 1.1340,286-470.8382 logy = 8.68+38.87 ohms por mile =1.18 obms per mile 22800012, Tees Chapter 4 ‘The mutual component of impedance is: 57 0.280-433.87 If all eurrent roturnod the shoath, and none in the ground, eo LOTH HL 184+3.68-+93.87—0.57-71.74 =438+j0.31 = 439 ohms per mile. If return current may divide between the ground and sheath paths, ~ — (0.286-+ 93.87)" Beh ST ss 12744184 1623—2.31 = 2.89-471.87 =3.44 ohns per mile, ‘The positive-sequence impedance of this eable is: = 0.987-+0.208 ohms per mile ‘Therefore the ratio of 2ero- to pasitive-soquence resistance i 2.9, and the ratio of zero- to positive-sequence reactance is 92. Zorosoquence impedance is often caleulated for all return current in the sheath and none in the ground, be- cause the magnitude of the answer is usually elose to that caleulated considering a paralleled return. ‘The actual native of a ground-teturn cirenit is usually indefi ittmay be mixed up with water pipes and ather co materials, and also because low-resistance connections between sheath and earth are sometimes difficult to establish, ‘Single-Conductor Cables—Fig. 15 also shows the factual and equivalent circuits for three single-conduetor cables in 8 perfectly transposed three-phase circuit, where the sheaths are solidly bonded and grounded. The im- ppedance expressions applying to single-conductor cables Aliffer in some respects from those for three-phase cables: eo f semret ret 0.8382 Foy per phase co) seeretretiteatta— Bra) ohms per phase per tile. @n whore: ¢, resistance of one eonduetor, chms per rail. roma, resistance of earth (200 Table 8), ohms per mile stance to equivalent earth return path (see ‘Table 3), mches GMR,,=geomeiric mean radius of the conducting path made up of the three actual condue- tors taken as a group, inches, EMR SEMD sactance of an individual phase conductor ‘st twelve-inch spacing, ohms per mile. sactanee of earth return. % “ F ogy, Be 0.8882 5 loam 73 ohms per mile. eo Table 3.) 0.2704 F togy (OMD™ 2.702794 J tone Mp); ohms pr aieChapter 4 GMD,.=geometrie moan distance among, conductor centers, VSS Se of me sonnet rt 0.8982 J Toe Girt per phase per milo (32) or gertretjlate—2ra) ohms per phase por nl Ee) where: GMR..~geometrio mean radius of the condueting both made up of the thre sboaths i parallel CPV) eerosstance of one sheath, ohms per mile 0.200 = Gp e ap orl sth ris inside radius of sheath inches renoutside radius of sheath, inches, 21 renclance of one sheath, chs per mile 94 L a 0.2194 66 NOE OF, f dD. =r 0.8882 Flog axe — oh eam reti0 8382 £ lone gages ohms per phase per mile. ow Fm = Set j(tetz~ 2x4) ohms per phase per ral 9) where: MD w-2+— geometric mean of all separations between sheaths and conductors VEE owe P ome from the eile eat of Fig. 19, eal a sequence impedance when both ground and sheath paths exist ist 2.2 2,— 2 ohms per phase per mila, (25) If current returns in the sheath only, with none in the ground: fo Fehés—Bew ohms per phase per mile (28) = retro 0.8382 logy ‘obms por phase CMR Ser ile 88) eretreti(ts—%) ols por phase per mile, (87) If current return in the ground oa f= eta) Hin =. ohms per phase per mile 29 Cables in Steel Pipes or Conduits When cables are {installed In iron conduits or steel pipes, the zero-sequence resistance and reactance are affected by the magnetic material beeause it closely surrounds the phase conductors and forms ¢ likely return path for zero-sequence current. No method of ealeulating this zero-sequence impedance is available, but some rather complete results are available from field tests on installed low-voltage cables, 14 shown Electrical Characteristies of Cables 7 in Table 2. Some special tests of the zero-sequence im= pedance of high-voltage pipe type eable have been made but the results ate not yet of a sufficiently De generally usable. 3. Shunt Capacitive Reactance Shant capacitive reactances of several types of cables are given in the Tables of Electrical Charneteristies, directly in okras per mile, Tn addition, shunt capacitance, and charging current ean be derived from the eurves of ‘geometric factors shown in Figs. 8 and 9, for any cable Whose dimensions are known, ‘The geometsic factors given im these eurves are identified by symmetrieal-component terminology. "The positive-, negative, and zore-sequence shut eapac~ itances for single-conduetor metallie-sheathed cables are all equal, and ean be derived from the curves of Fig. 8. ‘Theee-conductor shielded eables having round conductors, are similar to single-conductor cable in that each phase ‘conductor is eurrounded by a grounded motallie eovering;, therefore the positive-, negatives, and zero-sequence values fare equal and are dependent upon the geometrie factor relating a conductor to its own shielding layer. The geometric factor for throc conductor chielded eablos having seetor-shaped eonduetors is approximately equal to the geometric faetor, G, applying to round conductors. How- ever, if the sector shape of a shielded cuble is known, then the curve in Fig. 10, based op insulation thickness and mean periphery of insulation, is recommended as giving more acentate values of geometric fuetor For single-conductor and three-conductor shielded eax bles (see Fig. 8), CoCo 008 pts pr pase per mi 8 re keke Lente Jem amperes per phase per mile 40) ‘Three-conductor belted cables having mo conductor shielding have zero-sequence values which differ irom the positives and nogative-sequence; the appropriate geo- etre factors aro given in Fig. 9; 0.267% irra pr pase ye il, (A) = 8088 mrt pr pe pr ile 1) are rg= 77, megohms per phase per mile, (48) ara ’ seo 2G eps pe pase pe ils) out. k-be bes I= fe=22EH amperes pr pase pr mie. 8) 0.329-k-ke 19SEC amperes per pase per mle, (45)8 Electrieal Characteristics of Cubles When three-conduetor belted onbles have sector shaped conductors, the geometric factor rust be eorrected from the value which applies to round conductors, “This cor- restion factor is plotted in Fig. 9, and its use is explained below the curve Tn the foregoing equations, Ch, Cs, and Co are positive, negative, and zero feaquence eapacitances rey ze and 24 ate positives, Sequence capacitive reactances. Ie, fe aud Te are positive, negative, and zero Sequence charging currents o=line-torine system voltage, kilovots. Kedicloctric constant, according to the values in Table 4. It is important to note that in converting shunt eapaci- tive reaetance from an “ohms per phase per mile” basis to cotal “ohms par phase” basis, itis nevesary to divide by the eieuit length: wgativer, and zero ith in mies’ bMS Pot phase, an 6, Insulation Resistance. The calculation of cable insulation resistance is difficult, because the properties of the insulation are generally. predictable only within a wide range. ‘The equations pre- sented helow are therefore quite dependent upon an accurate knowledge of insulation power factor. For single-conduetor and three-conductor shielded cables, 1.790 remreeri™ pe gO" ohms per phase per mile, as) For three-conduetor belted eables, 0.5976 tymry= 2 STE -10° ohms per phase per mile. Fk coe & G9) te 's _.10¢ ohms per phase por mile. (60) Fikes & In these equations, ry, Fey and ry are positive, negative-, and 26ro- Sequience shunt resistances. lelectric constant (see Table 4). cos $= power factor of insulation, in per unit. In Table 5 are listed maximum values of insulation power factor, taken from specifications of the Association of Edison Tuminating Companies. These standard values will very probably be several times larger than eetual measured power factors on new cables, ‘Tamu 4—Dimnecraic Consranrs OF CaBiit INSULATION Tnsulaion Rangeof t | Typical & Solid Paper soao | ar OiFFilled 30-40 35 Ga Filled 3d-40 37 Varnished Cambie 40-60 50. Rubber 40-00 i 60 Chapter 4 “Tamu 5—Manawua Power FactoKs" oF CxRL INSULATION “Temperature oi Gas Cable | Solid Filled Filled (Dex. €- Paper | dom-pretsure) | ow-pressure) toa | 0.000 0000 009 70 0.015 0.0075 0.013 30 o.021 9.0000, 0.018 8 0.025 9.0097 6.022 0 6.030 0.0105 0.027 ‘the poec tetra cae unl bbw thaw vy wi ar IL. TABLES OF ELECTRICAL CHARACTERISTICS ‘The 60-eycle electrical charactoristies of the most usual sizes and voltage elasses of paper insulated cable are con- ‘ainod in Tables 6 through f1.. In each ease the positive-, ives, and zero-sequence resistances and reactances are tabulated, or else constants are given from which these quantities can be calculated, Also, included in these tables are other characteristics nseful in eable work, such 1 typical weights per 1000 feet, sheath thicknesses and resistances, conductor diameters and GMs, and the type ‘of conductors normally used in any particular eable, Incach of these tables the electrical characteristics have boon ealeulated by the equations and curves presented in the foregoing pages. Where sector-shaped conductors are used, some approximations are nevessary as pointed out previously. In Table 6 the positive- and negative-sequence reactance for seetored cables has arbitrarily been taken 175 poreent less than that of an equivalent round-eonduetor cable, in aceordance with Dr. Simmons" recommensTations The equivalent GMR of three conductors in sectored, ‘eables is necessarily an approximation because the GMR of fone sector cannot be determined accurately. ‘This condi tion arises sinee the shape of sectors varies and a rigorous calculation is not justified. ‘The variation in sector shapes robably is greater than suy error present in the approxi- mation given in the tables, The reactances calculated from these approximate GMR’s are sulliciontly accurate for all practical calculations ‘Table 7 for shielded cables is similar in form to Table 6 and where saetored eables are listed the sare approxima tions in GMR and reactance apply. Table 8 for three- ncluctor oil-filled cables is similar to both Tables 6 and 7 and the same considerations spply. In these tables for three-conductor cables, the zero- sequence characteristics are calculated for the case of all retum current in the sheath and none in the ground. As pointed out in the discussion of zero-sequence impedance, this is usually sufficiently accurate because of the indefinite nature of the ground return cireuit. Where ground must Le considered or where there aro paralleled tree-phase circuits, the impedance must be caleulated as illustrated in the examples given. From the quantities given in these tables of three-con- ductor eabiles, the overall diameter of any particular eable ‘can be calculated. D=215Kd+2T)+2t+L) 1)Electrical Characteristics of Cables Chapter 4 ‘Tarun 6—60-Cyoim CHARACTERISTICS OF THREE-CONDUCTOR BELTED Pa?en-INSULATED CABLES Grounded Neutral Service ange suas gas 888 | age creenacy 29 sueaTHt Ra) sage gag e222 8 ee sthf 288908 [sp eeze azga mia ‘apRO—SEQUENCE B28 2858 S88 &: E oe ee RASS S888 O88 ea cash outs =r5 | bana Seen base ess 288 RRR gURR Raa Ba RERE 848: Soon test SEES ‘Babg SE DES EES ARS BRGS 2595 HEE eeseauee seen SGRR RESR S88 S89 ‘Sa5s SEES 9783 & Sane 882 83 58 3888 2BE a8) $8 RES BEES SEE Eg guar gEa2 #R » gee! SRSQRES E288 2255 SAR8 RAS see SUES 5 oct Se8e a8EE He AARY ROES GURL geR | HARE EONS GEae SBE ‘S588 S255 £858 288 | ga88 Bees 2832 BR | | sab S82 BB83 Bz [sacs e2es sees ese) BAR GARG SSE SBR S288 9888 £898 883 ESS SER8 R893 8 awke B225 2483 ©: ap SES 28 S228 S888 S88 WE OGEE SED 228 a 9858 $82 on Tne ei a eg Salon bel vip ail etirn curceat ithe sheaih nowe a eround ‘i lone mts are ul to desea scabies: SAS Rov CS—Comint Set Sat arent (SMM | SESE ASST SELL OOS | HERE GELS COLE GOS | SEET TCHG OBES BAS | THRE ALCS CSET OBE 2888 886 =e or T mae T tae ome T an80 Electrical Characteristics of Cables Chapter 4 ‘Tauie 7—60-Cycue Cuanactenistics of THREE-CONDUCTOR SHIELDED Parzn-INsuLaTep CaBLes Grounded Neutral Series i 8 i £ 6 k é 3 e 2 ¢ g 2 i é SB) HB fem] se gan | as | gat ie | & SB HB | 8a | 338 Sa | | 88 ue aes . SS | Fe | OF | S38 | ase | Sas | seo | oo. 8 | a 8 | me | ose | odes | aces | octet oe ey oat S| ga | see | gue | eae | ge at um |e at . BS | S| BB ee PB) RB a) Be te |e) te 8 oo fa | Ou8 | ORG | Se | ae | Ss | aes | G28 He S| 8 HB GS NERS GR] Me |e te @ oh] | SRR] Shs) ER | Sue | Sas | BB | SR | ae | oe | BR | ae | See & ge | 3 RSS are Seo | a | BGS | a | ESS | a] on | BE | a Be | ce A tae] O08 | Be | oat | Ae | S| a | te | mos | S| yee | eee | eas | oar fom | ae) ome | at | gam | me | ag | oe see | S| SB | See | eat | ae | ots |S | oem | Pe | Oke | as |S | oe L air ly na RT rans ad tea eT he FA Ehcingermbnie ar nd to dna condctr tye SRS in which, according to Fig. 6, D=outside diameter in inches. d=diameter of individual conductor in inches, TT =conductor insulation thickness in inches. {belt insulation thickness in inches (when present). ‘L=lead sheath thickness in inches. ‘This equation refers to eables with round conduetors. For sectored cables there is no exact rule, bub a close approx ‘mation can be obtained by using an equivalent eable wi round conductors und ealeutating the diameter D by Bq, (11, and then subtracting 0.3 to 0.4 times the round con ductor diameter d,depending upon the shape of the sector A set of calculated constants is given in Table 10 for single-conductor cables, from which the positive, negative- and zero-sequence characteristics ean be quickly deter mined by using the equations given at the foot of the tabulation. These equations are derived directly from Rad O5-~Compant Sain those given for the calculation of sequence impedances in the sections under Electrical Characteristics. Since ag £ 12 aa02704 Joon gt chins per phase per mile a 704 Log 2 02701 Liogy 2 ohms per phase per mile (22) ~oores Lom, & 247 0.2704 F tone 5 ohms per phase per mile (18) and 7, and r, are conductor and sheath resistances respeo- tively, the derivation of the equations given with Table 10 becomes evident, ‘Tuble 12 gives the one other quantity, ta, necessary fur the use of Table 10, ‘These reactancs spacing factors are tabulated for equivalent cable spacingsChapter 4 Blectrical Characteristics of Cables 81 ‘Tapia 8—60-CyoLe CHARACTERISTICS OF THREE-CONDUCTOR OIL-FILLED PAPER-INSULATED CABLES Grounded Neutral Serviae 13) typeof Conductor Ex ae B 1s : nome | ot | ee | ta] Big ist 8) 8) eS | | #32 | ee | | sa | _|__ | a 8 | |e aa mila oz | Ble 233 | | 8 i | . si | a ai | 3| 2 | em) 83 | BB) | $a | poop | co | nm | oer ae | i 5 sae HES ES g os Rau a! a ine 33 2 HE 3 i ‘Tl elloving pmb ace we diane te cae type: C—Compact Rava: CS-Conpact eo from 0.5 to 36.0 inches, which should cover the range met jn practice, For all spacings less than 12 inches, za is negative. ‘The constants calculated in this manner apply to one three-phase circuit of single-onductor Jead-sheath cables, assuming. all zero-sequence return current to be in the sheaths, none in the ground. ‘The GO-cycle characteristics of single-conduetor il filled cables are given in Table 11, ‘This teble is similar form to Table 10 and the impetiance characteristics are determined in precisely the same way. Here agaia dhe sequence constants apply to ome three-phase circuit of three cables with zero-sequenee return current assumed to be all in the cable sheaths. Single-conduetor oiled cables have hollow conductors (the oil channel forms the core), consequently Table 11 includes eables of the two ‘most common inside diameters, 0.5 and 0.69 inches. In each of the tabulations, the voltage clase listed in the first column reiers specifically to grounded-neutral opera- tion. Frequently cable systems are operated with other ‘than a solidly grounded neutral, In low-voltage cables the ‘same insulation thickness is used for both grounded and ungrounded operation, but in eables rated 7000 volts and above, a greater thickness of insulation is recommended for a given voltage class when eable is oporated with an ‘ungrounded neutral. A good approximation of the cloe- trical charactoristios of those higher voltage cables when, operated with other than a solidly grounded neutral, ean bo harl by referring in each specific ease to the next higher voltage elas listed in the tables, “The constants of several typical eables ealculated by the wethols outlined are listed in Table 18. These typieal ‘eases are included to he used as a cheek on the general ‘magnitude of eable constants when making calculations for a spevitic case. Representative sizes and types of eable have been chosen to caver as many types of calculation a possible, IML, TABLES OF CURRENT CARRYING CAPACITY One of the most common problems in eable ealeulations is that of determining the maximum permissible amperes82 Electrical Characteristics of Cables Chapter 4 ‘TABLE 9~60-Cycun CHARACTERISTICS OF THREE-CONDUCTOR GAS-FILLED Papest-INSULATED CABLES (Sue DED TYPE} Grounded Neutral Service TI ] 7 Lae al enoenavence | eaearm | le a if ~ — - 1] ’ | i a], >| ye |e, 3 Tl ds a | is |g ajé ag a i, E Re ald ¢|) a) # | 2 el) | B jdt y #/ 4.) ze g ai | ze ela: | ae FEI ds S| 38 é of | ak Be | 32 82 | $8 | 3a é3 | Bw) t/a) ie fem pea] ae fas ae ae em ae ee ;8) 8] SB] im | ee gut] om | ae |eam [se | og | gg | ae | ra 2/8) & #| im | a POE] SIR BB] SPER EB] RR | Le |B geo 28 bas) BPE TR He EEG {= | ow om fe oan | i | ous | se | ne | om | ar oR gu ls ga [ig [em | ge] fig, a) 8) dB | § io im |t PB) Ee) UE | Be) Rae oi a| ie les gia Te. ea $i |e | is | oe oS) S| GB) ER He) 2 i2 2 | 8 | 18 | BS veg |e | ae | tm 238 a3 aap] om | fa Siete ane ame ‘Bisel lta Smet in eat; sne eon, ‘cing eple ar ad odinaesundater type: Sead per conductor for any given cable. ‘The limiting factor in able applications is not always the maximum permissible insulation temperature, Somotimes regulation, efficiency, ‘economy, ete., may dietate the maximum permissible am peres, However becauso tomporature rise is most often the controlling factor, the calculations of current-carrying ‘capacity are usually based upon this limitation, In Tables 14 through 19 earth temperature is assumed to bbe uniform at 20 dearees Centigrade. These tables were taken from a publication of the Insulated Power Cable Engineers Association and give maximum allowable am- peres per conductor for representative eable types, Cor rections for earth temperatures other than 20 degrees Centigrade are given within the tables. Special conditions may make it-advisuble to ealeulate a cable temperature problem in detail" taking into account variable loading, “hot spots” along the cable route, and other factors not contemplated in making up. the tabulated information. Approximations can also be obtained for the eurrent- ‘carrying capacities of other types of insulation by applying Porat ture Castano mat Yen splat ‘multipliers to the tables presented for paperinsulated ea- bles. ‘The value for varnished carmbrie-insilated cables ean be obtained by multiplying the value given in the tables for paper insulation by 0.01, the resulting figure being accurate to within five percent of the calculated value. Similarly, carrying capacities for rubber insulation ean be dotormined with the same degree of accuracy by applying ‘a factor of 0.85 to the figure given for an equivalent paper- sulatel cable. For special heat-resisting rubber thie factor becomes 0.95. Cireuits are frequently installed with esch duct eon- taining threo cables. The current capacity of these cir- ‘nits will be less than that tabulated here for one eable per duct, but will be somewhat higher than the eapacity of an equivalent shielded three-conductor exble of the same gonductor size and voltage rating ‘The number of overhead power cables is small per~ centage of the number in ducts, and for this reason space does not permit, inclusion of loading tables for eables im air Unfortunately there is no simple correction factor or curve that can be used to translate the figure for cables in ducts,pimphoy| B22 288s see eqs | Se S88% FEE 282 A828 2688 SASS 583 DEVE | eee asses sage gag| se eRoe aEEE Bens Seat aaEg gees E885 585 ages 2823 #4 rounded Neutral Service 'PAPER-INSULATED CABLES Electrical Characteristics of Cables ‘Tapue 10—60-Cyoue CHARACTERISTICS oF SINGLE-CONDUGTOR ConcanTaic-STRAND Chapter 4 age 2805 22a | Sasol a Beets 8 stomane for siraning, (Based on ail return current In sheath; none In ground) UES ESSE G82E 33 | i 7 Se 2 uae 233 | is eRe Bag £34 ERR = aa EE EE ait | -s8§ g868 Se2 | ag Egee ue aie FET ahaa fee | ° 9988 498 y é 3ai S228 ggg | G0 Seve eve jaeen wa | i § bee com | we | BE fens asus SSE wuss §Seg # fo38 $588 F2s2 Saag See SaaS sshd G2E8 2835 HESE faze 2: | Bae esea Oe82 B25 E82 383 i i i i. : #3 goag SESE oae8 gigi ¢ EY ae2 G8 E83 SESS EEE iis 35 gg 4503 #892 gees en | 225 8 = SS are Ea HEE HEY GORE TaE/HREE SE GEER HE | spas worshoneg #82 G80 SER Ba ae gore zy | Peet 8 eS nn q HiT! BS | as a mang BED 1 Et ers, 88g 222 ("7° -80Q S888 £8 oeay aedg gue[" -eag gue gas |E% = i wae ARSE ESE aa82 888 585° ESE aaan|72 2 § Tove wos S08 SG Gee cea lead Bnew Pree ene GEES Wage sua HS GEE! Bee ‘ero oreiva wr T ae I “ae aweChapter 4 ‘PAPER-INSULATED CABLES: Grounded Neateal Serviee Blectrienl Characteristies of Cables ‘apie 11—60-Cyeue CHARacTExistics of SINGLE-CONDUCTOR OIL-FILLED (Hottow Cone) a z R258 25 |B8SE 32 [8253 ies teu ld pram S82 S292 3 EGET FERS |S CABLES IN PARALLEL Also, mutual effects may develop between 2g G22 22 /gi2 2022 [2888 2882 2283 Iv. wartetle— ms) 2 S825 $2 958 S555 ESE) 3553 (Based on all return current in sheath; mone tn ground) i 8 z 4 5 a 4 5 f s 4 q E ‘Zero-Sequence Impedance: than one eable per phese is instslled in order to carry the oral current. cable eireuits which are adjacent, throughout their length but which terminate on separate busses. Depending upon, ‘This reduetion in carrying the type of eiteuit, the cable type and configuration, and. the system conditions being investigated, the problem muy. is frequently encountered, because in many cireui take any of several forms. {gegtgg2f glans tor sanding: Abovn value alate roan “A Se of Curves Sor Sti etn Tata 4 t i oobi] sees Sane #5 [ened sats [eaes 2828 | fees 282 | 3 4 bys Peuz7 # [se 283 [ase] pb. eee aang e559 $858 38 2883 ai| rb faa foes? SSS GEES ae leaae GEE aaae| Eid "a 822 $2252 S03 2803 38 Hi) ah 2 gf! 22852. Tse 2 e,aigi d Gh; ast83% Higeey & G22 Sa5Z5¢ Ri Giccege 22 252502 |: Geyesdti 222 2z2nts Si), MEPy Es: Epis] aed a Ee BEcees ili 2itsige G2! e25 i423 sz| si BETS E23 gFegsieies SEAL) Go ppd e's si PLa7 ees vomvo| "282 S558 38 [£22 2958 [£883 2882 | 2282 222) 422i: eB € B4b ahi sity sygonss| "HE HEH Ea saga ("sas fan | asa2 fz | 2 2° © 3 PEGiiLE =| ae ba ec ee eis caw | 2F s2ba 28228Chapter 4 Electrical Characteristics of Cables 85 ‘TapLe 12—-ReAcTANce SPACING FACTORS (x,)*, OHMS PER Mau aT 60 CYCLES Tr. In. 188 25 210. 160 105 ars | oar | 290 | 0107 170 | 0012 | 220 | agra | 30.0 | ot 175 | 0046 | 22.5 | 007s | a10 | ois 180 | 9.049 | 2.0 | oora | 320 | 0.110 1835 | 008 | 28.5 | ons2 | 330 | 0.123 190 | 0056 | 250 | ose | 340 | 0.126 19.5 | 0960 | 25.0 ‘oaa6 | 325 | —0.150 | 5.75 1.00 | —o.2 | 3.50 | -0.149 | 6.00 oz Hoss | 0.25 30 | 0.252 ~ou33 | 6.50 176 | —o2s4 | 425 | 0126 | 675 200 | 0217 | 450 | -o119 | 700 225 | —o208 | 475 | —o.1n2 | 7.25 250 | 0100 | 500 | 008 | 750 aan few hn Sie aig Ibis difficult to anticipate in detail the problems met in practice, but the examples outlined here indicate methods ENGTH IN MILES a deat ralaiomane i Frenne ene ie eee ee eae 5 caused by self and mutual impedances but such equations cae et an i. SO ee nes Co cea eee sual ts teste slon “ trode ear sen erm ern, er rama wben Tard sate ce comme» ma seeing oe gauges fu geval emi Seen nlite segues componente Meese ee ae ent ee ee eae su areas Sree sample of cleltion on pacled _ oe a en ee ee MEM Coes sat inp i, Baur st for paral eb, ih op Example 3). ‘TADLE 18. 60-Cycum CoNsTANTS oP TyrIcAL CADLES Iv O1me Per PuAst Pon MILE PORTING, AND. ier Cas tated 4 caging sn Shee > CR 9764 i psa nome ata pe Nest ie oe | eam | ootso | asz0 | oom | ost | 270 rete ae mare RMON tm coer) a2 see | oats | oe 200 | 2 | om | sem ‘Tipo Toye: Kore, 00 CM. 19/64 ie Taunton 18 | 0.488 | 0.98 | L166 sao 2a | osm | 3 a0 ise Goigg OPT yw Ar. 00 MCN. 0 ibe] a4 | 0.1m | Ox | 6360 sem | ss | one | soe SCendasoe enpertae ESC; Seth tnpanane 786 Electrical Characteristics of Cables Chapter 4 ‘Taste 14—Curxenr Carrvine Capacity or THREE-CONDUCTOR BELTED Pare-INSULATED CABLES in of Bally odode Det Bak ONE [rinse 1 six Tie [twee a ol * [ml »|»]s t/a] al eal 3 Pa TRV B elsigials | ig BS |B) le BR GB) By ey es B\s|/ee 8 BS 2 eRe eg elele ele. eee alalala a 8) 8) 8) gl gi 3) | BiB) 1 Bl ee MESES BE aS GE SE ARG, OE AR SE } S85 SS BP BP PYRG § Blagg 3 Bae 8 gle 2/3/88 | _ on ; _ Hye 5 a 3|s 8) gl gli) 8) B jg) ae 2/8 HB) | 8) 8 als 3) 8) 5] 8 We EE cee Sees etl ei aneE C5 companied 1 Mutt tainted tresta hy hac fated win yr tngpanane abe than 20°. Esample 3—Type of Circuit: A three-phase 60-cyole load current (positivesequence only), and its magnitude ‘cable rircuit connected between 1 sending and a receiving iz known, ‘The cable conductors ean be of different sizes, bas, using single im | tm) to m | om | om | of ° 4 | eal air gu After substituting the proper self and mutual impedance alues as defined later, these equations can be solved by the ‘method of determinants for current distribution, based on 4 total of 1.0 ampere positive-sequence current in the eit cuit, To obtain actual currents, the disteibution factors rust be multiplied by the actual load earzent in amperes Apparent conductor impedances: Using the current- distribution factors for each conductor to solve the eam plete voltage drop equations, an ‘apparent! impedance for each phase of the clreuit can be ealeulated. ‘This ap apeslanee ia valid only for the particular current division caleulated Apparent impedance of phase a a Tain Be Bes ohms Apparent impedance of phase b Ey “ Tipp ie gh oben aie, ohms, Apparent impedance of phase © Fe Be ap eat wpe ae Bem oie, ohms, Supplementary equations: The original ascumption of positive-sequenee current flow through the citeuit pre- cludes the existence of any net ground return current, This assumption simplifies the determination of the various self and sutual impedances, beesuse the effects of a Chapter 4 ‘ground retuen path may be ignored with very small error: By alert in) where 1 uit length in miles ‘ac. resistance of eonduetor a’, ohms per mile. reactance of conduetor a’, to twelve inch radius, ‘ohms per mile, = 70.2794 log Zon, Bos, Bot, Ze, and Zr are determined similarly, based ‘on the respective conductor characteristics. 2 Zeb j0.2704 logy 37 1(—22) where spacing in inches between conductors a’ and a”. The re- ‘maining mutual impedance are calculated similarly, using the appropriate spacing for each. A series of more complex examples of the above type of problem is deseribed by Wagner and Muller.* Example 4—Typo of cirouit: A three-phaso 60-cyele cable eizeuit connected between a sending and a receiving bus, using two dissimilar three-conduetor lead sheathed: cables in parallel Conditions: Exch cable contains three conductors that, by the nature of the cable construction, are symmetrically. transposed so that the flow of positive: or negative- sequence currents will cause no zero-seqquence voltage drops, Therefore, the sequence networks are not interde- pendent and an impedance value of each sequence may be ealeulated and used independently Problem: To find the zero-sequence impedance of the cntire cable circuit, and to determine how zero-sequence current divides between cables. abe ‘scary LENGTH —— ~ enn % & ze [tee laey |e tlt we ' ss : i th 2 , ‘rn o [EQUIVALENT GIRGUIT, TW ALL QUANTITIES {EXPRESSED IN ZERO SEQUENCE TERMS Fig. 17—Actual and equivalent zero-sequence clreutt for two. parallel three-conductor lead sheathed cables (see Example 4).Chapter 4 ‘Tapte 17—Cunnent Carrying Capacrty or THREE-Ct Electrical Characteristics of Cables 89 ;ONDUCTOR GAS-FILLED Parer-INSULATED CABLES T ‘l=[=lel=l=[=letate i [8 | a] 8] 8 B 8 alg e a m |S M8) S| 2) mis LS |e a) 8 | Bt # aa | a | at] ae] ae s | | Rel aB 3 2/218 8/8/82 ia] ie BEL ue | S| Bh By) BLE ag {| s/w) w]e | a By S | el Bl 2 #1 ad at PE a HEE 88 aS | V0 AE Sat aE 3 at a PREREREE ESE correc ae + say tant Circuit: Refer to Fig. 17. ‘The three actual conduetora in each cable have been reduced to one equivalent con- ductor in this figure, and all impedances are to be ex pressed in nero-sequence lerins considering the earth a3 a return path for each cireuit. Complete voltage drop equations: Ryn lye lege Lele tle lice the sheath currents ean be solved in terme of conductor currents by using the last two equations only. Also, itis ‘convenient to assume that the total zero-sequence current flowing into the sending bus is one ampere, which makes Tem LO~ Ly After making these changes, the following single equation, results: Le A rt 0 nae aoe, Modified equation: tll ~ Bee) } (Ber Bee aloe Bey) Ta Targ~TZe Ze Bes) Lerrkese ~howte) ‘This equation furnishes a solution for Je, from which Fer follows directly. ‘To find the zero-soquence imperlance of the entire cirouit roquirea that ono of the complete voltage drop equations be solved for By ‘Then Becker) fot Tesh by 10 = Ba, ohms. Supplementary equations: ‘The equations necssary to determine each impedance value are shown here: every impedance must be expressed in zero-sequence terms, with, the effect of earth as a return path included,90 Electrical Characteristics of Cables Chapter 4 ‘Tame 18—CUuRRENT CARRYING CaPactty OF SINGLE-CONDUCTOR SOLID Parex-INsuLATED CABLES eae ee a ee wba Lele [| «|= lw» |» |» | AMPEKES FER CONDUCTOR ‘Copper Tempers °C 2 S282 EEE GEE ¢ a ig: a2 S38 BE859 SERRE 388 & & 3 Ee SEER dezze enue fae... 2) #\# g B/E & ‘es lal a = mle as 3 g & 8) 8 # a g 8/8) 8 BIa|& 8 3 | i | 2] R) o i Ed #/ | 8 B|a)8 2) 8 g o| 3) 2 e 2 6| 8) 8 @ Fe = 8) 2) e 5 3 8 gle| gs g & ig B/G |g | el Tle uy oS | Made ae MiedoChapter 4 Electrical Characteristics of Cables a1 ‘TABLE 18—CURRENT CARRYING CAPACITY OF SINGLE-CONDUCTOR SOLID PAPer-INSULATED CABLES (Continued) AMPERES PER CONDUCTOR! 3 8) E)S)2/28 2 g o wl al am| a | a = 3 Be] ts | | SS |e a Fi 8 2,R eB) 8 B/ 2) 2 = we |e | ae | ae | ae | & | BR Pe a | a a So) Re] a] Rt | Rt ci | & i a | BE] eS ee | ge B a S| asad ontieiom | hg de eaagee Bilais 1341 vate a] Be) |e mR By 8/8) 8 @ els) se BBB 2 | 2| 8 Be) ee Bt) Be) 8 |) 8) s 8) | Elgg 8 E EB |e | HS | ae a 3 ao B ta le one pit ely (pan suited best FO i per dat see daly Leona in cutalde dus ony. sabi hadi ha op oe Wetted a un 0,Zam Urettetj (tet Z.—2ta)] chins, where I= circu length in miles, and the other terms are defined as for Eq. (19) Ze is defined siznilarly. Be=UBret1,+)(G24-+2)] ohms, where the terms are defined as for Eq. (28) Ze is determined similarly [reta(8y+2,)] ohms, where the terms are defined 28 for Eq. (26) is determined similarly, Ite je Bra)) ohms, where 5 ingfor Sth contnncentr spacing between cables, in inches ‘A more general version of the sbove type of problem, covering those cases where the cables are nol necessarily bussed together, ia deseribed by Cheek” Example 5—The use of complex GMR's and GMD's will very often reduce a complicated problem to workable terms. ‘The use and significance? of these factors should be studied Uhoroughly before attempting a solution by this method (soe Chap. 8). Taste i -Cummesrt Cansvina CaPactey ror SINGLE- CONDUGTOR OML-FILLED Pare-INSULATED CARLES (amperes por conductor)* Rated Line Voltege—Gronnded Neutral Giver Sta, | 84800 | 46000 | 69.00 | 115000 | 198000 er AWG. | “oua)’ | __ Maximum Copper Temperature Deg 7 ° P aco 0100 is zoo | 405 | ass | amo | Bus | sooo | aso | ao | a0 | aoa |e sooo | iia | ao | im | |e wooo | ps |g) om | Go| ato sooo | siz) oa | ins | a0 | aoe woos | oss | ss | eas | fas | Sry wooo | 2 | mo | ts) oss | aso room | te | iw | Ss | nom | ier | fee |i rococo | sre | smo | Sas 1290000 | sno | me | ose rose | 07 | 108 | Fae psig SL ae aa i te, vertigo Electrical Characteristics of Cables Chapter 4 ‘Tamur 20—Svecusne WeTwsrANo Larue Younees For CABLES WITH METALLIC COVERING" Solid-Paper Towulation OiLFilled Paper Trsulaton | Basie Inmalation) _Innpoleo Cinss | Tosulation | Insulation Withstand Iowulation Withatand ky | Level for /thickness) Voltage {Thickness} Voltage | Haipment | a a io | 4 =| i | i 68 350 | 878 138 50 | err Bese oe resnncnlains by Halprin sad Sarin the three- 1, and am Cireuit: Four paralleled cables similar to conductor belied cable deseribed in Example ranged in a duct bank as ilueteated in Fig. 18, Problom: ‘To find the overall zero-sequenee impedance of the cirouit, with sheaths and ground in parallel, or with return current only in the sheaths. GMR of threo conductors, GMR.=0.838 inches (from example 1), (2) Cable configuration. oF owe conovcron nov oraiRcu ‘CONDUCTOR BRANCH ots oat SOS (GMO) SEPARATION 38 ita) CONDUCTORS 0.038 loti ay 2500 ro TF aC eiNOOF EATS, () General equivalent circuit, Fig. 18-Four three-conductor cables In a duct bank (see ‘Example 8).Chapter 4 GMR of the four conductor groups, GMR.= ¥/(0.338)"(5)X10)*(15)'~8.479 inches. Equivalent spacing of three conduetors to their sheath, Saat GMD among the conductors and the sheaths, EMD (oe-4) = 9/(0.812) (GO) 5)#= 4.380 From Fig, 18(b), resistance of the sheath branch, 0.600 (.623)(0.10)¢@) Also from Fig. 18(b), impedance of the ground branch -2800%12 4.330 0,812 inches, ches. 0,818 ohms por mile, = 0.286-+0.838 log Electrical Characteristics of Cables Fig. 19—EMtect of 93 is gS 2 oo 2g ea Eas Ee BE open | Sins | a.” espn) : za 0 steel plates on current-carrying capactty of se aystem; fat config =0.286-+33.260 ohms per mile. single-conductor = ] T s | | 2 bed 4 | 5 be 4 i Fig. 20—Summary of some Impulse tests on paper-insulated cables (based on Information presented by Foust and Scott!®), and Eddy.!? 1 x 10 negative wave, high density olid insulation (Simplex Wire and Cable Co, 2 Held and Leichsenring,¥7 negative Wave, solid insulation, 3 Hele ive and negative waves, oll. 4 Held and Leichsenring, positive wave, ollé insulation, 5 An unpublished test, regular density paper, ofl-flled in= ion (General Cable Corporation; ‘and Scott, average of ve ces, 1x 10 positive wave, ular denaty pape, alld inaulation (General Blctic 1 4 reat a i 7 An Unpublished test, high density paper, ol-fited insu- lation (General Cable Corporation). x 8 Am unpublished test, solld insulation (Che Okonite Company) Foust and Scott, 1.5 x 40 positive wave, reqular density paper, solld Insulation. Foust and Scott, combination regular and medium den~ ‘ity paper, solid insulation. Foust and Scott, high density paper, solid insulation. Boust and Scart, madium density paper, solid insulation, Foust and Scott, 1.5 x 40 positive wave, combination reg ular and medium density paper, solid insulation. Foust and Scott, 0.5 = 4D positive wave, regular density paper, solid insulation. Foust and Scott, 0.5 x § positive wave, regular density Paper, eolld Insulation, 1 2 3 44 Electrical Characteristics of Cables ‘The zero-sequence impedance with sheath and ground in parallel, 0.848(0-286-48.260) 9, ae oe ia py +0247 470.0707 = 1.02940.276 ohms per phase per mile ‘The absolute value of this impedance is 1.00 obims per hawt por mil - ‘The zero-equence impedance considering. all retum curvent in the sheath and none in the ground, Zo= (0.247 450.0797) 40.848 = 1.008-470.0797 ohms per phase per mile. x ‘The absolute vale of this impedance is 1.1 ohms per phase per mile, or substantially the same as with the sheath and ground in parallel. In this ease the effect of high sheath resistance is minimized by the fact that four sheaths are paralleled, V. IMPULSE STRENGTH OF CABLES Power-transmission circuits are often made up of cables and overhead-line sections connected in series, and this coustruction may impose lightning-surge voltages on the cablo insulation. Even when circuits are totally under ground, it is possible that eable insulation will be stressed by transient overvoliages eansed by avvitehing operations For these reasons the impulse strength of eablo insulation information of some value for predicting eable perform- ance in un actual installation No industry-wide standards have beon established for cable impulse strength. Test data from various sources is "and some of these results for paper-insulated thown in Fig. 20, Severs! variables are inherent jn the curves, #0 hat the spread of the test points is wider than might be obtained with uniformly controlled test conditions. ‘The factors not. yet, completely investigated include the effect of normal insulation aging, the relation between actual voltage gradient within the insulation and the average gradient, wave shape and polarity of the test impulse voltage, and grado or compounding of insulation. Using 1200 volts per mil average stress as a sale with stand impulse strength for paper-insulated eables, as sug- gested by Halporin and Shanklin," the withstand voltages for representative cables may be listed as in Table 20. REFERENCES 2. Calculation of the Fletriesl Problems of Underground Cables, by D.M, Simmons, The Eleerie Journal, Vol. 20, May Wo ‘November, 1982. (The fst tile this tele eontaine w cone prehensive biliography for 1889 nd before.) 2. Symmetrical Components by C. F. Wagner and KB. Evans (a book), MeGran-Til Book Company’ 1988 8, The Tranamiscion af Bletrie Power, Vols. Vand T, Tavis (bro), Minos tnatitute of Pechnology, 19 4. Careoni-Rating and Impstance of Cables in’ Bulking and Ships, by H.C. Booth, B. E. Hutchings, and 8, Whitehead, TER. Journal, Vol. 83, Oetobor 1988, p. 407 5, Problems in the Mensirement of A-C Resistance and Resetanes of Large Canductore, hy EH, Salter, ADEE. Preniastons, Vol. 67,1988, pp. 1300-1390 by WA. Chapter 4 6. A-C Resistence of Large Size Conductors in Stel Pipe or Con- dust, by Rod. Wiseman, AJE.E, Transactions, Vol. 87, 1948, pp, 1745-1758. 7, Reactanee of Large Cxbles in Steel Pipe oF Conduit by W. A. Del Mar, AEE Tronsactons, Vol 61, 1048, pp. 1400-1412 4 Unbalanned Carmenta in Cablo Groups, by C.F. Wagner snd M,N. Mull, dn, Tho Electric Journal, Vol. 35, October 1958, p. 300, 8. Zero-Sequence Impedances of Parallel Three-Conductar Cabl by R.C, Chock, Hlecini Light and Power, October 1048, p. 10. ‘The Temperature Rise of Cables in a Duet Bank, by J. H. Neher, AALE-B, Technical Paper 40-184, April 1949 11, Determingtion of Cable Temperature by Means of Reduced Seale Bfodel, by Androw Gomant and Josoph Stier, A.B. ‘Transaetions, Val. 65, 19, pp. ATE 12, Impulse Strength of Cable Ineulation by B, W. Davie and WN. Pedy, ALLE. Transactions, Vol. 89, July 1040, p. 304 18. Somo Tmpulie-Valtage Breakdown Testa on OllTreated Papees Insulated Cables, by CM. Foust and J. A. Seott, ATER, Transactions, Vol. 89, July 140, p. 389. 1H. Impedance of Three-Phase Secondary Mains in Nonmetallic and Jom Conduit, by L., Beiger, FET. Bulletin, Fobraszy 1988 iiowtions for tmpregnated Pepertnsulated Lead-Covered "Solid" Type (7th and Sc editions), "OM-Filled” Type ‘ition), “Tow-Pressime Gas-Filled” Type (1st edition), Prepared ly Assocation of Edison Ulanninsting Connpanie 16, Current Caceying Capacity of Frapeegaated Paper, Rubber, and Varnished Carve Insulated Cables (Int editin}, compiled by ‘The Insulated Power Cable Bugincers Assocation, Publication No, P.20-206 17, The Behavior of High Tension Cable Installations Under the Bflcct of Voltage Inpulses, by C. H. Held and H. W. Leich= senving. Paper No. 297, CLGRE, Pars, JunesJuly 189. 18, Impuhe Stringthof Inslated-Power-Cable Cireuit, by Heri Halperin and G. B. Shanklin, A.lE.B, Transactions, Vol. 63, 1945, p. 1190. Books 460. Plcrie Colo, by W. A. Del Mat, MeGiraw-till Boole Com= pany, 1924 4 Elerical Choractarieice of Tronamiesion Civont, by William Nesbit, Westinghouse Techical Night Sehunl Pret, Bast Ptr burgh, Pa, Sid edition, 1928, 42. Underground Systeme Reference Bonk, NELA. Publication No, 050, 131 49. Symmetrical Components, by ©. P. Waguer aud R. D. Eves ‘MeCiraw-Hill Book Company, 1988 44. Impresnatal Paper Tnsulaton, by J.B, Whitehesd, John Wiley A Sonn 85, 48. Bcc Power Traneninsion and Distribution, by Le P. Woe ru, John Wiley & Sons, 1988 48, ‘The Principies of Blaeie Power Transmssion, by H, Waildizon, ‘Chapman & Hall, 1990, AT. Tho Tranamsnion of Hlectse Power, Vole 1 and Uly by W. A. Tools, Minos Inatiznt of Teshologs Impudavee and Capacilance {W Formulas and Tables for the Culoulation of Mutual and Bell Indurtanee, Mesa Rost and Glover, Burestt of Standards Scientic Papers, No. 8169, 1919; ako No, S820, 1918, 61, Prosimity Eee in Cable Sheaths, Dwight, A... Tranoae fuons, September, 11,9. 9, 62, Cloulation ofthe Blctrica Problemsof Underground Cables, by D. M-Simmons, The Bkatve Journal, Vol. 29, May, Jane, July, Ostaber and November 1892, pp. 237, 288, 397, 476, land 827 63, Calculations of Industance ad Current Distribution in Low- Voltsge Connections co Electric Fumaces, by €..C. Levy, ALLEL, Transactions, Vol 51, Deseanber 1022, p. 98.Chapter 4 (0, Resistance and Resctance of Three-Conduetor Cab, by BH. Salter, GB. Shanidin, and R. J. Wisoman, A.IE.K. Teanoac. tions, ol. $8, Decewber 1984, p. 1881 165, Irapedsnce Measurements ot Underground Cables, by RL 1 Ip ALLELE, Promenctions Vol. 85, Apri 68. Impedance of Three-Phase Secondary Muine in Nonmetallic and Iron Conduits, ty Le Brioger, F-E-1, Bulletin, February 1958, (67, Unbalanced Curren in Cable Groupe, by C. F. Wagner and H. Muller, Jr, The Bledric Journal, Vol 38, October 1988 p. 390 68, Carrent-Rating and Impedance of Cables in Buildings ard Ships hy H.C. Booth, FE. Hutchings, and 8. Whitehead, LEB. Journal, Vol 83, October 1088, p.497 60, Problems in the Measurement of A-C Resistance and Reactance of Large Conductors, by EH. Salter, AILB.E. Transections, Vor. 67, 1985, p. 13%, 70, Renctance of Largo Cablee in Staal Pipe or Cond by WA. Del Mar, LEE. Tronsneians, Vol. 67, 1948 p. 1400. TL A-C Resistance of Largo Size Conductors in Steel Pipe or Con- ‘dai, by Red. Wiseman, A... Tvenmacnons, Vol, 1948, pins, 72, ZeroSequence Impedances of Parallel Three-Condactor Cable, by H.C. Cheak, Hlecric Light and Power, October 18, p. 74 75. AVC Resistance of Segmental Cables in Stel Pipe, by L. Mayers hoff and OG. 8 Bogor, ny ALLELE, Praneactions, Vol. 68, 1080, 1. Jtlone and the Calelation of Tnduetance from Geo metre Mean Distances, by W. B. Boast, A.L Hu, Yransocions, Vol. 68, Part Hy 1080, p- 1631. ova Rating ard Heating 100, ‘Temperatures in Pletrie Power Cables Under Vsviahle Load- ing, by EA. Church, ALE, Tranetctans, September, 198), p. 889, Calculation of the Electrical Problems of Underground Cables, by D. M. Simons. ‘The Electric Jowraal, Vol. 20, August and September 1982, pp. 805 and 428, ‘Thermal Transiente and Oil Demands in Cables, by IK. W. Mille and P.O. Wollason, A.L.B2. Tranaacione, March 1983, Vo. 82, p. 98 eonomical Looding of Underground Cables, by EA. Chureb, ALLE. Transaction, Vol. 54, Noveriv 1935, p. 1168. Current Rating and Trnpedance of Cables jn Bulkdings and Ships, by HL. C. Booth, B E, Hutchings, and 8. Whitehead, LEE, Journal, Vo. 8, October 1638, p 407. Masimam Safe Opeeating Temperacures for 15 kv Papersdn- slated Cables. by C.W. Franklin and E.R. Thomas, 4.1 BLE. Transactions, October 1989, Vol. 5, p. 856. bail Ratings of Cable, by Herman Hslperi, A.D.E.B, rama ‘actions, Vol 88, October 1999, p. 535. Boonomica) Loading of High-Voltage Cables tnatalled in Ua- dergroond Subway Systems, by E.R. Thomas, ALB Traneacions, 199, Vol. $8, p. 61. Load Ratings of Cablet, by Herman Halper, ALB, Transactions, Vol 6 981, YO, Guide for Wartime Conductor Temperatures for Power Cablo Jn Service (committe port, A.L.B E, Tranenations, Vol. 63, Beptenber 149, 600. Current Conying Capacity of Tmprernated Peper, Rubber, and Varnished Cambris Insulated Cables (Ist edition), cou piled by The Insulaed Power Cuble Engines Association, Publication No. P-29-226, 1088. Determination of Cable Tenperaturn fy Means of Rasued Beale Models, by Andraw Gemant al Jowoph Stier, A.L.B-F. Transactions, Vo. 5, 148, p. 475. ‘Thermal Charartocston of 4 120 ky High-Pressure, Gas-Filed| Cable Installation, by W. D. Saivlersn, Josoph Sticker, and MH, MeGrath, AB, Transactions, Vol. 67, Part 1, 1983, past, rol 102, 18 104 108, 106. 17. 108. 100, 0. uy, na. Electrical Characteristics of Cables 95 NIA, The Temperature Rise of Buried Cables and Pipes, by J. HE. Neher, A.DE.E. Tronanetons, Vol. 68, Patt T, 1949, p. 9, ‘The Temperstute Rise of Cablee in a Duct Bank, by J. Be Neher, As. Trunacetiona, Val 8, Part Iy 198, p. S41 ‘Transiont Pemperatire Phenomena of 9-Conduetor Cabos, by F.0, Wollaston, 4..B.B. Transactions, Vol BS, Part, 1949, 1284 ‘The Thermal Resistance Between Cables and x Surrounding Pipe or Duet Wall, by F.H. Buller and J. H. Neber, ABE. Transactions, Vol. 69, Purt I, 1950, p. 312. ‘Heat Transfer Stady on Power Cable Duets and Duct Arson bie, by P- Grochler snd G. F. Barnett, A.LE-B. Tronsactions, Vol 60, Past 1, 1060, p. 957, Intation 140, The Behavior of High Tension Cable Tostallations Under the Etfet of Voltage Innpulses, by C. H. Hell and H.W, Lelehana- ring. Paper No. 207, CALG.RLE, Paria, June-July, 1999, Tinpulse Strength of Cable Insulation by E. W, Davis and W. N. Eddy, A.TE.B. Trovanctins, Vo. 69, July 1940, p. 304 Some Impalie-VoltagoFreakdown Testo Oil-Preated Faper- Insulated Cables, by C.-M Foust and J. A. Soot, LEE, Tronsnctons, Vol. 89, July 1940, p. 383. Impulse Strength of Tnsulated-Povver-Cable Circuits, by Her man Halperin and G. B. Shanin, A.Z.E.B, Transactions, Vol 83, 1011, p- 1100 PPowor Faetor Messurerents on Poly-phase and Multiconduc- tor Cale Using SinglePhaso Bridges, by B. W. Greate, AUE-E. Pransoctions, Vol. 68, Psxt I, 1980, p. 65D. General 160, Characeristice of Oilflled Cable, Shanklin and Buller, AWE. Travsoctions, December, 1981, p. Mi 151, Oikiled Cable and” Accesion, Atingon apd Simmons, ALEE. Transcctions, Deceber, 191, p. 12 382, 120 ky Compressiou-Trpe Cable, by'I. T. Fauostt, De 1 Komivee, I. W. Collins, and Re W, Atkintea, AE. Tran ‘actions, Vo. 81, September 182, p. 652. 169, 120 ker High-Pressure GaeFilled Cable, by TT, Fauoett, Us 1. Konives, H.W. Colin, and RW. Atkinson, A.D. Transactions, Vol. 61, September 1042, p. 658. 154, Lows, Medion, and High-Pressure Gas-Pilled Cable, by GB Shanklin, ALLIEE, Transaetions, Vol. 61, October 1942, p. 719 ble for Posy Transmission and Distribution, by C. 7. Matcher, lecric Ligh ond Powe, September 1948, p. 38, aud October 1946, p72. 196, High-Preener, Gus flied Cable Impregnated with BxtracTHigh Viseosy Oi by Josep Sticher, 0.1 Doan, BW. Atkinson, nd Louie Meyerhofl, «1.8.2, Transactions, Vol. 68, Par 1, 1049, p38. 151. Specificstions for Wire aud Cable with Rubber sad Rubber Tee Tnsslations, Tat edition, 1048, prepuand by. Tasslated Power Cable Enginoors Associaton 188, Specifictions for Vamished Cambric Insulated Ceble, 5th blition, 1018, prepared ly Iawulated Power Cable Kagioesce Association 188, Specifications for Tmpregnated Papersosulated Leal-Covered Cable: "Solid" Type (7th end 8th editions, 1047), "OHL-FMiea”| ‘Type (uh edition, 1947), "Lon-Pressure Gas Filled” Type (iat edition, 1918), prepared by Association of Faison Thais nating Companios. Biiiogrpiaes| 180, Cal-uation of the Eleotical Problems of Uadergzound Cables, bby D, M. Simmons, The Hlecie Journal, Vl. 20, May 1029, 287 114 wi. 143, 8. 181. Undovground Systems Reference Book, NELA Publication No. 050, 1981, Append I 18%, Rating of Cabler in Relation to Voltage, (Bibliography on Dielectrics) by D. M, Simmons, A.L.E-B. Transactions, Vol 41, 1922 p. 601.CHAPTER 5 POWER TRANSFORMERS AND REACTORS Original Authors: J.E, Hobson and R. L. Witzke 'N this chapter are included the fundamental theory, some of the physical characteristies of power trans formers and reactors, No attempt is made to give a come plete exposition of the material. Tt is expected that the listed references will he consulted for a more detailed con- sideration of each section. Although the fundaraental the- ‘ory presented here holds also for distribution transiormers, the standards of operation and present practiees regarding distribution transformer application are not inehided in this chapter. Grounding transformers are inchided since they are ordinarily aaceiated with power aystoms, I. THEORY 1, Fundamental Considerations ‘Refore going into the various problems invelved in the application of transformers and the methods used in ana- Jyzing their effect on system operation, itis well to review bnefly the fundamental theory of transformner action, ‘Two windings on a common magnetic core sre pictured in Fig. 1. Let the number of turns in the P winding bem, ig. 1Two-winding transformer. and the number of turns in the $ winding bem, Assume ‘that there is a flux in the core whieh links both windings and is a sinusoidal function of time, = be sin ot a ‘Then the voltage induced in the P winding at any instant. by the flux is a4. em m8 x 10-4 volts @ = Mise 608 ot X 10-4 Volts where w= 2rf hence, ata ts 08 wt X 10 volts Revieed by: R. L. Witzke and J, S. Williams and the rmy value of this vltage i 2nf, Bay AAdfrsA Base X10 volts (3) frequeney is eles por second ots scetiona! aren of magaetic iret in savare centimeters (assumed uniform). sstmaximun flax density in the core in ines per ouare eeatinete, Similarly, the ma voltage Indaced in the 8 winding bythe flux w wen by 4.44fmA Bou X10-* volts “ "Thus its evident that sinusoidal tox inking a cil in dees in it a voltage which aloo sinusoidal ead which Tage the Hux by 90 electra degree "To pply thobove principle to the operation ofa trans former sor ngnin to Fig. Land consider the 8 winding as ‘open and let @ sinusoidal voltage be impressed on the P winding. The current, J, that flows in the P winding Under tis condition (f5~0) is elle the exeting curent fd sts up am alternating ux abut dha winding, Which Consists of two para mutual x whose path Se whlly in the core and which, therefore, links both ndings, ad 2 Teakage fue wbove path is pul im wr and which tks only the P winding. The ratio of the leakage flux to the mutual fox depeatis on the relative reluctance of te respetive paths, wiych in tues funtion af the nate tion of the sore and the magnitae of the eurent, Te convenient to consider the valtage induced in the P winds ing, by tne fax inking ity ae male up of wo component, one product by the linkages resulting frm the tual flix and the other produced by lege fs: Tn the oF dinary comroercial transformer the lekage fuss eal aut ean be neglected forthe presot. ‘Then the smalls drop in the winding s also ignored, the voiage seduced in the P wining by the matual ux ean, with close approx Imation, beset wqust and opposite the imprased Voltage, Tf, as amvumod, the attr sinusoidal, then the mute flix must also be sinusidal andthe induced volege i given by By. 3), Ey ~4AdfnsA Baa X10 volts. saa X10 volts where, By hypothesis, all of the mutual flux which has just been considered in connection with the P winding must also link the S winding. Hence, a voltage is induced in the 'S winding, which is expressed by Eq. (1), By 4 AA frad Bane X10 volts,Chapter 5 Hf the circuit connected to the S winding ie closed, a current, I, flows and, in the manner alzeady deseribed in connection with the P winding, sets up a mutual and leak- age flux about the winding, The direction of this current, is such that the mutual flux produced by it, opposes that produced by the P winding and it, therefore, tends to, nullify the flax in the core. Consideration of the energies involved shows that an additional component, Z,/, must be added to the current in the P winding before the S winding. is closed, such that the magnetomotive force acting on the magnetic circuit remains unchanged after S is closed. In other words, the resultant fux in the core produced by. the combined action of the currents flowing in tho P and S. windinge must equal the mutual flux present when the S winding is open. Therefore, nilemrals — Rll oy remembering that the fux caused by J, is opposite that caused by Z, which accounts for the negative sign. In a well-designed transformer, the exciting current is small in comparison to the normal load current /',, hence we can assume the total current, [,, in the P winding to be equal to J’, and obtain rat Te ‘The leakage flux produced by [, induses & voltage in the ‘8 winding opposing that produced by the mutual ffux. ‘However, itis small as in the ease of the P winding, and, it neglected along with the resistance drop, permits writing the relation between the P and S voltages as » @ Ee +b, i) The seven equations developed ubove summarize the general relationships between the flux, the induced volt ages, and the primary and secondary voltages and currents involved in tranaformer action. However, they are based on a number of assumptions that, in analyzing the opera- tion of the transformer or of the eystem to which itis con- nected, cannot always be made. A more rigorous develop- ment that takes into consideration the effects of exciting ‘current, losses, and leakage fluxes is therefore required. Referring again to Fig. 1, and considering instantaneous currents and voltages, the classieal equations for the coupled circuits are yt e9= Reig Le! ® where Ry and Re are, respectively, the effective resistances of the primary and secondary windings: Ly and Ls are the self-induetances of the primary and cccondary windings: and Af is the mutual inductance between the two windings. ‘The positive direction of eurrent flow in the two windings i taken such that the fluxes set up by the owe currents will be in opposition, ‘The coefficients Ly, Zs, and M are not constant but vary with the saturation of the magnetie circuit! As previously Power Transformers and Reactors 7 stated, the total flux linking either winding can be div into two components, « leakage flux whose path is wholly or partly in aie and a mutual flux most, of whieh lies in the ‘on core. Furthermore, the mutual coupling between cir- cuits must have an energy component to furnish the icon les in the magnetic circuit, With the above considerations, in mind the equivalent cireuit representing the two coupled windings in Fig. I can be derived?. ‘The equivalent cireuit is shown in Fig, 2(a), where the mathematical artifice of an idea! transformer’ is introduced to preserve actual voltage and current relationehips at the terminals, and to insulate the two windings, ‘The ideal transformer is defined as having no losees, no impedanee drop, and requiring no exciting current, ‘The ratio of transformation for the ideal transformer is N, where (9) @ Fig. 2 Equivalent clrcuite for ewo-winding transformer. (a) Equivalent cireuit im ohms, with magnetiing current con- sidere (b) Kquivalent cireait in ohms, with allimpedancea on the przaary voltage bece (6) Bauivalent cfeuit ia chime, with the magoetising branch eglacted. (a) Equivalent eloult in obms, wlth the leakage Impedance re- fered tothe secondary vltege baz. (6) Equivalent iret in percent98 Power Transformers and Keactors "Tha dant sean branch in Zc moet he sen hat nie rt (4) pi path for the no load, or exciting current of the iransformer. ‘The variation in Mf during the cycle of instantaneous cur- rent and voltage variation is ignored, and a mean value is, used, The branches, Ze= Reso lr—M30) ad, Zs + so{ tau) are essentially constant, rogerdless of i stantaneous current variations, since their corresponding Teakage fuxes lio mostly in ir. Ze and Zs are components ‘of the leakage impedance between the P and S windings such that itt Laem tet sas (10) Zs is defined as the leakage impedance between the P and 3 windings, as mneasured in ohms on the P winding with the S winding short-circuited. Actually it is not possible to segregate Zey into two parts, Zr associated ‘with the P winding and Zs associated with the S winding by any method of test; for example, Ze, the portion of Zes associated with the primary winding, varies with ex citation and load conditions. Tt ie customary, in many, calculations involving the equivalent cireuit, to make 1p, Zr Zoe ~ a) ‘The ideal transformer can be shifted to the right, asin Fig. 2(b), to got all branches of the citeuit on the seme voltage base. Since the impedance of the shunt branch is, lange compared to Zs, it can be omitted for most ealeula~ tione involving teunsformer regulation, and the equivalent circuit becomes hat of Fig. 2(¢). A notable exeeption to thoae eases where the shunt branch ean be disrgarded is the ease of the three-phase core-form transformer excited with zeroequence voltages. ‘This will be discussed in detail Inter ‘The form of the equivalent ciruit given in Fig, 2(6) ean he changed to show the leakage impedance referred to the secondary voltage, by shifting the ideal transformer to the leit, aa in Fig. (@). For this condition Zp, the leakage impedance between the P and S windings aa measured in hms on tho'S winding with the P winding shor-cireuited, is related to Zrs as follows: ee ‘The equivalent circuit using percentage impedances, percentage voltages, and currents in per unit is given in Fig. 2{6). An ideal transformer to maintain transformation ratios is not required 2, Transformer Vector Diagram ‘The vector diagram illustrating the relationship between, the terminal voltages, the internal induced voltages and. the currents in the transformer of Fig. 1 can be drawn direotly from the equivalent circuit for the transformer. ‘This circuit is repeated in Fig. 3(a) and the various voltages, ass (12) Fig. 3 Equivalent circuit and corresponding vector dingram for two-winding transformer. and currents are identified there. ‘The primary and second- ary leakage impedances Zp and Zs are shown separately, and the primary and secondary resistances Rr and Rs are also indicated. Zy and Im represent the core-loas component, ‘and the magnetizing component respectively of the exciting, current [.. The vector disgram in Fig. 3(b) is draven for 8 1:1 ratio of transformation and for a load of lagging power factor. ‘The power-factor angles at the} winding terminals and the S winding terminals are designated in the diagtam, as Op and 8s respectively. Ii. ELECTRICAL CHARACTERISTICS 3. Transformer Impedances "The turns ratio of « tve-winding transformer determines the ratio beoween primary and secondary terminal volt ages, when the transformer load current is zero. However, when Toad is applied to the transformer, the load eurrent, cencotnters an apparent impedance within the transformer which causes the ratio of terminal voltages to depart from the actual turme ratio, This internal impedance consists of two components: (1) a reactance derived from the effect ff leakage Sux in the windings, and (2) an equivalent resistance whieh represents all losses traceable to the flow fof load current, such as conductor 1° loss and stray eddy-current loss. Tinpedance drop is conveniently expressed in percent, ‘and is the ipedanee-drop voltage expressed as a potoent= age of rated terminal voltage, when both voltages are referred to the same circuit; im three-phase transformer banks, it is usually appropriate fo refer both impedance- drop vollage and rated voltage to a line-to-neutral basis. Percent impedance is also equal to measured ohmic im- pedance, expressed as @ percentage of “normal” ohms. ‘Normal ols for a transformer cireuit are defined as the rated current (per phage) divided into ruted voltage (line- toxneutral). Representative impedance values for distribution and power transformers are given in Table 1; for most purposes, the impedances of power transformers may he consideredChapter 5 "Tabu 1—TRANSFORMER IMPEDANCES () Standard Reactances and Impedasces for Ratings 500 leva and below (for 60-eyso teansformore) Rated Vollage Class in Ev Sing |__28 |__| 8 Phase |Aver-[ Aver Aver] Aver-| Aver Aver |Aver-| Aver Kea | sge | age | age | see | age | see | sar | age Rating) Mee | tne | Me | Ime | Te | te | Ree | tne act: | pede | act- | pod | act- | pede | act- | pede noe | ance | ance | ance | ance | Snve | ence | ance % | % | % | % | % | % a firf2z/os|ex| | wo | 1s | 22) 15 | 24 | a4) aa 2s | 20 | 25 | 17 | 23 | 48 | 52 so | 21 | a { 21/25 | ao | sz wo | a1 | 3a) 29 | 32] so | sz sw _| 47 | ss 140 | 50 | 51/52 EES a tage Ne epic as equal to their reactanees, because the resistance euro ponent is so emall. ‘The standard toleraneos by which the finpedanees may vary are +7} percent of specified values for tavo-winding transformers and + 10 peroent for theee- ding, suto, and other non-standard transformers, the pervent resistance of transformers is less consistent, among various designs than is the impedance, and though, the curves in Fig. 4 show definite values for transformer resistance, considerable deviation from these figures is, possible. ‘Transformers can be designed to have impedances within closer tolerances than mentioned above, or impedances ‘outside the normal range, but usually at extra cost. A guide to the impedances of three-winding transformers ® sve tlow (i ide doesnot apply to auto-rane (1) Select a kva base equal to the kva rating of the WA i 0 * | obo a ere a 10 PBiuse PAR ARRS “SP He Fig. 4—Percent resistance of transformers, based on OA kva fatings. Power Transformers and Reactors 99 ‘Tapu 1—~Trawsronwer Derspaxces (Continued) (©) Standard Renge in Impedance: for Twe-Wieding Power Transformers Ratad at 55.C Ri (Both 25- and 6o-eycletraneformers) “apolance Limit ia Percent High Low. : Chas Winding | Winding oa Cs Tosulation | Insulation ow | oa i ae et | ee | tartar | Min. { largest capacity winding, regardless of voltage rating. All impedances will be referred to this base. (2) Select a percent impedance between themedium-volt- age and the high-voltage circuits (Zxn'%), lying between ‘the limits shown for bwo-winding transformers in Teble 1 (3) The percent impedance between the medium-voltage and low-voltage circuits (Zsei%) may lie between the limita of 0.35 (Zun'%) and 0.80 (Zun%). Seloct a value of Zn'% lying within this range.100 Power Transformers and Reactors Chapter 5 (1) Having established Zn and Ziex7%y the percent impedance between the high-voltage snd low-voltage cits ents (Zan.%) 38 dotermined as flows Bau = 110 Zunlo+ Ze) (13) ‘When impedances outside the above ranges are required, suitable transformer ean usually be supplied but probably ft inereased cost. 4. Regulation ‘The full load regulation of a power transformer is the change in secondary voltage, expressed in percent of rated secondary voltage, which occurs when the rated kva output fat a specified power factor is reduced to zero, with the primary impressed terminal voltage maintained constant. Percent regulation ean be ealeulated at any load and any power f Regulation =| pr-+ge+“2 where: “Regulati rapercent resistance or by an approximate formula: (px—gr)") operating current 200 rated current as) jon” is a percent quantity; “loact losses in ow, at rated Kv8, yoy rated kya |g6 RCE SUES OE i REGULATION CHART Teeevoa nu 7 tot E 20 DIRECTIONS:-4 sraaignT FS | RESISTANCE AND REACTANCE fag SEALES WILL GWE NTERCEPTSE Sheen he a Gs a st copycat Wil BY JF PETERS sz corvmonT eter NLCOE. Fe ios 35-3979 40k Fig. 5—Chart for calculating regulation of transformers,Chapter 5 impedance kva, z= percent impedance = PEPE ESS 100 z= percent reastance=V/z=p pcos 8 g=cin 9 8 power factor angle of load (taken as positive when current lags voltage). ‘The full-load regulation of a transformer can be de- termined for any power factor from the chart in Fig. 5; this chart is based on Eq. (14). ‘Typical regulation for three-phase transformers at full load and various power factors is shown in Table 2, These "Tants 2—ArProxiwark REcUtATION FoR 60-CYouR. ‘Tunbe-PHASE TYPE OA TRANSPORMERS A? FULL LOAD yo | Tazaing ercent. Regulation ee? | Power Porevnt Rea ass | actor | 1000 10.000 | 100000) ereent | kv kaka 3 | m a2 jae 90 33 a1 | 100 1 on | us| s0 5.0 as | 0 40 ar | 100 12 os o 0 ea 87 55 0 49 a) 42 100 a a | 06 138 80, 17 | 90 62 | 100 Le | am | oo | | | i 100 i toads teense Bel Taner figures also apply, but less accurately, to transformer banks made up of three single-phase transformers; in this cae the table should be entered with the thren-phase bank va rating, ‘The regulation of three-winding transformore can bo calculated direetly from transformer equivalent cireuita, if the impedance branches and loading for each cireuit aro Known. ‘The regulation of four-winding transformers may ao i ealoulated using formulas developed by R. D- 5. Definition of Efficiency ‘The efficiency of a transformer, expressed in per unit, is the ratio of real power oatput ta power input; Output) Losses Tnpat ~*~ Tuput Total losses are the sum of the no-load losses and loed losses. No-load losses aro eddy-current loss, hysteresis loss, I*R loss eaused by exciting current, and dielectric Eficiency (15) Power Transformers and Reactors 101 Toss; that is, all losses incident to magnetization at full, voltage with the secondary eireuit open. Load losses are PPR toss eaused by load current, eddy-current loss induced by stray fluxes within the transformer structure, and imilur losses varying with load current Norload loses are measured at rated frequency and rated secondary voltage, and can be considered as inde= pendent of Tond. Load losses are measured at rated fre quency and rated secondary current, but with the seeond- ary short-circuited and with reduced voltage applied t0 the primary. Load losses can be assumed to Vary as, the square of the load eurrent. 6. Methods of Calculating Efficiency Conventional Method-This method is illustrated ‘below for a transformer having 0.90 percent no-load loss, and 1.0 pereent load lose at full load. Pereent no-load loss, is determined by dividing the no-load lose in watts by 10 imes the kva rating of the transformer, and the percent Jond lose (total minus no-load) ia determined by dividing the Toad loss in watts hy 10 times the kva mating of the transformer. Note that the no-load loss remains con- stant regardless of the load whereas the load loss varies directly as the square of the load. Percent load tono9 7500 5000 2500 (a) Percent lo ose. 1.90 8022508) Persentnolontion 5050550) Sumof @) end). 150 10827562} Sam of (1) and). 101.50 76082 $075 25562) Dividing 100 tinea ) by 6) 148 140 148 220 @ Subteact (6) irom 100 98.52 $8.60 98.52 97.80 (ficiency) Slide-Rule Method This method the same transformer. ia illustrated for Percent load 100.00 75.00 50.00 2500 (1) Peccent no load oss. 9 80” 5050) Percent load Tos. roo 5e2 25, bee @) Bam of @) and (3). 150 roe 75 3a) Sam of (1) and (3). 10130 76002 B75 98582 () At this point the operations are continued on the slide tale, and are deseribed here for the full load point only 1. Set 5 (sum of no-load and Ioad losses) on D scale. 2 Set 101.5 over this on the (’ scale, 3. Now starting at the right end of seale D, read the first figure (ie, 1) a8 90, the next (i.e, 9) as 91, the next (i.c., 8) 98 92, ete., until 98.52 is read under the left end (Le, 1)'of sale C. ‘This 98.52 is the per- cent efficiency at full load. ‘This procedure is repeated in a similar manner for other Toads. ‘Nore—If the sum of the percent no-load and load loss st full load is I pereont or les, the frst gure at the right fend of D scale (i.e, 1) is read as 99 pervent and the second figure (ve. 9) is read as 99.1, the third figure (ie, 8) is read as $972, ete Ii the sum of the pervent noslosd and load loss is greater than 1 percent as in the ease Mlustrated above, the right ‘end is ead as 90 percent, In calculating the valnes for the ‘other points, judgment will indicate whether 0 or 99 to be used as the first gure on the right end of scale D.102 Power Transformers and Reactors Chart Method-The chart in Fig. 6 may be used to caloulate transformer efficiency at various loads. ‘The pro- cedure ip deseribed in the eaption below the chart. 7. Loss Ratio and Product ‘Maximum operating efficiency for a transformer results ‘when the no-load (constant) losses equal the load (variable) losses. This condition will likely occur at some load less than rated kva: [Fe 1 ENG Vi Chapter 5 a7) L=per unit kva load at which transformer operates ‘most efficiently. so-load losses, kw. Toad losses at rated load, kw. cs re od en io = ind sae a caxLt=Fe (16) no toad Tos EFFICIENCY CHART ° Ty FL LoHD Be va ™ E be fs F 9 bs be fa \ E E E 2 aos bE, bs Es E E Es 4 be E Ee E 995— i [sos E33 4 C 908 - 4 990. E ko - 4 - q Es Fa bs £ F @ en le E 5S f r Hy 2 bu Fa B E E Ls : 4 Feo Faso 2 Eos a bo» a 4 Ls | 2 380: Ee 4 [ Es a8. ES 4 bs fo r 5 ry c « 3B 3 Eo: a4 cr cr? fos ie oa E 8 + ba ba £4 aed Ee was] E J Fes Fas Fo 4 Es 1 4 be a ob. ba E aq E970 4 rs - 3 . 3 E, F ° j-90.0 q E 4 ba. b q bk r Fu wf Fs ba Fist 4 E od Ewe =f Ee E 47 ko bs sr bors boo, [ors 2¢0- Es * 3 open h BY a PETERS Fig. 6—Chare for calculation of efficiency. Directions: A stralght-edge placed between, ‘the known full load copper loss and iron lone points will give Intercepts on the efficiency ‘scales for various Toads,Chapter 5 Power Transformers and Reactors 103 " a Kv cuss] poem Poe a PRS : : | r| CS zo yf SSS i Eo4__| | | SS] z v © a 2" mere 5 oor cy 3 Sahay cass Pe Boe a t E 3 : - e 8 aol HH Eos 8 1s | 8 2s ae | be 50) 2 | “2mm 0 | m | 2 | ow is | soe} 0) | oun] koe | oa | | Ss ts | we) kw | oo”) oe | | at | at ro | ato wo} us | ao | 0 | w vs | so | a0] ae | ao | 1 hs) oe | as | ats | m| so | aw | ao | a0 | ao Bo] ge | fo | fo] & m | xo | do) ao | so | am wo] ms | oo | oo | tw jo | 50 | ao | io | 30 | Be a | ow | oo a | ows wo | so | a0 | om | 30 | ao us| oat | oe ns | a oo | 30 | go) ao | ao | a0 re | or | os | ams | 30 | go) ao] so | eo mo) ge | ie | ae [odo |] ae] ao | a0 na | ewe ros | 30 | 0 30 | 00 ts | oo | 0 [ito | 30. | ann 30 | ao tor | rs (tbe bo) | im 50 | two 199, 35, 690, 690, 1785 | 3.0 1550 | 1785. 3.0 1550, ml ower wacelorimta Presta value for ipulee tc of such appara SSA ae. 30 hy; 28 tins 20 ho 80 em, 28498588 hy he. 1 ca, 0 Tw vale pst th hop wave a ft (2 Xgoancndaadormers fr operatic with mua seily grounded o grounded vroush ve ay haw ren ation he weir, Wh (rere fea ia ore crratin,etptomera ent fe tensa oy Hesperia ete Pt ens st ef aan eve fr te ad alt spon Se pu ae tS ap nga elt ie coy TER aS REACT ve ome ave 120 eyeles for 60-cyele transformers, and 60 cycles for 25-eyele transformers, ‘Combinations and modifications of the tests described above are contained in transformer standard publications, for example ASA C57.22-1048, and these publications should be convulted for detailed information. 16. Insulation Class of Transformer Neutrals ‘Transformers designed for wye connection only with the sutral brought out may have a lower insulation level at the neutral than at the line end. ‘The following rales are n selecting the permissible neutral included ax a guide insulation level: (a) A solidly grounded transformer may have a mini mum neutral insulation class in aceordanee with eoluman, 2 of Table 6. (b) A transformer grounded through a neutral impedl- ‘once must have a neutral insulation clase at least as high, 8 the maximum dynamic volinge at the transformer neu- tral during system short-circuit conditions. In no case should the neutral elass be lower than that given in Column, 2, Table 6. (©) Ifthe neutral ofa transformers connected to ground ‘through the series winding of a regulating transformer, the neutral invululion elass must be st least as high as the ‘maximum raise or lower voltage (phase to neutral) of the regulating transformer. In no ease should the neutral class be less than that given in Column 3 of Table 6. (@) A transformer grounded through the series winding ‘of 8 regulating transformer and a separate neucral imped ‘ance shall have neutral insulation class at least as high as the sum of the maximum raise or lower voltage (line to neutral) of the regulating transformer and the maximum dynamic voltage across the neutral impedance during sys- tem short-cireuit conditions, In no case should the neutral insulation clase be loss than that given in Column 3 of ‘Table 6. (c) If theneutral of a transformer is connected to ground through s ground fault neutralizer, or operated un grounded but impulse protected, the minimum neutralChapter 6 "ana 6a nt aron Cua Ar ease Sam wo ® | @ Gates wintag | Smet | Grandad | “Tow Insulation | $1 oF | Through | Groand Fault ‘Through: i e ranaformer | Htmpulse Protected 12 u 1 88 | 8.66 z we] os a5 ses | wae | g ge | 8 é g | 8 2 a z " 8 Bs Pe 3 ‘ s x ‘ 287 15 | ow insulation class shall be in asrordance with Column 4 of ‘Table 6. ‘VI. TEMPERATURE AND SHORT-CIRCUIT STANDARDS 17, Temperature Standards ‘The rating of electrical apparatus is inherently deter- mined by the allowable operating temperatures of insa- Intion, or the temperature rise of the insulation above ambient tomperature. For transformers and voltage 1 ulators with Class A insulation, either sir or oil cooled, the rating is based on an observable temperature rise (by resistance or thermometer) of 55 C above an ambient temperature at no time in excess of 40 C, and the average during any 24-hour period not exceeding 30 C. ‘Trane formers and other induction spparatus are designed to limit the hottest-spot temperatures of the windings to nat ‘more than 10 C above their average temperatures under ‘continuous rated conditions. The limits of observable temperature rise for air-cooled transformers with Class B. insulation is 80 C by resistance measurement. 18. Short-Cireuit Conditions A proposed revision to American Standand C57.12-1048 (section 12.050) reads in part: “1, Transformers shall be eapable of withstanding with- out injury short eireuits om any external terminals, with rated line voltages maintained om all terminals intended for connection to sources of power, provided: (@) The magnitude of the symmetrical eusrent in any wind- ‘ng ofthe transforiner, resulting {rom the external short circuit, does not exceed 25 times the base current of the Power Transformers and Reactors 109 ‘winding, ‘The initial current is assumed to be come pletely displaced from zero. (b) The doration of the short circuit is limited to the fol- lowing time periods, Intermediate values may be dotermined by interpolation, Symmetrical Current, "Time Period in Any Winding in Seconds 2 20 times base 8 16.6 times base 4 143, of les, tines base current 6 2, Where kva ie mentioned in paragraph 2 the following is intended: When the windings have a self-cooled rating, the kya of the self-cooted rating shall be used. When the windings have no self-cooled ratings, the largest kva obtained from the ratings assigned for other means of cooling by the use of the following Factors aball be used ‘Type of Transformer Multi Woter-cocled ( Dry-Type Forcod-Air-Cooled (AFA) Foreod-il-couled (FOA or FOW) 43, For multi-winding transformers: ‘The have current of any winding provided with ex- ternal terminals, or of any delta-eonnected stabiliz~ ing winding without terminals, shall be determined froma the rated kva of the winding or from not less than 35 percent of the rated kva of the largest ‘winding of the transformer, whichever is larger. “Ta some eases, Uhe short-circuit current, as Himited by transformer impedance alone, will exceed! 25 mes base current. Tt must be recognized that such cases can occur with transformers manufactured according to these stand- ards and that the transformers built uncler these standards fare not designed to withetand such short-cireuit, current.” Under shorteireuit conditions the calculated eopper ‘temperatures for powor and distribution tranaformers shall not exeoed 250 C where Class A insulation is used assuming fan initial copper tamperature of 0 C, or 350 C where Class B insulation is used assuming ‘an initial copper ‘temperature of 125 C. VII. TRANSFORMER TEMPERATURE-TIME CURVES 19. Constant Load A “heat run” of a transformer on test is made to deter- mine the temperature rise of the various parts at rated load. If the test wore macle by applying only rated load, ‘with the transformer at room temperature, thirty hours oF more would be required before stationary temperatures were reached, Such a process would be quite inefficient of time, energy, and in the use of testing facilities, Accelerated, heat ‘runs are made by closing radiator valves, ote., and, applying loads in exeass of rated load until the expected temperatures are reached. Radiation restrictions are then removed, the load redueed to normal, and the test. con- tinued until stsble temperatures are reached.10 It is evident that the temperature-time charactoristies of a transformer eannot be obtained from the accelerated, hoat-run data. Information is secured from the heat run, however, which permits tho temperatures to be calculated, under assumed load conditions. Exact calculations are quite involved, but sufficiently accurate results can be obtained by the use of an approximate method due to 8. B. Griseom for estimating the temperatures reached ‘under variable load conditions, changing ambient tempera tures, ete. Certain simplifying assumptions can be made thal permit s quick estimate of the expected temperatures. Let sransformer load in kva. W = total losses (in kw) at load ‘T,=final temperature rise at load Lin degrees C above the temperature at (0. M-=thermal capacity in kw hours per degree C. ‘=radiation constant in kw per degree C. oil temperature rise in degrees C at time ¢ shove the temperature at t=0. thermal time constant in hours. ne in hours, H If the heat radiated is directly proportional to the ‘temperature rise of the transformer above the ambient, the radiation constant can be obtained from the heat run data for Wand Ty: ww Tr where the temperature at ¢=0 is taken as ambient. Since the total heat generated is equal to the heat, radiated plus the heat storod (heat consumed in caising the temperatures of the vatious parts) ke. (23) ar wernt a ‘This equation ean be volved for 7, giving w ag 25) (26) where = whe trnsfonies tne constant in bows, (27) ‘This derivation may be broadened to show that Ea. (26) is equally correct for the ease where the oil temper- ature rises T and Tp are those above the temperature at, 40, whether the value then is the ambient temperature oF otherwise. ‘The foregoing discussion has been based on the sssump- tion that the temperature throughout all parts of the trans- former is the same, This, of eourse, is not the case. When the transformer load is inereased, the copper temperature is above that of the surrounding parts, and when the load is deereased, the copper tends to he mare nearly the same temperature as the surrounding parts. Also, the top and bottom oil are at different. temperatures. Eq. (25) is, therefore commonly taken as referring to the top-oil Power Transformers and Reactors Chapter 5 temperature rise, that is, T and Tr are defined as before but refer to the top-ol specifieally. Further, the finsl top- oil temperature rise Ty is not directly proportional to the losses for all types of transformers as Bq. (28) would indicate, but is more correctly represented by the relation t= Teo sata ad) “ m=08 for type OA transformers. 0.9 for type OA/FA transformers. 0 for type FOA transformers Tr =final top-oil temperature rise at full Toad in degrees C, ‘The use of this relation when substituted in Eq. (23) indicates that for other than the type FOA transformer the radiation constant k and the time constant Hare not completely independent of load but vary according to a small fractional power of the total loss. However for con- venience in calculations this variation in f and HT is nom mally overlooked and the values obtained from Equations (23) and (27) for the full load condition are taken as constant. ‘The error introduced by the procedure is not large compared to that normally expected in transient thermal caleulations. ‘To determine the temperature rise curve for any Toad L therefore, the radiation constant f under full load condi tions is first, determined from the heat run date using. Eg. (23). The thermal espacity M is dependent on tho thermal capacities of the various parts of the transformer. For convenience it ean be assumed that the transformer parts can be separated into three elements: the core and coils, the ease and fittings, and the oil. Although the core and coils are of copper, iron, and insulation the speeitie heats of those elements do not vary widely. Sineo, also, there isa reasonably eonstant proportion of these elements, in different transformers, a single weighted coefficient of thermal capacity for the eoils and core is warranted. The following relation is accordingly suggested: 1 pg 206 (tof core and coils) -+0.04 (wt. of case and Sittings) +0.17 (wt. of oi)] @) Here the coefficients of the last two terms are aleo weighted to make further allowance for the fact that all parts of the eas» and fittings and the oil are not at a uniform tempera- ture. ‘The values of & and M found as above may bo substituted in Eq, (27) to obtain H. The value of T» for the desired load Z is determined next by substitution of heat run data in Eq. (28). The quantity W for the load L may be evalwated by the relation =| BE ; (full load copper loss} wri)» tom (wad less) (30) ‘The quantities H and T'r may now be substituted in Eq. (26) from which the top-oil temperature-rise curve may be plottod directly.Chapter 5 For example, a €000-kva, three-phase, self-cooled, 24 000-5040 volt transformer has the following full load. performance data as supplied by the manufacturer: Tron loss= 10 920 watts, Copper loss= 48 S40 watts. ‘Total= 54 460 watts ‘Top-il rise=40C rom heat-run test data). LY copper rise= 46.3 C. HY copper rise= 43.3 ©. We. of core and coils 25 000 pounds. We. of cage and fittings ~ 18 000 pounds, We. of oil 17 400 pounds. Prom this information the time constant H may be evalu- ‘ted and the expression for T obtained for the loarl J equal to the rated load. W_ssat T, 40 '= 1.36 kw por degree C. {0.06 X25 000-+0.04 x 18 000-++0.17X17 400] Tao! 5.18 kw hours per degree C. P= Tell eH) = 40(1~ 9), ‘The full load top-oil temperature rise curve shown in Fig. 16 was caleulated from this relation, ‘To plot the top-oil temperature-rise curve for half-lond conditions for this transformer the same time constant # is used as found above. From Eq. (28): r-wo(® 5)°X48.544-10 oe ~102( ‘The curve represented by this equation also appears in Fig. 16. ‘The rise of the hottest-spot copper temperature above the top-oil temperature is known as the hottest-spot cop- ber gradient and at fll load may be estimated from the relation Guay = Count (1) where: Guy = hottest spot copper gradient at full load in degrees C. Geaw=apparent copper gradient at full load in degrees ©. Av 10 C for type OA and OW transformers, =10 C for type OA/FA transformers. =5 C for type FOA and FOW tru with directed flow over evils ‘The apparent copper gradient: at full load (Gow) is the difference between the average copper temperature rise and the top-oil temperature rise, both of which are de- Power Transformers and Reactors ut 404 4 / 4 coxp [——~BtLiete ‘TOP-O1L. TEMPERATURE RISE-DEGREES CENTIGRADE 7 7 30 TIME=HoURS, Fig. 16—Top-oll temperature rise versus time, for a typical Transformer. ‘terminod during the heat-run. The average copper tem- perature rise above ambient at full load is required by. standards not to exceed 55 C for class A insulation. The use of that, value to obtain the apparent copper gradient will generally lead to overly pessimistic results sinee the setual value of the average copper temperature rise is oruully below dhe lisit, ‘Phetefure it is advisable bo use the vahie measured on the heat run and obtained from the manufacturer For any load 1, the hottest-spot copper gradient may be calculated from the relation Loe ms (an Taal a) From the pesformance date of the transformer previous vite Geuy=46.3-40=6.3 C for the LY winding, Gnan=6.3410=163 C. ‘The hottestapot copper temperature for fl-lod i thus 16.3 C shove the top-oil temperature, For, sty, hall-load, Bg, (82) must be used to obtain Gg =183X (0.5) =84 C. Tt isnot feasible in s study of this kind to kegp track of short time variations of copper or hottestspot tempera- ture, od it is suggested if itis desirable to show roughly how this varies, a time constant of 15 minutes be used. (32) GansPa ou Trainers ed Rar Chapt 6 ~ TT T | | 1] pS | 5 en ‘ J i - ! * ~ T 1 Lt 1 | Fig, 17—Step-by-atep graphical caleutation o 20, Variable Load A step-by-step analysis using Eqs, (28) to (32) ean be made to consider conditions of variable load, changing aarabient temperatures, etc, The method of approach is based on the fact that the initial rate of change af tempers- ture is the slope of a line joining the initial and final tem peratures, the two temperatures being separated by u time interval equal to the thermal time constant of the trans- former, 3 before Ty is caloulated fiom hwat run data and the total loss for each load condition through tho aoe of Eq. (28). The loss WW is obtained from Eq, (30). The final top-oil temperature is then found by adding Tp to the ambient temperature. Sinee the load is varying, the final temperature cannot be reached for each load condition and the step-ty-atep analysis must, be employed to obtain the top-oil temperature curve. Points on the hottest-spot ‘temperature time curve may then be obtained by adding the hottest-spot copper gradient (a for exch load to the top-oil temperature at the time corresponding to the load for which the gradient was calculated. Gi is ublained in the same manner as previously outlined. To illustrate the step-by-step method, the oil tempers- turestime eurve for the 6000-Kva transformer previously described will be calculated, stasting with an oil tempers ture of 55. for an assuna Toal eyle a tabulated it the rdjgcent colin, Figure 17 illustrates the uso of the calenlated data in the raphical step-by-step process to plot the curve of top-oil ‘temperature with time and the mannerin which thehottest- spot gradients are added to obtain the hottest-epot ton peraturetime curve, ‘The accuracy can be increased by Using shorter time intervals. noe temperatures under changing load conditions. Fa SS a |e 2 Wie | 4.76] 2.80 : ur far |e 8 19.2 ag2 | 54 4 me | BS 2/83] 88 3 Pa | a VII, GUIDES FOR LOADING OIL-IMMERSED POWER TRANSFORMERS 21. General ‘The rated kva output of a transformer is that load whiel it can deliver continuously at rated secondary voltage without excreding a given temperature rise measured under prescribed test conditions. ‘The actual test tempera- ture rise may, in a practical case, be somewhat below the established limit because of design and manufacturing. toleranes The output which a transformer can deliver in service ‘without uudue deterioration of the insulation may be more. cor less than ite rated output, depending upon the following.Chapter 5 design characteristics and operating conditions as they exist at a particular timet (1) Ambient temperature. (2) Top-oil rise over ambient temperature. () Hottest-spot rise over top-oil temperature Qhottest-spot eopper gradient). (4) Transformer thermal time constant, (5) Ratio of load Joss to no-load loss. 22. Loading Based on Ambient Temperature Aircooled ol-immersed transformers built, to meet es tablished standards will operate continuously with normal expectancy at rated kva and secondary voltage, pro- ing the ambient air temperature averages no more than 30 C throughout a 24-hour period with maximum air temperature never exceeding 40 C. Water-cooled trans- Fonners are built to operate continuously at rated output with ambient water temperatures averaging 25 C and, never exceeding 30 C. When the average temperature of the cooling medium is ] vps ‘GROUNDING TRANSFORMER, (GROUNDING TRANSFORMER ‘WINDINGS ORAAN PARALLEL, [ARE ON THE SAME CORE | (e) SGHEWATIC WINDING ARRANGEMENT OF AN INTER= (CONNEGTED-STAR GROUNDING TRANSFORMER OF THE TTWREE-PHASE. CORE-FORM CONSTRUCTION (4) EQUIVALENT CIRCUIT OF AN INTERCONNECTED-STAR ig. 27--Sear-delta and zig-zag grounding transformers,120 cieney become less apparent as the transformation ratio increases, 20 that. autatransformers for power purposes are usually used for low transformation ratios, rarely exceeding tol. XI, GROUNDING TRANSFORMERS A grounding transformer isa transformer intended solely for establiching « neutral ground eonneetion on a throo- phase system. ‘The transformer is usually of the stardolta fr invereonneeted-star (rig-zag) arrangement as shown in Fig. 27 ‘The kva rating of a three-phase grounding transformer, or of a grounding bank, is the product of normal line-o- neutral veltage (kv) and the neutral orground amperes that the transformer is designed to carry under fault conditions for a specified time. A one-minute time rating is often used for grounding transformers, though other ratings such as those suggested in AIEE Standard for “Neutral Grounding Devices” (No. 32, May 1047) ean be specified depending ‘upon the probabie duty to be imposed on the unit in Rated voltage of a grounding transformer is the line-to- line voltago for which the unit is designed, When operated at rated three-phase balanred voltage, only exeiting eurrentcireulates in the windings of a ground ing transformer. Current of appreciable magnitude begins to flow in the grounding eireuit only when a fault involving ground develops on the connected systein. Grounding transformers, particularly the zig-ang type, normally are designed s0 that rated neutral eurrent flows when a solid single-ine-to-ground fault is applied at the transformer terminals, assuming supply voltage to be fully maintained, Thisisequivalentto 100-percent zero-sequence voltage impressed at the transformer terminals resulting in the circulation of rated neutral current. ‘Transformers 50 designed are stid to have 100-pereent impedance based on rated kva and rated vltoge. Sometimes a resistor or other impedance is connected in the transformer neutral, and in these cases it may be desirable to specify that the grounding transformer shall hhave less than the conventional 100 percent impedance Equivalent circuits for star-delta and rig-2ag groundis transformers with external neutral impedance are includ od in the Appendix. ‘Because a grounding transformer is a short-time device, ite sizo and cost aro leas than for a continuous duty trans. former of equal kva rating. The reduced size can he estab- lished in terms of an “equivslent two-winding 55 C kva” U, by applying a reduetion factor K to the short-time rated kva of the grounding transformer, and this reduced kkva can be used for a price estimate, U,> Ue xX Ke for a three-phase grounding unit. (66) U,=3Uq XK; for a bank of single-phase grounding unite 7) where U, =equivalent two-winding 59 Ckva, three-phase. Usa = (ine-to-neutral kv) x (rated neutral amperes) ‘Values for K are listed in Table 12 for various types and Power Transformers and Reactors Chapter 5: Tae -ctons vox Darzauinine Bauivatanet ‘TwosWisoine 66 nva oF GROUNDING TRANsroRERS" stanDetea| ‘ig. 2g Comnsetion ratag | Orme” [ar | 38m Ton AE gE 48k | kr | ear 10 seconds (0.066 | 0.076 | 0.080 | 0.085 | 0.092 imioite | 0.170 | 0-404 | 0-110 | 0.115 | 0-118 | 0-122 2minutes | 0.240 | 0.139 | 0-158 | 0.360 | 0 107 | 0-17 Brmiowter | 0.295 | 0.170 | 0.187 | 0.198 | 0.204 | 0.212 ‘minutes | 0-340 | 0.196 | 0.206 | 025 | 0.335 | 0.245 S minutes | 0.380 0.204 | 0.2 0.220 | 0.242 | 0. Ki, For A Single Phase Unit (One of thee ia « bank) Iminue | 0.087 | 0.033 | 0.087 | 0.040 | 0.088 | 0.040 2minuies | 0.080 | 0.046 | 0.051 | 0.055 | 0.060 | 0.064 3 minutee | 0.088 | 0.057 | 0.084 | 0,088 | 0.074 | 0.080 minutes | 0.113 | 0.068 | 0.073 | 0.078 | 0.086 | 0-091 Sminuies | 0.127 | 0.073 | 0.082 | 6.058 | 0.005 | 0.102 Ps ae APRS a eee eo Gerais Garay oacnacica itis eee eee ce classes of grounding transformers; the table includes values, for both three-phase and single-phase units, though the single-phase type is uncommon. Conventional power transformers may be connected to serve solely as grounding transformers, but the current and, time ratings for grounding service are open to question depending upon the form and details of construction. When. these modified ratings are desired, they should be obtained from the transformer manufacturer. ‘Star-Delta Impedances—The impedance to zero- sequenee eurrents in each phase of a solidly grounded star- delta grounding hank made up of single-phase units is equal to Zrs, the ohmic leakage impedance between one primary (star) winding and the corresponding secondary (delta) winding Lon Des Percent sero-sequenice impedance is normally expressed in ‘terms of short-time kva and Tine-toline voltage: go; Foe Uo Bo vo xeet In a three-phase olardelte grounding transformer Zs may be smaller than Zs by an amount depending on the form ‘of core construction: a typical ratio of Za to Zes is 0:85, ‘though variation from this value for diferent designs isikely. Zig-zag Impedances—The impedance to sero- ecquence eurrente in each phase of « solidly grounded fig-aag bank ean be derivad on s. theoretical asia by reference to Fig. 27(), 9) Eye TeX Zea ent ep (60) Ee Fintan dn (1) Percent zero-seqquence impedance for the aig-2ag connes-Chapter 6 tion is normally expressed in terms of short- line-to-line voltage: XII. TAP CHANGING UNDER LOAD The modern load tap changer had its heginning in 1925. Since that time the development of more complicated transmission networks has made tag) changing under Toad more and more essential to control the in-phase voltage of power transformers, and in other cases to control the phase angle relation. ‘Tap-changing-under-load equipment is ap- plied to power transformers to maintain a constant second ary voltage with a variable primary voltage; to control the secondary voltage with a fixed primary voltage; to control the flow of reactive kva between two generating systems, cor adjust the reactive flow between branches of loop ci cuits; and to control the division of power between branches, of loop eireuits by shifting the phaso-angle position of transformer output voltage Various types of tap-changing equipment and cireuits are used depending upon the voltage and kva and also upon whether voltage or phase angle control is resjuired. Uniler-load-tap-ehangers are built for 8, 16, and $2 steps, with the trend in recent years heing toward the larger number of steps so as to kive a finer degree of regulation. ‘The usual range of regulation is plus 10 percent and minus 10 percent of the rated line voltage, with plus and minus 7A percent and plus and minus 5 percent being second and third, reepectively, in popularity. ‘The 32 step, plus and ‘minus 10 pereent, tap-changini-under-load equipment has such wide acceptance as to be considered standard for many types of transformers. PREVENTIVE. AUTO TRANSFORMER. TRANSFER SHITOMES 7 q 7 ‘SELECTOR SWITCHES SEQUENCE OF OPERATION Posrrionl [2] 3]<[ 5 | 7a] spol [alsa ‘SwiTeH=1[ 79] = =a [ofo t Stet i =a fe = sie =a oj0{0] alot il a e/a} jololot lato fo] =sl Jolefol [ofofel [ol Stet Tot 1et To} ‘O=SWiTGH Close Fig. 28—Seventeon position, single-phase, Type UT tap, ‘changer. =H a] lst lotto) Power Transformers and Reactors 121 31. The UT Mechanism Figure 28 illustrates schematically the operation of the type UT mechanism for changing taps under load. ‘Taps from the transformer winding connect to selector swit es 1 through 9. The selector switches are connected to lead transfer switehes R, S, and 7. The-connections for the tap changer positions are shown on the sequence chart of Fig. 28. The sequence of switching is so coordi- nated by the tap changing mechanism that the transfer switches perform all the switching operations, opening before and closing after the selector switehes. All arcing is thus restricted to switches 2, 8, and 7, while switches I to 9 merely select the transformer tap to which the load, is to be transferred. ‘When the tap changer is on odd-numbered positions, the preventive auto-transformer is ehort-cirenited. On all even- numbered positions, the preventive auto-transformer bridges two transformer taps. In this position, the rela- Lively high reaotance of the preventive auto-transformer to cireulating currents between adjacent taps prevents damage to the transformer winding, while its relatively low impedance to the load current permits operation on this position to obtain voltages midway between the transformer taps. 32. The UNR Mechanism Fig. 20 shows schematically the diagram of connections ‘and sequence of operations of the type UNR tap changer. ‘he operation of the scleetor and transfer switehes is ex- actly as desotibed for the type UT tap changer. But the type UNR tap changer also has a reversing switeh which reverses the connections to the tapped section of the winding so that the same range and number of positions REVERSING SwrTOHES PREVENTIVE AUTOTRANSFORVER ‘SWITCHES. ‘eugeror smrenes BosiTion| 1 [eTS1 ST s]eTF [ooh jehaTAPEpE ‘SMTGH=I[OLO] | | iolo lolol ioj9l6|9jolo10076} oI lojo}o[9) ol jot ‘olelo}_lolalo: a ‘of lolol Jol io) ‘O-EWITEH CLOSED Fig. 29 Seventeen poaltion, single phase, Type UNR tap ‘changer122 can be obtained vith one-half the number of tap sections, for twiee the range can be obtained with the same number of taps. The reversing switeh is a close-before-open switch, which operates at the time there is no voltage across its contacts, 33. The URS Mechanism ‘The type URS load tap changer is applied to small power transformers and Jarge distribution transformers. The transfer switehes are eliminated, and each selector switch, fserves as a transfer switel for the tap ta which it is con- nected. ‘The schematic cireuit diagram and operations sequence chart ia shown in Fig. 30. evERSING SurtCH——p—————} seLectoR-Teansren-2) SWITCH ‘SEQUENCE GF OPERATION ah labhcderomtap hia aneenh| 2 PBR aloo Fig, 30—Thiery-three position, single-phase, Type URS tap. changer. Physically, the stationary selector awiteh contacts are arranged in circles, one for each phase. The moving se- lector switch contacts, as they rotate about a eenter shaft, both select the taps and make eontaet with them. The Feversing switeh operates when the selector switches are (on position 17, at which time there is no current through ‘the reversing switches and therefore no areing on them, ‘The URS tap changer, like the other load tap changers, cean be equipped for hand operation, remote manual opera” tion, or for full automatic operation under the eonsral of relays, XIII REGULATING TRANSFORMERS FOR VOLTAGE AND PHASE ANGLE CONTROL _, Consider two systems A and B in Fig. 31 connected by a single transmission circuit, L and B may both be generat ing units, or one of them may be a generating unit and the other a load. Should -I generate 10 000 kilowatts in excess of its own load, there can be but one reault, the 10 000 Power Transformers and Reactors Chapter 5 soe o re) Fig. 31—Power Interchange between systems: (a) Two systems with ts, (bs Vector diagram of voltages during interchange of power. 0) Intrustion of aa in phase voltage, BC, to careos for exees- sive voltage drop. Kilowatis must go over the tie line to B. An inerease in generator output by -l must be accompanied by a corres- ponding deerease in output (increase in input) by B if there Js to be no change in system froquency. The of power from 4 to B results in a difference in magnitude betsseen terminal voltages and also a shift in phase angle, 9s illustrated in Fig. 31 (bi. 40 ‘s the terminal voltage at A,BU isthe terminal voltage at B, 4 Bis the veetorial vole= ‘age drop from .{ to 8, created by the flow of load current { and # is the phase~angle difference between terminal volte ages. In actual practice the phase angle is uot allay appa rent, but thedrop in voltage, 12", is often objectionable. An attempt to maintain satisfactory terminal voltages at. 4 aand B will often result in undesirable circulation of reactive kkva hecween the systems, The tlow uf power from -l 10 B, or vice versa, is determined by the governor settings. ‘The flow of reactive power over the intervonnevting ‘ine is determined by the terminal voltages held by the machine excitations at 4 and B. Excessive voltage drop between the systems can be readily corrected by transformer taps, of a fixed nature or by tap-changing equipment, introduc~ senes ° TRANSFORNER, ¢ . m ¢ a j T 4 THEORY i Bair Tmansronwer Pig. 32--Regulating cransformer for voltage control.Chapter 5 ing an in-phase voltage, BC, to compensate for the voltage drop and bring the terminal voltage at B to desired value Figure 32 is a simplified sketeh of a regulating transformer for voltage control, using an exciting autotransformer with, automatic tap changing equipment indicated by the arrows, Consider three systems intercormected with each other 0 that the interconnections from A to B, from B to C, and from C to A form a closed loop, as in Pig, 33 (a). An ) c & A 2 2 ° ° tb) te) Fig. 83—Power Interchange with three Interconnected ‘eyateme. entirely new cloment, enters, and adjustment of governors will not entirely control the flow of power over any one of the interconnecting lines, An attempt to adjust load on, the tin hetween two systems results in & change of load on, the other two tie lines. With the tie line from B to C open, and with power transmitted from 4 to B, the terminal voltages of A and (will be equal and in phase, with no power being transmitted from A to C, or vice versa (see Big. 33 (b}). There now exists between # and C a difference in voltage and a difference in phase angle, If the ti lin hetween B and ( is closed under these conditions there is a redistribution of power flow between A and B, part going over the line from A to B, and a part of the power going irom A to B over tho lincs A-€ and C-B (eee Fig. 83 (e)) ‘The distribution of power, bath kw and reactive kva be- tween the various lines is determined solely by the relative impedances of the interconnecting lines If at the time of closing B-C' an adjustment of trans former taps were made, or a regulating transformer for voliage control were inserted in. the Toop, it would be possible to make the voltage at C equal in magnitude to that at. B but it would not have the same phase relations ship. ‘There would still be a flow of power from 4 to C and from C lo B. ‘Conditions similar to that just described scour on inte connected systems involving loop circuits. To control the circulation of kw and prevent overloading eertsin lines it is often necessary to introduce a quadrature voltage, any pluce in the loop, by the use of a regulating transformer for phase-angle control. This differs from the usual star- Power Transformers and Reactors 123 delta power transformation in that the angle of phase shift, of eurcont and veltage is not fixed but depends on the tap position. Figure 34 is a schematic diagram of a typical regulating transformer for phase angle control. we (Regula le control. nf transformer for phase In general the distribution of real power flow over tho various interronneetions found in, loop circuits ean be eon wolled by regulators for phase-angle control. The flow of reactive kva ean be controlled by regulators for voltage control. The proveding statements follow from the fact that transmission-cirenit impedances are predominantly reactive. The voltage regulator introduces series in-phase voltage into the loop, and quadrature current (reactive va} is cireulated around the loop since the impedances are reactive. The rgulator for phase-ange control intron ‘duces a quadrature series voltage in the loop resulting in the low of currents lagging the impressed voltage by nearly. 90 degrees, or the viteulation of in-phase eurrents (kw). For the ease of correcting the voltage for line drop, & simple voltage contrel equripment can be used. This simply adds or subtracts a voltage in phase with the system volt- age. For the ease of phase-angle control, the equipment ‘ean be identical excopt the voltage selected fo add or sub- tract is in quadrature. As the earlier discussion showed, thore are eazes whore both voltage and phase angle control semies ° ‘TRANGFORWER Dy exornins wows ig. 35~-Regulating transformer for independent phase-angle sand voltage control.Fig. S6-—Regulating transformer for voltage control, rated. 20 000 kva, 12.7 ky, plus or minus 10 percent. are required. There are a number of combinations of con- nections to accomplish this, one of them being shown in Fig. 33. Where the voltage and phase angle bear a close relation, one mechanism may suffice. However, where com- pletely independent control is desired, two mechanisms tovo regulating windings and one series winding, oF vith one regulating winding and two series windings ave recoseary. If it ia desired #9 close the loop, and the Bow of both real and reactive power over the various lines forming the loop must be controlled, tne economical loea- tion for the control equipment is at the point of lowest load to be transferred. This may dictate the location in a loop, unless when in tying several companies cogether the boundary between systems determines the location. ‘The voltage to be added or the phaseangle shift that must bye obtained can be determined by calculation, considering the impedances of the tie line and the load conditions in the loop. When such caleulstions become involved, the use of the network calculator provides a quick and aceure ate tool for obtaining the olution Sevoral common connestions used for regulating trans- formers providing voltage control, phase angle control, or combined voltage and phase angle control, are tabulated in the Appendix under Equivalent Cireuits of Power and Regulating Transformers, The equivalent eixcuits of the regulating transformers to postive-, nogative-, and zero- sequence are given. It should be noted that the equivalent circuits for phase-angle control regulators involve an ideal transformer providing a phase shift of voltage and current. Positivesequence voltage and current are aways shifted by the same angle in the same disection. ‘The angular shit for negativesequence voltage and current is always equal to the angular shift for postive-sequence, but sin the oppo site direction. Zero-sequence currents and voltages do bot tundergo an angular shift in being transformed. For ex- Power Transformers and Reactors Chapter 5 ample, refer to F-7 in Table 7 of the Appendis, which is the regulating transformer for phase-angle consral shown in Fig, 3, For pesitive-sequence. neglecting regulator imps Bl=NesEynV en a=tan ty In For negative-sequence, neglecting regulator impedance: ViF8n For zero-sequence, neglecting regulator linpedance: For this regulator zera-sequence voltage and current are not transformed; Z) flows through the regulator as though it were a reactor. Tt happens with several connections of regulating trans- formers that zero-sequence voltages and eurrents are not transiormed at all, as ia F-7; or are transformed with a different transformation ratio than for positive- or nega- tive-wequence quantities as in G-1. This phenomenon, and the use of the sequence equivalent cireuits for regulating transiormers has born discussed in papers by Hobson and Lewist, and by J. E XIV. EXCITING AND INRUSH CURRENTS. IF normal voltage is impressed actons the primary ter minals of a transformer with its secondary open-cireuited, a small exciting eurront flows. This exciting current con- sists of two components, the !oss component and the mag nevizing component. ‘The loss eorpponent is in phuse with the impresoea voltage, and ite magnitude depends upon the no-loed losses of the transformer. The magnetizing, oom- ponent lags the impressed voltage by 90 electrical degrees, ‘snd its magnitude depends upon the number of turns in the primary winding, the shape of the transformer satura- tion curve and the maximum ux density for whieh the transformer was designed. A brief discussion of cach of these components follows 34, Magnetizing Component of Exciting Current If the secondary of the transformer is open, the trans former ean be treated as an iron-core reactor. ‘The dif- ferential equation for the cizeuit consisting of the eupply and the transformer can be written aa follows: (69) de enki e-instantaneous value of supply voltage ‘instantaneous value of current R= offeative resistance of the winding nstantanoous ux threading primary winding rimary turns where,Chapter 5 Normally tho resistance, 2, and the exciting current, é, are small. Consequently the Ri term in the above equa- tion has little effect on the flux in tho transformer and can, for the purpose of diseuesion, be neglected. Under these conditions Bq. (69) ean be rewritten: 70) the supply voltage i a sine wave voltage, = VIE sinlat 1), ay where, Z-rns value of supply voltage o=2rf Substituting in Ba. (70) Vib a ae sin(at+n) nits Solving the above differential equation, = VBE coslul +N Hh va) In this eotation, 28 as (ot 8) ia the normal steady-state flux in the transformer eore. The second term, ¢,, represents a transient component of Bux the ‘magnitude of whieh depends upon the instant at. which the transformer is energized, the normal maximum flux and the residual flux in the core at the time the transformer is wraesseo outage, sreaoy stare (0) rmararonven son conve Power Transformers and Reactors Ly anv 125 energized. Under steady-state conditions this component. js equal to zero; the magnitude of 4, is diceussed in Soc. 38. From Eg, (72) it can be seen that the normal eteady- state ux is @ sine wave and lags the sine wave supply voltage by 90 degrees. ‘The supply voltage and the normal flux are plotted in Fig. 37 as a function of time. If there were no appreciable saturation in the magnetie circuit in a transformer, the magnetizing eurrent. and the Aux would vary in direct proportion, resulting in a sinu- soidal magnetizing current. wave in phase with the flux However, the economic design of a power transformer re- ‘quires that the transformer iron be worked at the eurved ppart of the saturation curve, resulting in sppreciable sat- ‘uration. Under this condition the magnetising current is not a sine wave, and its shape depends upon the satu- ration charaoteristica (the B-H curve) of the transformer magnetic eirénit. ‘The shape of the current wave ran be determined graphically as shown in Fig. 38. In Fig. 38(b) are shown the impressed voltage and the fux wave legging the voltage by 90 degrees, For any flux the corresponding value of current can be found from the B-lf curve, Pole lowing this procedure the entire current wave cnn be plotted, ‘The current found in this manner dees not con- sist of magnetizing current alone but includes a. loss com ponent required to furnish the hysteresis loss of the core. Trowever, Lis component is quite sinall in comparison to the magnetizing component and has little effect om the ‘maximum value of the total current. A study of Fig. 88 shows that: although the flux is a sine wave the current is a distorted wave. An analysis of this current wave shows that it contains odd-harmonie com- ponent of appreciable magnitude; the third harmonic component is included in Fig. 38. In a typical case the harmonics may be as follows: 45 percent third, 15 pereent fifth, three percent seventh, and smaller percentages of higher frequency. ‘The above components are expressed in, percent of the equivalent sine wave value of the total ‘exciting current. ‘These percentages of harmonie currents will not change much with changes in transformer terminal voltage over the usual ranges in terminal voltage. In Fig. 89 are shown the variations in the harmonie content of the exciting current for a particular grade of silicon steel. (0) vous. Fuue axa ewer maven Fig. 88—Gruphieal method of determining magnetizing current.Power Transformers and Reactors Qik | mr Tot [ ety 2 ct t ea a5 PERCENT oF eeuvaL ame wwe Fig. 40—Raciting current 4s. terminal voltage. The above Fig. 9—Harmonie content of exciting current fora particular ‘Brade of silicon steel. 35. Loss Component of Exciting Current ‘The no-load losses of a transformer are the iron losses, fa stnall dielectriv loss, and the eopper loss caused by the exciting current. Usually only the iran lasses, Le, hys- teresis and eddy current losses, are important. ‘These Josscg depend upon frequency, maximum flux density, and the characteristies of the mognetie citvuit In practice the iron losses are determined from labors tory tests on samples of transformer steel. However, the formulas given below are useful in showing the qualitative tlfeet of the various factors on loss Tron loss (73) Wes KP EB og walls per Ib Wy hysteresis lass W,=eddy current loss Fafrequency ‘= thickness of laminations Bags = masimurn flux density Ky, Ko and = are factors that depend upon the quality of the steel used in the core. In the orignal derivacion of the hysteresis loss formula by Dr. Steinmota, v was L6 For modern steels 2 may have a value a high ws 3.0. The itoa loss in a 6D-eycle power transformer of modern design 48 approximately one watt per pound. ‘The ratio of hyse teresis loss to eddy curvent loss will Ie on the order of 30 vith silicon steel and 3; with oriented stel. These Siures should be aed af a rough guide only, as they vary oor siderably with transformer design 36. Total Exciting Current As discussed above, the total exciting current of a trans- former includes a magnetizing and a las component. ‘The economic design of a transformer dietates working the iron at the curved part of the saturation eurve xt normal volte 22; hence any increase in terminal voltage abwve no, will greatly imerease the eve ting current. In Fig. 40 the exciting current of typical transformer is given as a function of the voltage applied to its terminals, ‘The exciting current increases lar more rapidly than the term= curve applies for one particular design of tranaformer: the ‘thapeo! the curve may vary considerably depending upon the ‘drade of stool and the transformer design, al voltage. For example, 108-pereent terminal voltage results in 200-pervent exciting current, 37. Typical Magnitudes of Exciting Current The actual magnitudes of exciting currents vary over fairly wide ranges elepending upon transformer size, volt= age class. ete. Ta Table 13 are given typical exciting currents for power transformers, ‘The exciting eurrents Lireetly with the voltage rating and inversely with rating, Tamer 13 ‘TypteaL BxcrriNe CunaeNT VLvEs soe SincuE-Puase POWER TRANSFORMERS in percent of fl load current) “The following values should be considered ae very approximate for ‘average standard design und re predicated on prevailing perform ‘anve eharaetorstos, Test values will a rue come below these values iit phasor minus variation mist be expected depending ‘upon purchaser's requirements, Shout cloter estimating data be reguited, rhe matter shoud be rfored tothe proper manulactueee’s design eines ‘Voltage Class (Pull Tnautationd 6 Kv 25 Ky 60 Kv 68 Kv 161 Kv/290 Ke F871 8% 49%) | jas jae jas) | sa jaa [as | boas [ar 2s sag womo igo aaa ise ome 25.000 [ole igh lan Jae as somo || 392. Rahat Faia 38. Inrush Current When transformer is first enorgized, a transient ex citing current tows to bridge the gap between the con- ditious existing before the transformer is energized and the conditions dictated by steady-state requizements, For any given transformer this transient eurrent depend tupon the magnitude of the supply voltage at the instant the transformer is energized, Ube residual flux ia the core,Chapter 5 sand the impedance of the supply circuit. Often the mag- nitude of this transient, current execeds fullsload current, and may reach 8 to 10 times fulload current. These high inrush eurrents are important principally because of their elect on the operation of relays used for differential pro- tection of transformers. In studying the phenomena that oceur when a trans- former is energized it is mare satisfactory to determino the flux in the magnetic cireuit first and then derive the cur- rent from the flux. This procedure is preferable because the flux does not depart much from a sine wave even though the current wave is usually distorted. ‘The total flax in a transformer core is equal to the nor- mal steady-state flux plus a transient. component. of flux, fas shown in Eq. 72, ‘This relation ean be used to determine the transient flax in the core of a transformer immediately Vie after the transformer is energized. As Y=" represents the crest of the normal steady-siate Mux, Eq. (72) can be rowsitten, 6. £08 (wlth) toy wa) where At (75) where qu=transformer residual ux =the coe N=steudy-state lux at (= ‘9 initial transient lux. In the above equation the angle 2 depends upon the instantaneous velue of the supply voltage at the instant the transformer is energized. If the transformer is ener- gized at zero voltage, ) is equal to 0, whereas if the trans- former is energized where the supply voltage is at a posi= tivemaximum value, is eyusl to 90 degrees, Assume that ‘a transformer having zero residual flux is energized whon the supply voltage is at a positive maximum, For these con- ditions @ and cos \ are both equal to 2670 60 gu is also equal to zero. Tho transformer flux therefore starts out under normal conditions and there would be no transient, However, if a transformer having zero rusidusl is enengized, at zero supply voltage the following conditions exist: Gn 608 (a) +h 76) ‘The flux wave represented by Eq, (76) is plotted in Fig. dla. ‘The total flux wave consists of a sinusoidal flu, wave plus a d-e flux wave and reaches @ erest equal to twice the normal maxirour flux, In this figure the trao- sient flux has heen assumed to have no decrement; if loss 1s considered the transient lox docreases with timw and the crest value of the total fux is less than shown. In Fig. 41 (b) similar waves have been plotted for a transformer baving 60 percent positive residual flux and energized at zero supply voltage. Sixty percent residual Dux hits been Power Transformers and Reactors 17 roTaL FLUX TRANSIENT FLIX on Nee OY oe oncerte (a) PRIMARY CLOSED AT ZERO VOLTAGEZERO RESIOUAL FLUX TOTAL FLUK 169, esioual yee Sorrow 2. aretoe stare . voLTAse (b) prmuny cLosco aT 2680 VOLTAGE= 60% POSITIVE RESIDUAL FLUX. Pig. 41— Transformer flux ducing transient conditions. assumed for illustration only. Flux waves for any other initial conditions eam be caleulated in a similar manner using Eq. (74), 39. Determination of Current Inrush Afier the flux variation has been determined by the method described, the current wave can be obtained graphically as shown in Fig. 42, In the ease illustrated it was assumed that a transformer having zero residual flux was energized at zero supply voltage; the Mux there fore is equal to twice normal erest flux. For any ux the corresponling current. ean le obtained from the trans- former B-H curve. Although the maximum fux is oaly twiee its normat value, the current reaches a value equal to many times the maximum value of the normal trans former exciting current. This high value of current is reached beewuso of tho bigh dogroe of aaturation of the transformer magnetie civoait In the above discussion loss hes been neglected in order iplify she problem. Loss is important in an actual or because it deortases the maximum inrush cur- rent and reduces the exciting eurrent to normal after a1s Power Transjormers and Reactors Chapter 5 Nave anaent oe ru ware ayo Pu Fax Fig. 42—Graphical method of determining inrush current. period of time. The losses that are offective are the resist~ ance loss of the supply eireuit and the resistance and stray Jossed in the transformer, Figure 43 is an oscillogram of a typical exciting-eurrent inrush for a single-phase trans- former energized at the zero point on the supply voltage wave? The transient has a rapid decrement during the first few eycles and decays more slovly thereafter. ‘The damping coefficient, 2/1, for this cireuit is not constant because of the variation of the transformer inductance with saturation. During the first few current. peaks, the degree of saturation of the iron is high, making L low. ‘The inductance of the transformer increases as the saturation iy Nae decreases, and hence the damping factor becames mallee as the current decays. 40, Estimating Inrush Currents ‘The calculation of the inrush current to a power trans former requires considerable detailed transformer design information not readily available to the applicati ineer. For this reason reference should he made to the menufecturer in those few eases where a reasonably a2- curate estimate is required. An order of magnitude of ‘Taste 1—Arrnoxtuare Inauset Cunness 10 60-C¥eLs PowER TRASSHORMERS ENERGIZED FROM THE HIGH-VOLTAGE Sie ‘Transformer Cove Shell Racing Kva F 0 ot inrash currents to single-phase, 60-cycle transformers can ‘be obtained from the data in Table 1d, The values given are based on the transformer being energized from the high-voltage side at the instant the supply voltage passes through zero. Energizing a core-form transformer from the low-voltage side may result in inrush currrents ap- proaching twice the values in the table. The per unit inrush current toa shell-form transformer is approximately the same on the high- and low-voltage sides. ‘The inrush eurrents in Table 14 are based on enorgizing fa transformer from a zeronteactance souree. When it it esized to give some weight to source reactance, the inrush ‘current ray be estimated from the relation I ieny I=. or where Io=Tnrush current neglecting supply reactance in per unit of rated transformer eurrent. Xm Effective supply reactance in per unit on the transformer kva base. ‘XV. THIRD-HARMONIC COMPONENT OF EXCITING CURRENT 41. Suppression of the Third-Harmonic Component As discussed in connection with Fig. 39, the exciting current of a transformer contains appreciable harmonicChapter 5. current. ‘The third harmonic is by far the largest harmonic component, being as high as 40 to 50 percent of the ‘equivalent sine-wave exciting current. If the fux in a transformer magnetic cireuit is sinusoidal, the exciting current must eontain a third-harmonic eo: ponent. If this component cannot flow, because of trans former or system eonncetions, the flux will contains third- harmonic component. The third-harmonic flux will, in turn, induce a third-harmonic voltage in the transformer windings. The magnitude of the third-harmonic vollage induced in a transformer winding, when the third-harmonic current is suppressed, will vary between § and 50 percent depending upon the type of transformers usec. With single phase transformers or with three-phase shell-form trans- formers the third-harmonic voltages may be as high as 50 pereent of the fundamental-frequency voltage. In a three phase core-form transformer the reluctanee of the third- ‘harmonic flux path is high (see Sec. 36); consequently the third-harmonic flux in the transformer magnetic circuit is, ‘small even ifthe third-harmonic component of the exciting current is suppressed. The third-harmonic voltage induced is therefore emall, usually not moro than five percent. Ina three-phase system, the third-harmonic currents of ‘each phase a7e in phase with each other and hence con- stitute a zero-sequence set uf currents of Lriple frequency Likewise, the third-harmonie voltages will constitute « ero-sequence set of voltages of triple frequeney. Thus, although a third-harmonic voltage may be present in the Tine-to-neutral voltages, there ean be no third-harmonic component in the line-to-line voltage. ‘The paths permit- ting the flow of third-harmonic currenta are determined by the aystem and transformer ero-sequence ciretita It has been shown that third harmonies must oceur in cithor the exciting current or the voltage of a transformer. “The exciting current will take the shape imposed by the particular connections weed. It is always preferable to have at least one delta-connected winding in a thive- phase transformer bank, The delta connection will furnish ‘8 path for the flow of third-harmonic curvents and will minimize the third-harmonie current in the external eir- cuits. This is very desirable because third-hurmonie eur rents in the external circuits may, under some conditions, ‘eause telephone interference. A discussion of telephone Power source TRANSFORMER, TRANSMISSION LINE Power Transformers and Reactors 129 interferonee, as affected by transformer connoctions, is given in Chapter 23, See. 11, 42. Effect of Transformer Connections ‘The application of the above principles will be illustrated by consideration of a number of typical connections. Ta Fig. 4 is shown a three-phase transformer bank connected, ‘TanLst 15—InPLUBNCE OP TRANSFORMER CORNECTIONS ON ‘Tinmo-Haltmowse VoLtaGRS AND CORRENTE @ Fig. 44 Connections which influence the flow of third-har- ‘monte exelting current. wom [EEREEISTT comers ' Al} AL [see nore 1 2 4+ Ader 2 A oe 8 Stiowococe | * | + | Generators on crounveo| A | A | « « 3 ° ALA]e ns r AyAfte rs ° AlA}« + 38 ° AU | AC | see nore 1 0 SlAtere o a ne 18) "| cee Seoeecnas” | | # fT 13] Bans oR Lance capaci- | 2 | A ja « 8 . alalees E alale«s 6 Old fu. tena id doves130 to a transmission line, the tine in turn being connected toa poner souree. If the star-star eonneetion in Fig. Ha) is used the third-hermonie component of the exciting eur- rent fo suppressed nd a third-harmonic component will ‘thorefore be present in the line-to-neutral voltages. With the primary neutral and the generator neutral grounded, asin Fig. 44(b), a path is furnished for the third-harmonie exciting currents, If the impedanco of this path is low, litle third-harmonie voltage willbe present on the system. However, if the fine is long and ig closely coupled with telephone circuits, telephone interference may result. Tf the transformer bank is close to the power source no tele= phone interference should result from the use of this connection. ‘When a delta-connected winding is present inthe trans former such as in Fig 44(0) and (a), the deltx conneetion furoishes a path for the thind-barmonie currents required to climinate the thirdcbarmonie voltages. If the primary is ungrounded or the generator is ungrounded, no third- hharmonie current will flow in the line. If the primary is grounded and the generator is also grounded, little third- harmonic current can fow over the line. With this eonnes- tion the magnitude of the third-harmonic current in the Tine depends upon the relative impedances of the supply cireuip and the delta cireuit. This current is usually to small'to cause any troublesome interference The same general comments appiv when three-winding transformers are used. If one winding is delta connected Tittle or no third-harmonic current vill Sow in she supply eireuit and little or no thirtsharonie voltage will be Prevent on the eystem. In Table 15 is given a summary of a number of typical transformer connections with a brief deseription of the elect of the connections on the third-harmonic currents and voltages. XVI. TRANSFORMER NOISE ‘Transformer noise is a problem beeause ofits disturbing effect upon people, Noise may arise from several sources. of foree induced vibrations, including, (1) Magnetostriction, the small change in dimensions of ferromagnetic materials caused by induction. (2) Magnetic forces tending to pull jointed core mem~ bers together. (3) Magnetic forces acting between two conductors, __ or betwoen a conductor and a magnetic member. (4) Fans, pumps, or other transformer auxiliaries. ‘The most persistent of these sources of noise is magneton striction, which depends upon flux density and cannot be eliminated by tight core construction. The only means of reducing magnetostrictive force now at hand is to reduce ux density in the core. Noise arising irom any of the sources listed above may be amplified by mechanical resonance in the tank or fite tings, and careful design is necessary to avoid such re- inforeement of the original sound. Standards" have been established for permissible sound pressure levels for various types of transformers, in terms, ‘of decibels referred to 0.002 dynes por square centimeter: Power Transformers and Reactors Chapter 5 Pp. 7 db=20 loses gag 78) where P, the sound pressure, is expressed in dynes per square centimeter. Transformers designed to have sound levels below standard levels are available, but at extra cost, because the magnetie material is worked at an Induction below norma. It is quite difficult to predetermine @ sound level which will prove satisfactory in the surroundings where a new transformer isto be installed. Loeal conditions affect sound ‘transmission, reflection, and resonance to a great degree, and these factors are hard to evaluate prior to transformer XVII. PARALLEL OPERATION OF TRANSFORMERS 43. Single-Phase Transformers ‘Transformers having different kva ratings may operate in parallel, with load division such that each transformer carries its proportionate share of the total load. ‘To uebieve accurate load diviaion, it is necessary that the transformers be wound with the same turns ratio, and that the percent impedance of all transformers be equal, when each per- centage is expressed on the kva base ofits respective trams- former. It is also necessary that the ratio of resistance 10 reactance in all transformers be equal, though most power this respect ‘the impedance transformers will lixely be similar enough to permit calculations based on onl magninute, The division of curtent between tranoformers having ‘unequal tums ratios and unequal peccont impedances may be calculated from an equivalent circuit similar to the one shown in Fig. 45. Eithor percent impedances or ohmic Wig. 45-—Equientene circuit for parallel connection of single- ‘hase two-winding transformers. impedances may be used in an equivalent eireuit for paral- leled transformers. The circuit in Fig. 45 contains ohmie impedances and aetual turns ratios; this method is perhaps more appropriate when the cireuit involves unequal tur ratios, because the use of percent values in this type of cirouit involves extra complieations, Solution of this eic= cuit, with « load current [ assumed, will indicate the division of current betsceen transformers. Also, solution of this circuit with total load current: set. equal t0 zero will indicate the circulating current caused by unequal trans- former ratios, For satistactory operation the circulatingChapter 5 ceurrent for any combination of ratios and impedances probably should not oxeoed ten porcent of the full-losd rated current of the smaller unit “More than two transformers may of course be paralleled, snd the division of load may be calculated from an ox tended equivalent eireuit similar to the one in Fig. 45, 44. Three-Phase Transformer Banks ‘The same considerations apply for the parallel operation of symmetrleal three-phase transformer hanks as have been outlined for single-phase transformers. Tn addition it is necessury to make sure that polarity and phase-shift be- tween high-voltage and low-voltage terminals are similar for the parallel units. A single-phase equivalent circuit ‘may be set up on a line-to-neutral basis to represent one phase of a balanced throo-phase bank, using the theory of aymmetrical components, When three-phase transformer banks having any con- siderable degree of dissymmetry among the three phases are to be analyzed, it is necessary either to set up a com plete three-phase equivalent cireuit, or to interconnect, ‘equivalent soquence networks in x manner to represent the unbalanced portion of the cireuit necording to the rules ‘of symmetrical components. 45. Three-Winding Transformers Currents flowing in the individual windinge of parallel Ueecrinding bate can be determined by solving. an pauivatent eeu, sch as that shown ia Tig. 46. The hts vate t i i ingle Fig. 46—Equivalent clzeult for parallel cor Phase three-winding transfor ‘terminal loads, as well as winding ratios and impedances, affect the division of eurrents among the windings of a. three-winding transformer, so all these factors must be xnown Lefore a solution is attempted, 46. Three-Winding ‘Transformer in Parallel With ‘Two-Winding Transformer ‘The equivalent cirenit for a three-winding transformer paralleled with a two-winding transformer is given in Power Transformers and Reactors Fig. 47—Rquivatent circuit for a single-phase three-stindlng. transformer paralleled with a two-winding unit, Division of currents may be esleulated from this cir curt, if che load currents J and 1," are assimed. Parallel operation of two stich transformers is not uaually. stisfactory, since a change in tertiary load will aller the, distribution of load between the other two windings. If the impedances are proportioned (0 divide the load prop~ erly for one load condition, the load division between trans formers at some other loading is likely to be unsatisfactory. ‘An exception is the case wherein the a” cireuit of Fig. 47 represents a delta tertiary winding in a three-phase bank, ‘with no load connected to the tertiary; in this instance the transformers can be made to divide currents similarly at all loads, It is possible to design a three-winding transformer 60 that the losd taken from the tertiary winding does not seriously afleet lond division hetween the parallcled win ings of the two transformers. Tf the impedance Zp is made eqqial to zero, then current division xt the a’ terminals will be determined hy Ze and Z only, and this impedance ratio will remain independent of tertiary loading. Tt is diffeult tw abinin zero as the value for Zp, particularly if this Winding is of high voltage; however, values near zero can, be obtained with special design at inereased cast. Such a design may result in a value of Zy which is undesirable for other reasons, XVIII. TRANSFORMER PRICES. 47. Two-Winding Type OA Transformers Estimating prices for Type OA, oil-immersed, self-cooled, 60-cyelo, two-winding transformers are given in Fig. 48. ‘The estimating prices per kva are based on net priees as of December 1, 1949. As prices change frequently, the eurves should be used principally for comparing the prices of Uifferent voltage classes, comparing banks made of siugle- phase and three-phase tnits, ete. If the insulation level of the low-voltage winding is 15 ky, or higher, the prices in Fig. 48 should be corrected in actordance with Table 16. Price additions are also required, when the rating of either the high- or low-voltage winding. Ja 1000 volta and below. Transformers designed for star connection of the high- voltage winding may be built with a lower insulation level182 Power Transformers and Reactors Chapter 5 ‘TABLE 16—ADDITIONS 70 BE MADE TO PRICES IN Fic. 48 WHEN Low-VoLTAGE WINDING INSULATION LEVEL 1S 15 KV OR HIGHER Price Addon ia Perera’ Tow Voge Whang | | ‘Single-Phase Equivalent 55 C kva. j ‘Phase Equivalent 55 kya, Yeston elrcadd hatig St Sok aie cas | — — og vt 50110 | 1801 to | 380140 | 7001 to | 13501 and] SOL to | 3601 co | 7001 to | 14001 to) 27001 Fa vad | 300” | “reo0 | “an above | mo) | Foww” | “roo” | "2200 at aoe s way Wy Oe | Oe | om aR, om | om | om 25 , 7 4 ee 7 3 2 ia 38.5 o 7 ce! o | 6 oo] 4 | i eB Bo a fat | eles m | ao | | fo |e la | ul | om wo) ome ee mm | | 2 2 x te | Boe Re | ae ser 250 i 4 a | |e 41 at the neutral end than at the Tine end of the winding. ‘Table 17 summarizes the possible savings in eost with these signs. Reference should be made to seetion 16 for a. discussion of the minimum insulation lovel that should be used at the transformer neutral 48, Multi-Winding Units If a multi-winding ransformer is designed for simul tancous operation of all windings at their rated capacities, the price of Uke unit can be estimated from the curves given for tro-vinding transformers by using an equivalent SGeWESCI CITI anu 17—Paten Reowercn ron GrownorD NEVTEAL Sevice & Prey Wind ae inlingTamistun) Testa Bech — ations | eas 2 t tine Ea Sent Bad 3 S HHH pe ; 4 92 Fe | 30 8 aml BF we ‘ SSS us | se 2 us me) aS 5 = is @ | to BLL 138 8 | 60 06 aS Ho Ey wis $0 ret-puase eae reve mf bee $0 5 Eneee Paste 3 i 1s \ ev T | = 6 6 no ‘ 8 a 53 § tet oe | eek] 18 15-138 } 3 “I 196 | 1 2.0 gen me) ase ts : ioe bo-1s & g be mein | 38 2 230 1s | 00 2 2 asin 7 = 230 92-138 50 2 =| tal 30 ma | aS, Be "eansronwen aaring-ava Pan 33 cates iiss 33 Fig. 48-—Curve for estimating prices of al-immersed, ‘two-winding, type OA power transformers,Chapter 5 twowinding eapacity equal to the sam of the rated ca- pacities of the various windings divided by two, If a multi-winding transformer is not designed for simultaneous operation of all windings at their rated capacities, the price of the unit can be estimated from the curves given for ‘vo-winding transformers, using an equivalent two-winding, capacity equal to Equivalent ATUB-A) 9) Where A= }(Sum of the simultaneous loadings) Bee (Sum of the maximum rated capacitie the various windings). Tn addition, 5 pereent must be added for \ree-winding ‘trunsformers; 7.5 percent for four-winding transformers; and 10 percent for five-ninding transformers, 49, Estimating Prices for Other Types of Cooling, ‘Tablo 18 is a summary of the approximate eost of threo phase power transformers employing auxiliary souling sy5- tems, All cost figuros aro oxprossed in por unit of OA "Tamu 18—RELATIOE Cost oF TaRex-PHAse TRANSroRMnRs ‘win Social, Coonan fs in pr unit, asd on tho cost of at OA transformer ing a rating oat to the maximum of the special unit being eonsid- parc Each cout Tnsulation Clats—KV Type Rating __ vais fo oo | o> | a] 108 101| e250 ‘Oarea | 1 frosh-o7j-oah.08)1.081.091.091-05 1.07 2 {Cog ot ona eat og 091 051 09 5 locsalo-asjo-a)o elo 7)09}0290.991 00 10 fosob.sip.a2).ca sp a9y.ago 960.96 20 p.sso ssp sop s00 a0 920 910 020.0 50 "ist. so. sno.sn0.91).910 990.92 00 jo -sip.ss)o'sH0.900.200 910.91 oasrasros} 20 ).740.79).77).790.700.810.900.s90.53 0 [0 r3p 730 23.770. 900.s10 19.52 100 p.rap.t4p.tlo a t80.810.90 oa | 29 es o7p.zso.zs0.750.770. 190.810 82 ia | loetocoslo. to. 710.130.790.810. 82 200 o-080.079.670.080.710, 790.75 ow | 2 |1os.090.200.o).09 3.07 020. 970.08 5 |o‘sio-o20 92 oala-0s 0,070 990-490 1h ‘so.ao so 980 010.90 ne Ssh ao se 0 09. 2 lo'sso.xs0 so ago. 87 109 0.85)0.850.850.820 820.8510 st fa /" p.s9p.cao.ci apo.090 730.710 72 |e | 00, 610.wB0 620.680.600.60 BAAR AG las Sea ec Power Transformers and Reactors 188 transformer cost, where the OA rating used to determine the base cost is equal to the highest rating of the force- cooled or specially-cooled unit. ‘The kva ratings listed im the second column of Table 18 aro the highest ratings of foreed-cooled units; for example, the kva rating listed for OA/FA/FOA transformers is the FOA value. XIX. REACTORS 50, Application of Current-Limiting Reactors Current-limiting reactors are inductance coils used to limit, current during fault conditions, and to perform this function it is essential that magnetic saturation at high eurrent does not reduce the coil reactance. Tf fault current, is more than about throe times rated full load current, an iron core reactor designed to have essentially constant magnetic permeability proves overly expensive, therefore air core coila having eonstant inductance are generally used for current-limiting applications. A reactor whose induc tance increased with current magnitude would be most effective for limiting fault, eurrent, but this eharacteristi has not been practically attained.ig, $0—Dry-type air-core reactor, Air core reactors are of two general Lypes, oil-immersed (Fig. 49) and dry-type (Fig. 50). Oil-immersed reactors can be cooled by any of the means commonly applied ta power transformers. Dry-type reactors are usually cooled, by natural ventilation buit can also be designed with foreed- alr and heat-exchanger auxiliaries where space is at a pretium. Power Transformers and Reactors Chapter 5 Oil-immersed reactors can be applied to s cireait of any voltage level, or either indoor ur outdoor installation. The advantages of nil-mmersed reaccors also include: 1. A high factor of safety against flashover. 2. No magnetic field outside the tank to cause heating. for magnetic forces in adjacent reactors or metal structures during short~ Bae (82) Zaye NiZet Za) Za%= 5 SZe%+ 20%) also Ze=RetjXe Zea Reg tiXee~RrtRati(XrtXe) 63) Zos%o= RraetiXea%e ee where Xvs is the leakage reactance between the P and S windings (with T open-cireuited); and Res is the total effective resistance between the P and S windings, as ‘measured in obms on the P winding with S short-circuited and T open-circuited. Rost and Nos are the respective quantities expressed in percent on the kva and voltage of the P winding “ The equivalent circuits completely represent the actual transformer as far as leakage impedances, mutual effects between windings, and losses are concemed (except ex citing currents and no load losses). It is possible for one of the three legs of the equivalent circuit to be aero oF negative 58. Equivalent Circuits for Four-Winding Trans- ‘The equivalent cireuit representing four windings on the same core, shown in Fig. 54 (a), is given in Fig. 54 (b) using ohmie quantities, This form is due to Start! and Power Transformers and Reactors Fig, 4--Four-winding transformer, (a) winding diagram, ) equivalent eiruit here again the magnetizing branches are omitted. The branches of the equivalent cireuit are related to the leak age impedances between pairs of windings ws follows: Bun S( Zest bee jalov -K) aren K) a) 1 1 Kae Pret fier Zov~ sian Kes Poet gifer—Zorm yal ‘The windings will ordinarily be taken in the order that makes Ki and K; positive 90 that Z, and Z; will be posi- tive. The leakage impedances are defined as before; for example, Zrs is the leakage impedaneo between the P138 and $ windings as measured in ohms on the P winding with the $ winding short-circuited and with the T and YV windings open-eireuitod. The equivalent eirouit in per ‘cont has the same form a Fig. 54 (b), omitting the ideal transformers v 205-1 teat Devo Pho KG), ee Kilo Zeer B80 Similar equations, derived from Fy. (84), apply for the other quantities in the equivalent eiseuit, Veg. op Tieton ee. (85) XXI. SEQUENCE IMPEDANCE, CHARACTERISTICS OF THREE-PHASE TRANSFORMER BANKS 56. Sequence Equivalent Circuits The impedance of three-phase transformer banks to positive, negative, and roro-soquence currents, and the sequenee equivalent circuits, axe given iu the Appendix, under Equivalent Cireuits for Power and Regulating Transformers. The equivalent cireuits were developed by Hobson and Lewis? The same notation as defined in the early part of this chapter is used to denote leakawe jimpedanees in ohms and in percent. ‘The impedance to negative-sequence currents is always ‘exqual to the impedance to positive soquence currents, snd the equivalent cirouits are similar except that the phase shift, if any is involved, will always be of the sume magni tude for both poeitive- and negative-sequence voltages and currents but in epposite direetions. Thus, if the phase shift is bar degrees for positive-seqence, the phase shit for negative-soquence quantities will be —a degrees. The impedance of a three-phase bank of two-winding transformers to the flow of zero-sequence currents is equal to the positive-sequence impedance for three-phase shell- form units (or for a bank made up of three single-phase units) if the bank ie starslar with both star points ground- ‘| 4 Fig, 85~Zero-sequence exciting currents and fluxes in a three ‘phase core-form transformer, Power Transformers and Reactors Chapter 5 ed. Ifthe bank is connected star-delta, with the star point grounded, the zere-sequence impedance viewed from the Star-connected terminal for shell-form units, or banks of three single-phase units, is equal to the positivesequence Jmpedance; the zero-eequence impedance viewed from the delts-connectod terminals i infinite ‘The impedance to the flow of zero-sequence currents in three-phase eore-form unite is generally lower than the pesitivessequence impedance. Figure 59 ilustrates that there is no return for the zcro-sequence exciting flux in such unit, except io the insulating medium, or in the tank and metallic connections other than the eore. The flux Hinkages with the aero-sequence exciting currents are therefore lov, and the exeiting impedanee to zero-sequence currents correspondingly low. Although the exciting im- pedance a pesitivesoquence currents may he several {ousandl percent, the exciting impeslance to zeresequence euurrents ita three-phase care-form unit wil ie in he range from 80 to 300 pereeat, the higher values applying to the larzest power transformers. Low exeiting impedance tunder zero-sequenee conditions is reflected in somo re- ‘duction in the through impedances to 2ero-sequence cur- ront flow. star-star grounded, three-phase, two-winding tunit of the eore-form, or starstar grounded antoirans- former of the three-phase core form acts, becauso of this characteristic, a if it had a tertiary winding of relatively high reactance, In small core-form units this charactoristio is particularly effective and ean be utilized to replace tertiary winding for neutzal stabilization and third har monic excitation “The aero-sequence exciting impedance is afectal by the insgaitude of excitation voltage, and itis also affected by tank construction. For example, the zerosequence excit- ing impedance of a 4000-kva, 60 000-2400-volt. unit. was measured to be 84 percent at- normal voltage before the core was placed in the tank; it was measured to be 86 pervent al normal voltage alter the core and coils were placed in the tank. In this case the tank saturated but acted as a shortcireuited secondary winding around the transformer, tending to limit the area of the flux rotura path to that between tank and windings. The sero- fequenve exciting impedance is measured by connecting the threo wizlings in parallel and applying 4 single-phaso voltage to tho paralleled windings. “The zero-sequence exciting impedance of three-phase core-form units is generally much lower than the positive- bequence exciting impedance, and roueh lower than the zero-sequenco exciting impedance of three-phase. shell form units or three single-phase units. For this reason it is necessary to cousider the zerosequenee exciting ime pedance in deriving the zero-sequence impedance char fetcrsties for vertain cannections involving eore-form ‘The exciting impealanee to aerosequence currents hiss been denoted by Zse, Zon, ete, where the first sub- seript refers to the winding on which the zero-sequence exciting impedance is measured in ohms. Following the same notation, Zen is the exciting impedance of the S ‘winding to zoro-sequuence currents expressed in percent on the kva of the S winding. The mumber of branches required tocdefine an equivalent cireut of three-phase two- or multi- ‘Winding transformers isthe eame in genoral as has boon do-Chapter 5 seribed for single phase transformers. Anotableexception 19 this will exist in the formulation of the zero-sequenee im- pedanee of core form transformers with grounded neutral Tn this case an extra impedance branch must be provide the equivalent circuit, this branch being always short cireuited to the neutral bus, and having a valise dependent upon the zero-sequence excitation impedances of the wind- ings as well as the grounding impedance in the transformer neutral. If the three-phase bank connections are unsyin~ ‘metrical asin the case of the open-delta connection, mutual coupling will exist between the sequence networks. 57. Derivation of Equivalent Circuits In the derivation of equivalent circuits for three-phase transformers and banks made up of three single-phase transformers, it is convenient to represent each winding of, te transformer hy a leakege impedance and one winding ‘of an ideal transformer, ‘This method may be used in the development of circuits for two- and three-winding ‘transformers. Two magnetically-coupled windings of a single-phase transformer having n; and n turns, respectively, are shown, schematically in Fig. 36(a). ‘The customary equivalent civeutt used to reprosent euch a single-phase transformer is, shown in Fig. 56(b) in which Z, and Zs are components of the transformer leakage impedance, with a more or less, arbitrary division of the leakage impedance between Za and Za. Za is the so-called “magnetizing shunt branch.” Sineo the numerieal value of Zy is very large compared to Zs and Za, for most caleulations Fig. 5615} is appro’ imated by Fig, 58(c) where Zy is considered infinite Either of these circuits has serious dei eview representing the actual transformer; the voltage and cu: rent (ransformuition effected by Cransformuer action is not represented in the equivalent cireuit. and the eireuit terminals a and a’ are not insulated from each other as in the actual transformer. These disadvantages are evidenced, particularly when analyzing transformer cireuits wherein, several windings or phases are interconnected. To aver- come these deficiencies itis expedient to use the equivalent cirenit shown in Fig. 36(d) which combines the circuit of Fig. 56(b) with an ideal transformer. The ideal trans- former is defined as having infinite exeiting impedance (zero exeiting current) and zero leakage impedance, and serves to transform voltage and current without imped- tance drop or power loss; the ideal transformer thus re- stores actual voltage and eurrent relationships at the terminals @ and a’. ‘The eireuit of Fig. abe) is obtained from Fig. 56(d) by converting the impedance Zs to the Ey’ voltage base (by multiplying Zy by the squar of the voltage ratio). ‘This process may be thought of as “sliding the ideal transformer through” the impedance Zp. If che exciting, or no load, current may be neglected (Zw con- sidered ‘as infinite)’ the circuit of Fig. Sie) becomes Fig. 5641) Finally, if Zy is considered infinite, the clreuit of Fig 56(0) becomes Fig, 50(q), in which the two parts of the leakage impedanee, 2, and Zp, combine into the complete leakage impedance Zps, where eam Dhl (80) Power Transformers and Reactors 29 @ Fig, Stops in the derivation of the equiva ‘ewonwinding transformer.M40 In most developments the circuit of Fig. 58(g) will be found most convenient, although in some eases it heeomes desirable to have part of the leakage impedance associated. with each winding, and the eixeuit of Fig, 56(f) may be used. ‘To be perfectly definite, Zee is understood to moan the leakage impedanee, as messuzed in ohms, with the $ wind- ing short cireuited, and voltage applied to the P winding, ‘When the test is reversed, with voltage applied to the ‘winding, and the P winding short eircuited, the impedance ie denoted by Zyp. Tt is obvious from the development given that, when Z may he considered infinite, en 58. Derivation of Equivalent Circuit for Star-Delta Bank In Fig. 87 each transformer winding is represented by an impedance and one winding of an ideal transformer, the ‘transformer having ms turns in the P winding and ne turns in the S winding. ‘The windings shown in parallel are assumed to be on the same magnetic core. ‘The voltages te MT (a) sonemaric vinonam tome? : oS (0) POSITIVE- SEQUENCE, EOUNALENT CIRCUIT ] tf (@)NEOATIVE SEQUENCE EQUIVALENT GRGUIT 4 (@ZERO-SEQUENGE EOUIVALENT GiRGUIT Fig. 57Equivalene circuits ofa star-delta transformer bank Power Transformers and Reactors Chapter 6 gp and es represent the voltages across the P and § wind- ings of the ideal transformers. "Assuming positive-sequence voltages Eye, Eig) and Boe applied to the terminals abc, and a three-phase short circuit at thea’d’e’ erminals, the following relations can be written: 488) Designating the creuits connected tothe abe and ae terminus as Greuts 4 and 5, respectively, E im", 7 rt |) Ze=Zee ioe (R) 2 is defined as the impedance between circuits 4 and 5 in-alinaon the circult 4 voltage base, Zee isthe imped nce between the Pand 8 windings as measured by apply sng voltage to he winding with the 8 winding shor ead With. postivesequence voltages applied to the ate terminals end the a erminals open eruite, Ben t'Bag EyneEly Ennab Baal, Le (89) 90) ME —0) = VEE 8 re Pt, Bi Vm As postive-sequence quantities were used in this analy- sis, the final equation ean be expressed as follows Lotting N NE ge. B= NEE, en where Ey and Ey are the positive-sequence voltages to ground at che transformer terminals. ‘The above relations show that the Hine-to-ground volt~ ages on the delta side lead the corresponding starside voltages by 30 degrees, which must be considered in a Complete postive-sequence equivalent cieuitfor the trans- former. A consideration of Eas. (88) will show that the currenial’, Lf and Zé alzo lead the eurrents Ty, fy and Ta by 30 degrees. nen Moot, m (92) eg Leon, To Bem,‘Chapter 5 The complete positivessequence circuit in Fig. 57(b) therefore includes the impedance Zys and an ideal tr former having a tras ratio V and a 30-degree phase shift, ‘A similar auslysis, made with negative-sequence volt- ages and currents, would show that bem (93) 4) ‘The positives and negative-sequence cireuits are there- fore identical excepting for the direction of the phase shifts introduced by the star-delta transformation. ‘The zero-sequence circuit is derived by applying a set of ByaNEO™. sero-sequence Voltages to the abe terminals. In this ease Exg=Bug= Beg= Ey ly = lo Exe eo+Zelu e4—1eZis=0 because no zero-sequence voltage ean bbe present between line terminals. (5) G ras a St) ates Jpn z, a1f( ‘was obtained with positive-sequence voltages and currents, (06) If zero-sequence voltages are applied to the a’B'c’ ter minals, no current can flow because no return cizcuit is present. ‘The zero-sequence impedance of the transformer bank is therefore infinite as viewed from the delta side. res, which is the same impedance as 59. Derivation of Equivalent Circuit for Aurotrans- former with Delta Tertiary ‘The basic impedances of an autotransformer with «delta tertiary may be defined in terme of tho leakage impedances between pairs of windings, with the third winding open cireuited. The impedance between the primary and secondary, or common and series, windings of the trans- former in Fig. 58(a) may be obtained by applying a voltage across the P winding with the S winding short circuited, fod the T winding open cizeuited.”Rafersing to Fig. 50, see(S)iz ntl Baeytl2e =1(Z420) 7) ton Bita 28.2 Similar relations con be derived for the impedances Power Transformers and Reactors Sf tanta Fed (0) POsiTivE-SeOUENCE EQUIVALENT CIRCUIT Zu iw ig Ve w, i ey Ry twee (6) NEGATIVE-SEQUENGE EQUIVALENT cincUIT Ze ew (#) ZERO-SEQUENCE EQUIVALENT CIRCUIT Fig. S8—Equivatene circuits of a three-winding aucotrane former,ue Power Transformers and Reactors Chapter 5 (101) Representing the circuit transformation ratio (L-+m) by tax (Not) tna (aoa) N ‘The impedance betwen, crits 4 and 6 may be ob: tained by applying positive-sequence voltages to ter- ig. $9—Representation of the primary- to secondary-wad= mils be, with terminals 26? open and "B=" shorted ing inpedance of ak autotransforaer- Girouited, La this ease: eahee iw l./ra between the P and 7, and $ and 7 windings, resulting in he, i the act of equntions atts the aot of en on to a0) Baemept labe a1 [Zea] 08) With pesitive-sequence voltages applied to terminals ‘These equations can be solved for the individual winding @¥', terminals ate open and terminals a”B"C" short impedances Ze, Zs and Zy. cireuited, th ‘oe =hls getty Lam WZortmi Zeon * ean EB (PV Tey aon am WZortntZre—miZer] (0) eben ve tel Gyan ma Bentgt ee Hide d ™ 1 bee mn rived in toms ofthe impedances between windings, eng the same procedure as employed in the derivation of the impedances of the stardelta bank in section 58, With positive-sequence voltages applied to terminals Expressing Z>, Zs and Z in terms of impedances be- ‘abe, terminals a'b’e’ short circuited and terminals abc” tween windings as given in Eq. (99) ‘open cireuited, the following relations ean be written: isNoart Bien Bly Big0 te tum (emer +()zer—mtes, 108) H-Tetmti=0 ‘The above equation isthe impedance between cireits Band on chs on the cieut 5 voltage bee. An Za and cote (ld Ilo Lond. Za ate chase nnpeaces onthe eteut 4 base, 3 is cone ‘venient to express the circuit 5 to circuit 6 impedance on Eliminating ¢, and I’, from the above equation: the same base. Dividing by (I++m)*, Ie Zs eo(L+m) =" et Zs) —TaZe “ (Lm) = 5 et Za) — Tals a Lew (100) weChapter 5 ‘The transformer can be represented by the positive: sequeneo equivalent circuit in Fig 58(b). The relations Ddetween the impedances in the equivalent circuit and the impedances between cirouits ean be expressed as follows: Lut Zina Zu Bart Zia Zon %, Bart Zum ign (108) cao) an In the above equations Zim, Zn Zus, Za and Zy are in ‘ohms on the circuit 4 (abe terminals) voltage baae. Zu is fin ohms on the circuit 5 (a''e’ terminals) voltage base. Zes and Zpz ate in ohms on the P winding voltage base ‘and Zsr is in ohms on the S winding voltage base. ’ is defined as 1++n,, which is the ratio of fine-to-line or line- tomeutral voltages between circuit 5 (a’¥'e’ terminals) and cireuit 4 {abe terminals) ‘The phase shifts between cirouit voltages can be de- termined by applying positive-sequence voltages to ter- minals abe with the other two circuits open eireuited, Under these eonditions, une Eye (1-4) Bur= N’Bxa, Which shows that the one {deal transformer has au 1’ ratio but no phase shite eo Bag! ~ Ba! = Bella) eamaty m8, Eadie Bg BB gem 2 Tater Tae aia) Defining "as", 5 Eu! N' Buge®, Power Transformers and Reactors M3. ‘The second ideal transformer therefore has an N"" turns ratio and a 30 degree phase shift Negative-Sequence Circuit—S similar analysis made with negative-sequence voltages would show that the impedances in the equivalent circuit are the same as in the positive-seyuence circuit, and that the terminal volt ages are related as follows: Han" Bila N EA, (13) ‘The positive- and nogative-sequence cireuite are there- fore identical excepting for the direction of the phase shift introduced by the star-delta transformation, Zero-sequence circuit—The zero-sequence characte istics of the transformer ean be obtained as follows: 1. Apply zero-sequence voltages to terminals abe with terminals a'¥e' connested to ground and the delta opened ‘This permits evaluation of the zeru-sequence impedance between circuit Land circuit 5. 2. Apply zeto-sequence voltages to terminals abe with, the delta closed and terminals a'b'e’ open circuited. 3. Apply zero-sequence voltages to terminals wc’ with, the delta closed and terminals abe open cixeuited. ‘The general procedure in writing the necessary equa tious is similar 20 that followed in the positivense analysis given above, and the zero-sequence analysis in seetion 87. It will be found that the zero-sequence im- pedanees in the equivalent circuit shown in Fig. 58(d) are the same as the positive-sequence quantities, that is, Zn Zus Zuo (us) ZonZu If the neutral of the autotransformer is ungrounded, the zero-sequence equivalent circuit is altered considerably as shown in Fig. 60. In this case rero-sequence current flows OE Fig. 69—Zero-sequence equivalent circuit of an ongrounded ‘hree-winding autotransformer. between terminals abe ard a’b'e’ without transformation, urrent in the S winding is balanced by circulating eur rents in the tertiary, with no current, low in the P winding ‘The zero-sequence impedance is therefore determined by the leakage impedance between the S and 1’ winding Applying zero sequence voltages to the abe terminsls, with the @''’ terminals connected to ground and the tertiary closed, 12xua Power Transformers and Reactors Chapter 5 EugeT2a~ es sidering the autotransformer with delta tertiary (case D-1 ss in Table 1), the equivalent circuit impedances can be =n [% (i Na +] ‘obtained from the impedances between cirouite as follows: ieee) coal tent] ais Percent Quantities Tie manufacturer normally exe raSontl 0 Wa 26 a presses transformer impedances in percent on a kva base Wher iy % reporting to the rte ve of the svc involved ‘ape ahs son be converted om by th om eaten zat] 8) ‘The reaulting impedances will all be in percont on the cirenit 4 kva base ais) REFERENCES 1. Elecrie Cieoulte—Theory and Application by 0. G. C. Dab (book) Val 1, p34, MoGraw-Hhill Rook Company, Tae, New mpedanee in percent, San kva=3-phase kva rating of eireuit. 2 Regulating Transformers ia PowerSstom Analysis, by J. B. Hubsou aud WA. Losis, A.LELP. Transactions, Vo, $8, 1998, E=line-to-line cireuit voltage in ky. ne ‘Using the nomenclature employed in the derivations, 8. Fundamental Concepts of Synchronout Machine Renetances, by B. Re Prontice, AL-E. ¥. Transaction, Vol. 6, 1031, pp. Lox lOEu%e BP org 22of Supplement Te 4. Simplified Computation of Voltoge Regulation with Four Wind Eq~line-torine voltage, in kv, of circuit 4 Sap Toners, by RD. Evang, Buccal Breen T Osta 193 Uyethreephase kva rating of eiruit 4 5. Surge Prot nsformen, by HV. Potnen, ALB, Tron Zao mimpedance between cireuits Land 5in percent on etins, Septembor 1352 jp. SIHS8K and dheosion, pp. 8a va rating of eieuit 4 ‘oo ; 5 6. Rnwscan Standards for ‘Transformers, Regulators, and Ree Za=inpedance petween eteuits 4 and 5 in ohms on” eicm, Amerian Standards Asvosiation, ABA C3, {O18 the circuit 4 voltage base. 7. Loading Transformers by Copper Temperatuce, by HV. Put Similar relations can be written for the other impedances man and W, M, Dann 4.4.8.2, Transactions, Vol. $8, 1039, involved. PP. 903-509. jivalont Cheuit tmpelanee of Regulating Transformers, by JE. Clem, ABE, Pransnctins, Vol. 88, 1888, pp. 871-873, 9 Thoory of Abnormal Line o Neutral Transformer Voltages, by Tt should be noted that the impedances, as used in this chapter and in the Appendix, are expressed in terms of the fettage ova ring he seal ating Sosa he A Neal Terr Yan Ueliteabseipt Forctample tn ctor uaese, © Nlafies) AEE Dranertn 3, Mah tuts wllage Pic, Wheteas Za ull i ainsom the a0. Notion tasoor SBMA Pease No 4818 dik vltnge be "hoe epotaneossan bemasmaet Smt 2 feet, Gees Soe in el tos (117) Eauivaone Ciruits for Power and Regulating Transformers, by JOB. Mobson and W. A Lewis, Bledric Journal Preprint, i ss fs January Ua ‘The equivalent circuits can be hased directly on percent 4, J, and P, Transformer Book, by Stigant, Oth Edition, 1995, quantities as shown in Table 7 of the Appendix. Con- "Johnson and Phillis, London,CHAPTER 6 MACHINE CHARACTERISTICS Original Author c. Wagner present enormous proportions with large generating, stations and connecting tie lines, machine perform ance was largely judged in terms of the steady-state ehar- acteristics. ‘The emergence of the stabilicy problem gave rise to the analysis of the transient characteristies of machines and was largely responsible for our present knowledge of machine theory. A further contributing urge was the nead for more accurate determination of short= ireuit currents for the application of relays and circuit breakers, The v salient-pol Bee the growth of the public utilities into their nhle character of the air gap of the conventional synchronous generator, inotor, and condenser ‘with its concentrated field windings requires that their analysis follow a different line from that for machines such : which have a uniform air gap and attacked this ag inductivin ‘moter distributed windings. Blondel originally problem by resoiving the armature mm ‘two components, one in line with the axis of che poles and the other in quadrature thereto. When the si ‘tansients associated with system stabil ig. 1 Cut-away view of umbrella Senerator. Revised by. C.F, Wagner 1 Fig. 2—Cut-away view of contentional waterwheel generator, this conception was quickly recognized as an invaluable tool, Since that time the method has been extended by subsequent investigators notably Dohercy and Nickle, who introdueed into the industry several nev constants, such as transient reactance and eubtransient reactancs describe machine performance under transient conditions. ‘This chapter treats of the charsctoristis of synehronous and induction machines in the light of the development af the past twenty-five years, It will consider steady-state and transient conditions for both salient pole and eylin- drical rotor machines under both balanced and anbalanved conditions. There follows 9 discussion of the charactes- isties of induetion motors under auch transient conditions as might contrilaite co the short-eireuit current of a system ‘and might influence the choiee of a eireuit breaker. I, STEADY-STATE CHARACTERISTICS OF SYNCHRONOUS MACHINES The two general types of synchronous machines are the cylindrical rotor machine or turbine generator which has ‘an essentially uniform air gap and the salient-pole gener- ator. Figs. 1 to 3 illustrate the outward appearances and ‘cross-sectional views of typical modern machines. ‘Typical saturation curves for a hydrogen-cooled turbine generator, a watorwheel-generator and a synchronous eon: denser aro shown in Figs. 8, 7, and 8 respectively. Because of the necessity of matching the speod of water- wheel-generators to the requirements of the waterwheels it is didicult to standardize units of this type. However, 15Fig, 3 Steam turbine generator Installed at the Acme Sta. tion of the Toledo Edison. Company, 96 000 kw, 88.percer power factor, SB-percent SCR, 18 800 volt, S-phase, 60-cye! sgront strides have been mace with large 8600-rpin_ con densing steam (urbine-generators. ‘These find their groat- est applieation in the eleetvie utility industry. ‘Table 1 of Chap speeifications®* for thew machines, Tho concept of per-unit quanti ing the charaeteristies of machines of difforent. capacities and voltages. However, care must be exercised in the ease ff generators to use the sume referenee vulue for Held eur 1 gives some of the is valunble in compar Fig. 4 Cut-away view of hydro ‘ten-cooled turbine generator Machine Characteristics Chapter 6 rent. Depending upon the application, cither the field current for rated voltage in the air gap or the actual field eurrent for rated voltage, including saturation, is used. Unsaturated Cylindrical-Rotor Machine Under Steady-State Conditions ‘The veetor diagram of Fig. 9 is the well-known disgram, of a cylindrical-rator machine. Consistent with the policy of this hook, familiarity with this dingrara ia assumed. Let it sufi to indicate the significance of the quan- tities, The vectors e, and # represont the terminal voltage to neutral and armature eurrent, respeetively. Upon add- ing he scumtuey resistance drop, rv, and armature leakage ‘ance drop, 216 to ée the vector ey is abtained, swhich represents the voltage developed hy the airgap lux ®, which leas ex by 90 degrees. This flux represents the net. fluc in the air gap. ‘To produce this fus a field current, Tx ‘required. The eurrent /, can be taken from the no-loadut Chapter 6 ‘Machine Characteristics . T T ' T 1a] t 4 + ; j SST Le | fia val — AL ete \ 1 i i yO a z| s| FP) Te So Ea —o/ | £ § ef | | gy of | | | | | 3 i | | I 11s 03830 as 40 a FIELD AMPERES-PER UNIT Fig. (Saturation curves for typical hydrogen-cooled turbine PER UNIT AG VOLTAGE 3525 FIELD AMPERES-PER UNIT Fig. 7—Saturation curves for typical waterwheel generator. saturation curve of Fig. 10 as being the current required to produce. But, the armature current produces an mmf by ite so-called armature reaction, which js in time phase With it and in terms of the field can be expressed as Af. To produce the net mmf represented by the current, fy, the field current must be of stich magnitude and the Beld structure must adjust itself to such position as to equal Is In other words, Z; has now auch position and magnitude that [rand Av added in veetorial sense equals f. The triangle 0.48, formed by drawing AB perpendicular to ior Ai and OB perpendicular to OC, is similar to the triangle ODC; OB has the same proportionality to OC and AB to Av as e; has tof. Neglecting saturation, OB, designated 445 «1, is thus the open-eireuit Voltage corresponding to the PER UNIT FIELD AMPERES Fig. 8 Saturation curves for typical hydrogen-cooled con- denser. field curront 1); i: i the voltage taken from the air-gap Tino of the no-load saturation eurve for the abscissa vorre- sponding t0 f,. The side 4B of the triangle, since it is pro- portional ro Lv and consequently proportional to thearma- re current. can be viewed as a fletitious renctance drop. Te ig called the drop of armature reactance and is desig- nated yi, The roaetance drops ri and xf en he com= bined into a single term called the synchronous revetanee drop and there results, zany @ It follows from the foregoing that the internal voltage, s, is equal to the vertor sum of m,n and j rai. The Geld current, Js, can be determined for any condition of loading (neglecting saturation. of course) by merely caleulating e;and taking Ir from the air-gap fine of Fig. 10. At nofoad the axis of the field winding, the line OC, leads | the terminal voltage by 90 degrees. At zero powerfactor, the vector diagram reduces to that shown in Fig. 11, which shows that, except for the effect of the resistance drop, the foregoing statement woul still be true. Ast is only about fone of two percent in practical machines, the statement Pig. 9 Vector diagram of cylindrical-rotor machine,us uo toa ‘SATURATION CURVE rare zERo POWER-FACTOR SATURATION CURVE. te Pre Need a os FELD CoRRENT oud ero power-factor character- ‘2 generator. i Fig. 11—Vector diagram of eylindrical-rotor generator at zero ‘power-facto can be accepted as true for all practical purposes. How= er, as the real load is applied to the machine the angle 3 {nereases from zero and the lead of OC ahead of es increases, from 90 degrees to 90 degrees plus 8. ‘The angle 8 is a real angle; it can be measured without much difficulty. Ivis convenient for some purposes to resolve the reac- tions within tho machine into two components, one along, the axis of the field winding and the other in quadrature thereto, Tn Fig. 9, the armature eurrent is divided into the two components, fy, and, iq in whieh the subseripts are significant of their respective components. When this is Machine Characteristics Chapter 6 done, it ean be seen that zi ean likewise be thought of as arising from the two components of im the form of tyaia and sig, respectively, in leading quadrature to ¥g and tg In the exee of a cylindrical rotor machine, tag and rq both equal to 2, but ease will soon be developed for which they are not equal ‘The synchronous renctance, za, can be obtained most conveniently from the no-load eurve and the full-lowd zero powerfastor curve. In Fig. 10 OA is the field current, required to circulate fullload current under short-circuit, conditions, the terminal voltage being zero. Tn this ease all of the internal voltage (the ri drop ean be neglected, justifiably) must be consumed 9s synchronous reactance, ‘rop (rg!) within the machine. If there were no saturation, ‘the internal voltage ean be determined by simply reading. ‘the terminal voltage when the short-oirouit is removed, sintaining the field current constant meanwhile. ‘This voltage would in Fig. 10 be equal to AB. Thus the une sattated syuchronous reaetauee per phuse fy equal to th phase-to-noutral voltage AB divided by the rated current ‘When the saturation eurve is expressed in per unit or per cont it is equal to AB; but where expressed in generator terminal voltage and field amperes, ts eal to (100) sn pen or! in pre oH 2. Unsaturated Salient-Pole Machine UnderSteady- State Conditions Given the proper constants, the performance of an un- saturated salient-pole machine at zero power-faetor is the same us for a uniform airgap machine. For other power- factors, conditions are different. veotor diagram for suck machines is shown in Fig. 12, As before e and # are the in per unit. Pig. 12—Vector diagram of salient-pole machine, terminal voltage to neutral and the armature current, respectively, and ey is the “voltage behind the leakage eactunce drop.” ‘The flux is required to produce e.. ‘This flux ean be resolved into (wo components bz and ®,.. The flux $4 is produced by J; and iy, the direet-axis eom- ponent of Ai, and @y is produced by Ai, the quadrature- axis component of At. Here the similarity ceases. Because ‘of the saliency effect, the proportionality between the mmf’s and their resultant fluxes is not the saroe in the two, axes, When saturation effects are neglected ‘by can be ro- garded as made up of a component produced by 7 acting,Chapter 6 FUNDAMENTAL COMPONENT TSEC TP Py Fig. 18 Flu resulting feom a slousoidal mmf in (2) dine ax () quadsatare axis alone and a component produced by is. ‘The component produced by Tr can be regarded as producing the internal voltage es. The mmf produced by 1, has a general sina- idal distribution in the direct axis as shown by Fig, 13() ‘The resultant flux heeause of the variable reluctance of the airgap has the general shape indicated. Ttis the sinusoidal component of this flux that is effective in producing the Yas ig drop shown in Fig. 12. Ta the quadrature axis, the ‘component. of mon ix ikewise sinus in nature asskown in Fig. 13b, and gives rise to the distorted flux form. In proportion to the mm the sinusoidal component of fx is much less than for the direct ass, The effect of this com= ponent is reflected in the zy i drop of Fig. 12. Tn general a; i8 tbh soaller than 2y. ‘The armature resistance and leakage reactance drops can also be resolved into its two components in the two axes much af zi of Fig. 9 was resolved. When this is done the internal voltage ee can be obtained by merely adding and rig and then j iy and ja to the terminal voltage ec. ‘The notation 4 is used to differentiate the internal voltage in thie development from that used with the cylindrical rotor machine theory Another form of the veetor diagram of the machine is presented in Fig. 14, which shows much better the relation bevwveen those quantities thet are most useful for ealeula- tion purposes. IF rom B the line BP of length z4iis drawn perpendicular to, then sinee angle CBP is equal to 68, the distance BC is equal to aif eos (9-3), oF aig. BY comparing this Eine with the corresponcting line in Fig, 12, ican be seen that the point P determines the angle 8. This relation provides an easy construction for the determing tion of the angle § having given the terminal voltage, the amature current, end the power-factor angle, . Further, Fig. 1¢—Determination of Internal angle, 5, and excitation of fan unsaturated salient-pole machine when loading is knows. Machine Characteristics 49 the projection of BF upon OG is equal to zufu so that OC becomes equal to ¢y, the fictitious internal voltage, which is proportional to fy The armature resistance is usually negligible in deter- mining either the angle 8 or the excitation and for this ease 4 sin 5=Eela™ tgi cos(6-+8) @ Upon expancling the last term and solving for 5 @ From Fig. 14, the internal voltage eam ey con Shzui sin (G43) @ ‘The unsaturated synchronous reactance, 2, ean be de- termined from the no-load and full-loud aero power-factor ‘eurvoa just 43 for the machine with uniform air gap, The quadrature-axis synchronous reactance is not obisined so FOR SIGNIFICANCE OF QUANTITIES IN PARENTHESES REFER TO FIGIB Fig. 15~Determination of internal angle, s, and excttation fof'a saturated salient pole machine when losding is known. easily but fortunately there is not as much need for this quantity. It can be determined from a test involving the determination of the angular displacement of the rotor as real load is applied to tie machine and the use of Eq, (2), which gives evsin 6 "Tons (+0) 8 or it can be determined by means of a slip test. The slip testis described in the A.LLE.E, Test Code for Synchronous Machines! of 1945 for « determination of 24. The test for the determination of re is identical except that the mini- ‘mum ratio of armature voltage to armature current is used. 3. Saturation in Steady-State Conditions Short-cirouit ratio ia a term used to give a measure of the relative strengths of the field and armature ampere turns. It is defined as the ratio of the field current required to produce rated armature voltage at no load to the field current required to circulate rated armature current with the armature short-cireuited. In Fig. 10 the SCR is equal!. When no aturation je present it is skmply the reciprocal of the synchronous impedance, £4. It is impossible to specify the hest specific SCR for a siven system. Jn the past it hes been the practice in Eu- rope to use somewhat smaller SCR's than was the practice in this country. In recent years, however, the trend in this country has been toward smaller values. ‘The Preferred Standards for Lange 3600-rpm Condensing Steam Turbinc- Generators® specifies SCR of 08, ‘The desire for smaller SCA’s springs from the fact that the cost is sinalier with smaller SCR. On the other hand, static stability is not as good with smaller SCR. Regal tion is also worse but both of these effects are alleviated in part by automatie voltage regulators, Por most economical design « high SCR mackine usually has a lower 24’. ‘There- fore, both because of its lower 24’ and higher WR? a high SCR has a higher transient stability. This is not usually ‘significant factor particularly in condensing turbine ap- plications, because transient stability is not of great ico- portance in the systems in which they are installed. Tt may be quite important. for hydro-generators; the Boulder Dam machines, for example, are designed for SCR's of 24 and 2.74 ‘The effects of saturation arise primarily in the deter- mination of regulation. ‘Tests indieate that for practical purposes both the eylindrieal rotor and. the salient-pole ‘machino ean be treated similarly. Consideration will be given first to the characteristics for zero-power-factor lond- ing. Fig. 11 shows that for zero power-factor, the ry drop of the machine is in quadrature to the termina! voltage sand internal drop and can have litte effect upon regulation. It will therefore be neglected entirely. ‘The determination of the rated-eurrent, sero-power- factar curve can be developed as follows. Take any te minal voltage such as AF of Fig 10. The voltage behind leakage reactance is obtained by adding to this voltage th Jeakage reactance drop, Sk, which gives the line PQ. ‘The distance PR then gives the field current necessary for magnetizing purposes. In addition, however, feld current is required to overcome the demagnetizing effect of the armature current. This mmf is represented on the eurve by the distance ST, giving MT as the Geld current required to produce the terminal voltage OM with rated current, in the armature. Other points on the rated-current zero-powerfactor curve can be obiained by merely moving the triangle RST along the no-load saturation ‘Upon sliding the triangle RST' down to the base line, i can be seen that the total field current required to circulate rated current at short circuit which is represented by the point A, can be resolved into the current OD necessary to ‘overcome leakage reactance drop and the current DA re- ‘quired to overcome demagnetizing effects. Neither leakage reactance nor the field equivalent of armature current are definite quantities in the sense that they ean be measured separately. ‘They may be celeulated but their values are dependent upon the assumptions made for the ealeulations. Synchronous reactance, 2, is a definite quantity and is equal to the distance 40 expressed in either per unit or percent. When either 21 oF 25 in assumed, then the other Machine Characteristics Chapter 6 becomes determinable from Bq. (1) or from the triangle Just discusse ‘The foregoing analysis ie not strictly correct, ae it nog lets cortain changes in saturation in the pole structure, ‘The loakage from pole to pole varies approximately. pro- portional to the field current and the point 7 was deter- mined upon the basis that this leakage was proportional o the field current MS. ‘The increased field leakage at the higher excitation produces greater saturation in the field poles and this in turn increases the mmf required to force the flux through the pole, The net effect is to increase the field current over that determined by the method just discussed causing the tivo eurves to separate more at the higher voltages. The concept of the determination of the curve of rated. current af zeto-powerfactor by the method just described is valuable and in an attempt to retain the advantages of this method the concept of Patior reactance, zp, i8 intro- duced. The Potier reactance is the reactance that, used in a triangle of the general type described, will just fit be- twoen the two eurves at rated voltage. Tt ean be deter- ‘mined from test curves, see Fig, 16, by drawing DE equal TERMINAL VOLTAGE oe ELD CoRR Fla, 16—Zero power-factor characterlatica of generator to OA and then EF parallel to OB. The distance FG is, ‘then the Potier resetunee drop. Potier reactance is thus a fictitious reactance that gives accurate results for only one point, the point for which it is determined. For most max chines it is sufficiently accurate to use the one value ob- tained at rated voltage and rated current, Potier reactance, docreases with increased saturation. Sterling Beckwith” Proposed several approximations of Potier reactance, the ‘wo simplest are: 240.03(ry—2) and 208.24. For other loals at sero-power-faator, the conventionalChapter 6 method is to divide the lines BA and FD of Fig. 16 in proportion to the armature current. Thus for three-fourths rated current the regulation curve would be the line HLF in which BH and FJ are threeiourths of BA and FD, respectively. For powerfactors other than zero, several methods are available to determine the regulation, They all give sur prisingly close results, particularly at lagging power factors. The problem tay take either of two forms; the devermination of the terminal voltage when the Toal eur= rent, load power-fartor, and excitation are given, or the determination of the excitation when the load current, load power-factor, and terminal voltage are given. ‘The resist- ‘ance drop is So small that it is usually neglected. (@) Adjusted Synchronous Reactance Method’ — ‘This method utilizes the no-load and the rated-curront zero-powwerfactor curves. Ta obtain the excitation at any. other power-factor for rated current, an arbitrary excita- tion ia chosen such as OC of Fig. 17." The no-lead voltage NO LAD sarunarion CURVE 20% Power FACTOR ye Powen-racr3 BRLE, a Fig. 17—Determination of regulation curves for power-factors ‘other than zero by the “adjusted synchronoun reactance method.” Calis then regarded as an internal voltage and the distance AB as an internal drop of pure reactance, which is laid off {in proper relation with the terminal voltage as indicated by the power-fnctor of the load. The construction is as follows: The adjusted synchronous resetance drop AB is, laid off to make an angle with the X-axis equal to the powerfactor angle. A line equal to the distanee AC is then seribed from the point 4 until it intercepts the Y-axis at, "Described as Metéod (e) Para. 510 ia Refarencs 10. Machine Characteristics 18h / [wo coro. J sartnarion cuave rower-racton / eee foe \ N/a ms “sy . FAGTOR RATED CURRENT, emma yourace ve aN CA a aNOLe: , e FELD CORRENT Fig, 18 Determination of excitation, including the effects of the point F. The verival distance F is then the terminal voltage for the partietiar excitation. Pollowing chis pro cedure another excitation is rhesen ane "he construction repeated from whieh thevlotted lines obtained. ‘The intor- section ofthe line with the norasal voltage gives the exe:tae tion for the desired powerfactor at rated load. Uf the ‘machine is not operating at rated current. the zero-power- factor curve corresponding to che particular current should be used. (b) General Method—For lack of a better name th method has been called the “General Method." Ti is based upon the assumption that saturation (sineluied by reading the excitation requirements fram the no-load saturation curve for a voltage equal to the voltage behind the Pocier reactanee drop, ‘The method is deseribed in Fig. 19 with all terms exe pressed in per unit. The voltage, ty i@ the Potior internal Voltage or the valtage hehind the Potier reactance drop. Fig. 19 Determination of eld current for round rotor machine with saturation included by adding s in phase ‘wich152 "The distance jk represents the synchronous reactance drop, zai. If there were no saturation the synchronous internal, voltage would be Of. When using per unit quantities throughout this is also equivalent to the ficld current, ‘This method includes the effect of saturation by simply adding the increment in field current for this voltage in, fexcess of that required for no saturation, to Ok in phase ‘with ep, giving 8 a result, UA, When per unit quantities are not used the construction is a little more complicated. Tt involves the construction of e, separately so that s ean bbe obtained in terms of field eurrent.. This quantity is then ‘added to the diagram for no saturation in terms of the field, current, In Fig. 18, fret lay off from the terminal voltage, (0a, and then the z, drop ac at an angle with the horizontal, ‘equal to the power-factor angle. Oc then represents ¢». BY scribing this back to the ordinate and reading horigontally, the excitation corresponding to this voltage is obtained. ‘The effect of saturation is introduced by the distance s. ‘Whe field curvent required if there were no saturation is obtained by de construction Oj and jk where Oj represents the excitation, ap, required to produce the terminal voltage fat no load and jh the excitation, of, for the synchronous reactance drop, read from the abscissa. ‘These vectors eorrespond to e, and jk, respectively, in Fig. 19 excopt ‘that they are in terms of feld eurrent.' If kA, equal to the saturation factor, 2, ie addod along a line parallel to Oc, the total excitation Oh is obtained. (©) Round Rotor Potier Voltage Method"—This method is the same as (b) except that the effect of satur tion +, in Fig, 18 is, for the sake of simplicity laid off along, Ok, making om the desired excitation, Av oan be seen, there is little difference between those two methods. This method gives the best averall results, especially at loading power factors, ‘The particular name of this method was assigned to distinguish it from the next method. @ Two-Reaction Portier Volsage Method -This ‘method is similar to that of (c}, exeept that the toro-rean- tion method of construction shown in Fig. 14 is used to devermine the excitation before including the saturation actor s. Kig, 15 shows the entire construction, For the ssuke of comparison with other methods, the construction ig also shown in Fig. 18. The construction is the same 1s (©) except that the line Oz is made to pass through the point ginstead of k. This arises because x, isemaller than x4. 4, Reactive Power Capacity ‘The capacity of a synchronous machine to deliver re- active power is dependent upon the real power that it, delivers, ‘Two limitations from the heating standpoint are recognized: (1) that due to the armature, and (2) that due to the field. Figure 20 shows the reactive power eapability, of a standardized 3600-rpm stearn turbine-generator. Real power is plotted as abscisea and reactive power aa ordinate. All the curves are ares of circles. ‘The line centering about the origin represents the limit imposad by the condition of constant armature current whereas the other are by con- stant fleld current. With regard to the latter, the generator ccan be likened to simple transmission Tine of pure react ance, x2, with the receiver voltage held at constant value, x, the terminal voltage of the generator, and with the “Described as Method (a) Par. 1.520 in Reference 10. Machine Characteristics Chapter 6 kvan-PeR ONT as ‘ew Pen uur Pig. 20—Reactive power capacity of etoam turbin 20.000 kor, 25529 va, OAS pt. 0.8 SCR, at 0.5 pslg hydrogen. sending voltage held at a constant value ¢y, As shown in Chaps. 9 and 10 the power cirele of a line of such charaeter- isties has its center in the negative reactive axis at and tos has its center in the negative renctive axis at its internal voltage, ¢, must be such that ite radius, passes through the point of rated real power and rated Teactive power. Actually, however, the center is usually Tocated atu point equal to (SCR) times (rated kya). This in to take caro of saturation ofeets. Since, however, with tome 2s equa to SCH, cn mn ho this condition both relations reduee to an equivalence, The leading kvar eapacity (underexcited) of air-cooled condensers is usually about 30 percent of the lagging kvar eapaoity but for hydrogen-cooled condensers about 42 percent. Il, THREE-PHASE SHORT CIRCUIT In addition to its steady-state performance, the acti ‘of a machine under short-circuit conditions is important, "The presence of pathe for flow of eddy eurrents as provided by the solid core in turbine generators and by the damper windings in some salient-pole machines makes the trest- ment of these machines, from a practical viewpoint, less complicated than that for salient-pole machines without damper windings. For this reason the three-phase short- cireuit of these types of machines will be discussed first. Armature resistance will be negleeted except asit influences, decrement factors,Chapter 6 5. Three-phase Short-Circuit of Machines with Current Paths in Field Structures Consideration will be given to a simultaneous short ireuit on all phases while the machine is operating at no-load normal voltage without a voltage regulator. ‘The general nature of the currents thal appear is shown in Fig. 21. They ean be divided into two parts: a. An slternating component in the armature and as- sociated with it an unidirectional component in the field. ‘These two components decay or decrease together with the Fig. 21—Three-phase short circuit in aallent-pole machine ‘with damper windings same time constants. The alternating armature component ‘ean be regarded as being produced by ita associated uni- directional component in the ficld, All phase components of the alternating current are essentially the same except that they are dispiaced 120 electrical degrees.* b. An unidirectional component in the armature and an temating component in the field or in the damper scind- ings, Tn this ease, likewise, the alternating current in the ficd winding ean be regarded as produced by tho uni- directional component in the armature. 6. Alternating Component of Armature Current This component ean in birn_ be resolved inta several components, the rims. values of which are shown in Fig 22, They are: a. The steady-state component ». The transient component ¢. The subtransient component Each of these components will bo discussed separately. Steady-State Components—The steady-state com: ponent, as its name implies, is the current finally attained. Because of the demagnetizing effeet of the lunge short circuit current, the flux density within the machine de- creases below a point where saturation is present. Satura- ‘The machino used in thie eave was 4 salientpole machine. Ax will be seen Ister, such machines also contsia a second barmoaie component of current. Thin type of raachine was chet to abi ‘ore cleaely the presence of field and damper currents Machine Characteristics 153 Tine Fig. 22—Symmerrical component of armature short-cireult ‘current (Chree-phase short circuit from no-load rated ‘ste). Values are em tion is important only as it affects the field current neces- ary to produce normal voltage at no load. The steady- state value of short-cireuit current is thus equal to the line-to-neutral voltage read from the air-gap line for the value of field current required to produce normal voltage divided by the synchronous reactance in ohms. Transient Component—Ii the excess of the symmet- rieal component of armature currents over te steady-state component be plotted on semi-log paper, it ean be seen, at this excess, exeept for the first few eyeles, is am ex ponential function of time (the points lie ina straight line) Eaxtending this straight line hack to zero time and adding the steady-state component, the so-called ransient com- ponent, {,, or armature current is obtained. ‘This compo- nent is defined through a new reactance, called the tranatent reactance by means of the expression, ‘The manner in which this quantity is related to the ex- ponential and steady-state verms is shown in Fig. 22 Tn discussing this component, the presence of the damper-winding currents of salient-pole machines and, rotor eddy currents of turbine generators can, for the moment, be neglected. Before short-cireuit occurs the fx associated with the fiold windings can be broken up into two componente (see Fig, 23), a component # that crosses the air gap and 2 component dy, a leakage flux that. ean be regarded as linking all of the feld winding. Actually, of course, the leakage flux varies from the base of the pole to the pole tip. ‘The flux is so weighted thet it produces, ‘the same linkage with all the field (urns aa the actual Jeak= age flux produces with the actual turns. It is approximately. proportional to the instantaneous value of the field current Z,, The total fiux linkages with the field winding are then those produced by the flux (P-+2)). AAs the field structure rotates, a balanced alternating voltage and current of nor- mal frequency are produced in the armature, Because the armature resistance is relatively small, its cirenit can be rogarded as having ® power-factor of zero. The symmet- rieal current thus produeed develops an maf thet rotates synchronously and has a purely demagnetizing, as eon trasted with cross magnetizing, effect on the field fluxes It is a wellknown fact that for the flux linkages with154 i \ | Fig. 23—Alr-gap and leakage fluxes at no Toad. cireuit to change instantly, an infinitely large voltage is necessary and the assumption is justified that forthe tran- sition period from the no load open-circuited condition to the short-circuited condition, the flax linkages with the field winding ean be regarded as constant, ‘This s equivalent to saying that the flux (b-+4,) remains constant. In order that this flux remain constant in the presoneo of the de- insgnetizing effect of the armature current, i is necessary tia the feld eurrent 7+ incresse to overcome the demag- netising effect of the armature curcent. If Z;inereases then, %;, which is proportional to it, must likewise increase. Tt follows then that & must decrease. Consideration of the steady-state conditions has shown that the airgap voltage, is proportional to the airgap flux &. ‘The armature current for short-circuit conditions is equal to. If and conseciuently e) had remained constant, during the transi tion period, then the transient component of short-cireuit current would be merely the no-load voltage before the short-circuit divided by the leakage reactance and the transient reactance would be equal to the armature leakage reactance 2. However, as just shown, the sicgap flux doerenees and, therefore, the armature current is Tess. Te follows then that the transient reactance must be greater than the armature leakage reactance. It is a reactance that includes the effect of the increased field leukuge ‘eeasioned by tho increase in field current, Under steady-state conditions with m: armature eurrent ean be viewed as prodtced by a fi internal voltage equal to xs ia whose magnitude is picked from the airgap Tine of the no-load saturation curve for the particular field current. At the first instant of short- circuit, the ineressed armature current, t, ean likewise be viewed as being produced by a fictitious interaal voltage behind synchronous reactance, whose magnitude is xy! ot 2254, if the short-circuit be from rated voltage, no loed. ‘This voltage provides s means for determining the initial ‘value of the unidirectionsl eomponent of field current by picking off the value of Z;on the air-gap line of the no-load saturation curve corresponding to this voltage. If it were Machine Characteristica Chapter 6 possible to inerease the exciter voltage instantaneously to ‘an amount that, would produce this steady-state field cur- rent, then this component of short-cireuit current, would romain sustained. It is important to grasp tho significance of this truth. ‘There is always a constant proportionality between the alternating current in the armature and the Unidirectional (often ealled dirget-eurrent) component of current in the field winding, whither the operating condi- tion be sweady-state or transitory “The initial value of armatare current, ax stated, grade ually decreases to the steady-state and the induced current in the feld winding likewise decreases to its steady-state magnitude. ‘The inerements of both follow an exponential curve having the sane time constant. Attention will ext hoe given to considerations affecting this time constant Ia constant diect voltage is suddenly applied to the field of a machine with the armature open-circuited, the current builds up exponentially just as for any circuit bev- ing resistance and inductance in series. ‘The mathematical expression of this relation ist in which en [- *] xis the exciter voltage. 1s the resistance of the field winding in ohms. Ts, is the open-circuit transient time constant of the ‘chine or of the eireuit, in question in seconds. ‘is time in seconds. ‘The time constant is equal to the inductance of the field winding divided by its resistance. Tn the ease of the short- circuited machine, it was shown that at the first instant the flux linkages with the field winding remain the same 18 for the open-eireuit condition, but that the direct com- @ ponent of field current increases to 74 times the open- circuit value before short-circuit, Since inductance is defined as the flux linkages per unit current, it follows then that the inductance of the field circuit under short- circuit must equal & times that for the open-circuit con- dition, ‘The short-citeit transient time constant, that is, tho time constant that detarmines the rate of decay of the transient component of eurrent must then equal Te = 227%, in seconds % ‘The component of armature current that decays with this time constant can then be expressed by Gal ide When # is equal to Ta’ the magnitude of the component hhas decreased to «~ or 0.968 times its initial value. This instant is indicated in Fig. 22 ‘Subsransient Component In thepresence of damp- cer windings or other paths for edily currents as in turbine genorators, the airgap flux at the fist instant of short cirouit is prevented from changing to any great extent. ‘This results both from their close proximity to the air gapChapter 6 and from the faot that their leakage is rauch smaller than that of the field winding. Consequently, the initial short circuit currents of such tnachines are greater. If this excess of the symmetrical component of armature currents aver the transient, component is plotted on semilog paper, the straight line thus formed can be projected back to zero time. ‘This zero-time value when added to the transient, component gives the subtransient eurrent, ij". This oub- transient current is defined by the sublransiont reaciance in ‘the expression ‘The subtransient reactance approaches the armature Teak age differing from that quantity only by the Teakage of the damper windings. Since the exeess of the armature currents represented. by the subtransient components over the transient com ponents aze sustained only by the damper winding cure Tents, it would be expeeted that their decrement would be etermined by that. of the daroper winding. Since the copper section of this winding is so much smaller than that of the fold winding, itis found that the shori-eireuit ssubtransiont time constant, 7’, is very small, being about 0.05 second instead of the’ order of seconds as is characteristic of the transient component. ‘The component Of armature current that decays with this time constant is (id! “ig) and can be expressed as a function of time as Wine ‘Thus the time in seconds for this component to decrease to 0.368 times its initial value gives 7" as indicated in Fig. 2, ‘Tests on machines without damper windings show that Decause of saturation effects the short-circuit current oven inthis ease. be resolved into a slow transient component ‘and a much faster subtransient component. ‘Tho influence of current magnitudes as reflected by saturation upon the transient and subtransient reactance is discussed in more detail under the geveral heading of Saturation. 7. Total Alternating Component of Armature Cur- rent ‘The total armature current consists of the steady-state value and the two components that. decay with time con- stants 7,’ and 7". It can be expressed by the following equation WIDTH Hf ing FO 48 o ‘The quantities are all expressed as rms values and are equal but displaced 120 electrioal degrees in the three phases. 8. Unidirectional Component of Armature Current ‘To this point consideration has been given to flux Link ‘ages with the field winding only. ‘The requirement that, these linkages remain constant at transition periods de- termined the alternsting component of armature current. Since these components in the three phases have a phase displacement of 120 degrees with respect to each other, Machine Characteristics 155 ‘only one can equal zero at a time. Therefore at times of, three-phase short-circuits, the alternating component, of current in at least two and probably all throe phases must change from zero to some finite value. Since the armature cireuits are induetive, it follows that their currents cannot change instantly from zero toa finite value. ‘The “theorem fof constant flux linkages” must apply to each phase sepa- rately. The application of this theorem thus gives rise to ‘an unidizeetionsl component of current in exch phase equal and of negative value to the instantaneous values of the alternating component at the instant of short circuit. In this manner the armature currents are made continuous 23, shown in Fig, 24, Each of the unidirectional components 7 ‘syMweTRICAL, ‘COMPONENT Fig. M—The Inclusion of a d-c component of armature Corrent whoee eriatence la necessary to make the armature ‘current continuous at the instant of ehort clreuit. in the three phases decays exponentially with a time com stant 7, called the armature short-circuit time constant. ‘Tho magnitude of this time constant is dependent upon the ratio of the inductance to resistance in the armature circuit. As will be shown the negative-seqtience reactance, 2, of the machine is a sort of average reactance of the iafuialure with the feld winding short-circuited, 20 that it is tho reactance to usein determining T,. ‘There exists then the relation —s Batre in which ry ia the d-e ensistance of the armature, ‘The quantity 2ef merely converts the reactance to an indue- tance. "The maximum magnitude which the unidireetional seconds ® PRQUECTION AXIS FOR PHASE B +hhor ease © Co} Fig. 25--Representation of instantaneous currente of a three phase aystenn, fa) Three separate vectors projected om s-azia, (b) Single veetor projected on three axes. (o156 ‘component can attain is equal to the maximum of the alternating component. Therefore, Fanon VEE ® A symmetrical three-phase set of currents ean be repre sented as the projection of three equal-spaced and equal length vectors upon. a stationary reference, say the real axis. They can also be represented us the projection, as it rotates, of one vector upou three stationary axes, spaced 120 degrees. ‘These axes can conveniently be taken as shown in Fig. 25, as the horizontal-axis and two axes hav- ing a 120-degree relation therewith. Since the initial mag- nitude of the unidirectional components are the negatives of the instantaneous values of the alternating components ‘at zero time, then the unidirectional components can be represented as the projection of a single vector onto the ‘three equal-spaced axes. This fact is used at times to determine the maximum magnitude which the unidiree- tional component ean attain. By its use it is unnecessary to await test in which the maximum happens to oceur. ‘This method is in error, however, for machines in which tq and 4" ate radically different. 9. Total RMS Armature Current The rms armature current at any instant is Via Fine ‘The minimum current thus oceurs in the phase in which the unidireetional component is sero and the maxiroum ‘occurs when the unidirectional component is a maximum, that is, when maximum dissymmetry occurs, Since the maximum value that the unidirectional component can attain is V/2°™, then [=] +[eerfeveest co) Cf cour arms valuo a ite neme iplis, isan average aunty and ie usally taken over a eyelet hall epee af Une. ‘The foregoing expression asnumnee tht both the Aerating snd the uniiectional componente do not do. Crease, Doense of the natural deccement, during the st eyete’ In reality the decrement i wualy uficient 10 minke the elect noeeabie. In applying event breakers itis onal to use a factor 1.6 instaud V3. This factor Secs smal deorement 10. Effect of External Impedance Tf the short-circuit occurs through an external impod- lance raj dna, UNG Tans is not too large, their effect cat be intraduced by merely increasing the armature constants, by these amounts. ‘Thus the components of shart-cireuit ‘current become ay (42) (13) Machine Characteristics Chapter 6 ‘The short-circuit time constant is affected in « similar manner atte, Babtas For the armature time eonstant, the external reactance must be added to the negstive-sequonco reactance of the machine and the external resistance to the armature resist- ‘ance of the machine. The expression then becomes tebe 2a Because of the much lower ratio of reaetanee to resistance in external portions of cireuits, such as transformers or ‘transmission lines, in the vast majority of eases T'. for faults out in the system isso stnall as to justify neglecting, the unidirectional component of current. 11, Short Circuit from Loaded Conditions ‘The more usual ease met in practice is that of a short- cireuit on machines operating under loaded conditions. As before, the short-circuit current in the armature can be divided into two components, a symmetrical alternating component, and 2 unidirectional eomponent. Alternating Component—The alienating compo- nent in turn ean be resolved into three components: (1) steady state, (2) transient, and (3) subtransient, Bach of these components will be discussed individually. ‘The lose on the machine affects the weady-slate com ponent only as it influences the field current, before the ‘short cironit, ‘The field current can be determined by any of the methods discussed under the heading of “Stoady- ‘State Conditions.” Saturation will be moreimportant than for the no-load condition. ‘The steady-state short-circuit current, is then equal to the line-to-neutral voltage read rom the airgap line for the field current obtained for the loaded condition divided by 2. In the discussion of the determination of the transient component fro the no-load condition, it was stated that ‘the quantity that remained constant during the transition period from one circuit condition to snother, is the flux linkages with the field winding. For the short-cireuit from Toaded eonditions this same quantity ean be used as a basis for analysis. Consideration will be given first to a load before short circuit whose power factor is zer0, lagging, fand whose current is ign. ‘The flux linkages before short Greuit will be determined by a superposition method, obtaining fist the linkages with the field winding for zero armature current and any terminal voltage and then the flux linkages with ermature ourrent, i, and zer0 terminal voltage. ‘The total flux linkages is the sum of the two ‘values #0 obtained Let Ya be the flux linkages with the feld winding at no-load at rated voltage. For any other terminal voltage such a5 ¢, the flux linkages y will be equal to Te "Tin seconds as) re jn seconds (as) ra) Paw By definition the transient reactance of a machine is equal to the reactance which, divided into the line-to- neutral rated voltage, gives the transient component of (16)Chapter 6 short-cireult current. at no-load normal voltage. If this short-circuit current is designated as iy, then fase a7) At the instant of short-circuit from no-load at rated volt- age, the flux linkages with the field winding, 4, remain ceonstent, ‘The demagnetizing effect of the armature cur- rent is overcome by an increase in the Geld current. Thus the armature earrent ij, with its associated Geld eurzent ‘which is alvays proportional to it, ean be regarded as pro- ducing the ux Unkages yx with the field winding. For any other armature current, ia assuming always that the armature is short-circuited, se fua Takages with the Geld ¥;. Combined with Eq. (17), if, can be eliminated giving ytd be whi expression was derived from considerations applying only to the instont of transition, its application ie more general The only necessary considerations that rust be sutiafied, are that the armature be short-circuited and that the field ite winding are equal to this current contain a component of current to overcome the conditions of short eet, eo in general Hb perntasble to replace inthis exprowion hy tu. "Tho fl Hnkages with the Beld winding for the steady-state shortrret Wa condition thus become tax; By application of the superposition theorem, the total flux linkages with the field winding can then be regarded 136 the sum of the flux linkages produced by the terminal voltage, namely — rand ths by te armature 2 iy ue arma eared lps wl by 9 deta he Eidos dre aide” dat ces We ogee dst a rent with zero terminal voltage, namely tu tingle mtoctatin) A. 8) Since the flux linkages with the fiold winding produced by a unit of current iy under short-circuit conditions is exalt 27 thn he tens capone te a > > yy e e fen Fig. 26—Construction for the determination of internal volt- Machine Characteristics 15T circuit current i” can be determined by dividing these linkages into the total flux linkages just determined. ‘This gives a9) ‘The numerator of this quantity can be regarded as an internal voltage, e, which is equal to the terminal voltage plus a transient renctance drop produced by the lond current. ‘When the power factor of the loads considered is other than zero ingging, the veetor sense of current and terminal ‘voltage must be imtreduced. ‘This ean be accomplished by computing efor the operating condition in the same man ts i INTERWAL_VOLTAGE +] a TREAGTANCE IN PERCENT Fig. 27—Machine internat voltage ax s function of reactance, Foll-load rated voltage. ner that eq was determined in Fig. 14, except that-za should bbe replaced by 24’. ‘The voltage e:’ should then replace ectsa' ign in (19), However, for nearly all practical pur Doses itis suficiently accurate to replace e,’ by the ampli- tude of a quantity «, whieh is usually referred to as the voltage behind transient resetanee to distinguish it from, similar internal voltages for which leakage, ynchronous or subtransient reactance is used. The construction for this quantity is shown in Fig. 26 and to assist in the ready evalustion of the amplitude the curves in Fig. 27 are. provided. The transient component of short-circuit current, is then (20) ‘The suberansient component of short-circuit current ia obtained in a manner sirilar.to the transient component except that the subtransient reactance is used in the ealen- lation of the internal voltage ¢;". For loads of zero-power- factor lagging the subtransient, reactance drop, "ian,158 caused by the armature current is direetly additive to the ‘torminal voltage and for zero-power-factor leading directly. subitactive, For other power-factors e” can be obtained, from Fig. 27 by using xy”. The cubtransiont component of, short-cireuit ourzent is then (an) Unidirectional Component—In the three-phase short-cireuit from no load, the unidireetional component, ‘of current, was introduced to prevent non-contintious, transition of the instantaneous value of current from the no-load to the short-cireuit condition. The unidirectional curtent performs a similar role for the short-circuit from Joaded condition. Before the short-eirouit the armature ‘current is equal to iar end has some position with reference to ¢ such as shown in Fig. 28. ‘The subtransient eom- a ig. 25~Showing that ls for a short ccult from load ls equal to the megative of -/2 times the diflerence between {” and I. ponent, i, lags e" by ninety dogrees so i and ia will be determined with respect to each other. The v/2 times the vector difference between these Uo quantities (singe they fare rms magnitudes) gives the unidirectional component necessary to produce smooth transition. ‘The mngnitide of this quantity varies between this amplitude and 200 depending upon the point in the eyele ut which shorl= ‘ereuit occurs Other Considerations—Time constants aze not inflie enced by the nature of loading preceding the short-cieuit ‘Total rms eurrents can be determined by the relations already given. 12, Three-Phase Short Circuit of Salient-Pole Ma- chine without Damper Windings For most applications it is sufficiently accurate to treat the salient-pole machine without damper windings just as other machines. It must be recognized, however, that this is only an approximate solution. Among ather complica- tions, in reality a strong second harmonie is present in the armature current, Doherly and Niekle! have developed expressions for the armature currents for x three-phase, shott circuit from no loud. ‘These are given below. Machine Characteristics Chapter 6 WF cos (2nft+-a) +5 cos (2xft+-a) tate a tena, safe Se Toone l22) eat tos (infi+a)~ 77 008 (2eft--a—120") +E ons Qrft-ba- 120°) oe Ts 608 (Anft-a—120) mae cos (a4 1207) (23) (24) Where 25) ratty os aemAngle which indicates point on wave at which short-cireuit occurs. ‘The instantaneous field current, Za, is Le en Where T,=Initial value of field eurrent. mete tea ta] III. UNBALANGED CONDITIONS 13, Phase Currents for Unbalanced Short Circuits As explained in the chapter relating to Symmetrical Components, the unbalanced operating conditions of a rotating maebine can. for most purposes be deseribed in terms of three eharuelesistie constants: the positive-se- quence impedance, the negative-sequence impedance, and the zerosequence impedanee. The shortcireut currents can be resolved, as before, into the steady-state, transient, and subtransient components. ‘The difference between these components decreases exponentially as before. ‘The components of armature current and the time constanta for the different kinds of short-cirenita are given below for short-crcuits at the terminals of the machine, For three-phase short-cteuit (28) Te Tas For terminal-to-terminal short cireuit, the a-e components oof the phase enrrents are given byChapter 6 viet" BT 30 Boke i wh a 6 the negtiveseguene impedance of the Tor leninal-o-neutral shor cout, the we omponents of the phase cureents are given by Bai" ye, Hints US Titatn Be, aaltoct ten, ‘atartie Tay bate 9) in which 2» is the zero-sequence impedance of the machine. ‘Phe subtransient timo constant, 7", does not change sig- icantly with different conditions and, therefore, the ingle value is used for all conditions. The unidirectional components and the rms values are determined just as dleseribed under the general subject of “Short Cireuit from Loud.” The above values of e;, «' and e," will naturally be those values corresponding to the particular load condition. ‘The ratio of the phase currents for terminal-to-neutral to threo phace short cireuite ean be obtained from Faq's (90) and (28). Thus, for the phase currents Terminal-tomnoutral short cirouit xa" Machine Characteristics 159 up by the armature current. If the axis of the short- cirouited field winding lines up with the axis of pulsating field then the current is large and if the rotor is moved through 90 electrical degrees then the current is much smaller. The first position corresponds to the ease of » transformer in which the secondary winding is short- circuited, the field winding in this ease corresponding to the secondary winding of the transformer. This is the position in which the subtransient reactance, 23”, is deter- mined. Ttis equal to one-half of the voltage from terminal- to-terminal divided by the cnrrent. For thesecond position the field winding is in quadrature to the pulsating Geld ‘and consequently no current flows in the field winding. ‘The armature current is then determined by the magnetiz- ing characteristics of the air gap in the quadrature axis. ‘The subtransient reactance, 2", is determined when the field is in this position and is equal to one-half the quotient of the voltage divided by the current, ‘The reactance for intermediate positions varies between these two amounts in apcordance with the curve shown in Fig, 29. GURORATURE ANEW zB +++ NO DAMPERS I axis, 8] “Threo-phase short circuit au"-txt ‘The negative-sequence impedance, x is usually equal to x4”, but for many machines 7% is less than x4”. For ‘these eases, the terminal-to-neutral short-circuit current is greater than the three-phase short-cireuit current. The generator standards require that the machine be braced ‘only for currents equal to the three-phase values. Tn order that the terminal-toneutral current not exeeed the three- hase current a reactor should be placed in the neutral of ‘the machine of such value as to bring the zero-sequence impedance of the circuit equal to xe”. Thus, the neutral reactor, 2, should be 14, Negative-Sequence Reactance ‘The negativesequence impedance of a machine is the mpadance offered by that machine to the flow of nogative- sequence current. A set of neqative-sequence currents in the armature ereates in the air gap a magnetic field that rotates at synchronous speed ina direction opposite to that of the normal motion of the field structure. Currents of double frequency sre thereby established in the field, and in the damper winding if the machine has ome. "The imaginary component of the impedance is called the nega tive-sequence reactance and the real component the nega tvesequence resistance. ‘These will be discussed sepa rately, in the order mentioned. Tf single-phase voltage is applied across two terminal of a salient-pole machine without dampers while its rotor is stationary, tho resulting current is dependent upon the Position of the rotor with respect to the pulsating field set ve SE avENGE T ol “OGKES- ROTOR WETHOE (COPPER OAUPERS 1 a a 2 ANGULAR POSITION OF ROTOR MM DEGREES Fig. 29-Rolation between subtransient and negative-se- "quence reactance, When a set of negative-sequence currents is made to flow through tho armature with the field short-circuited and rotating in ita normal direction, thon the field winding takes different positions successively as the armature field rotates with respect ta it. The nature of the impedances in the two extreme positions, that is, where the field wind: ing lines up with the magnetic field and where it is in quadrature with it, should be somewhat the sare az" and x,", the only significant difference being the fact that, in the determination of z4"” and 24", currents of normal frequency were induced in the field, whereas, in the nega- tive-sequence ease the currents are of twice normal fre- queney. One would expect therefore that the negative sequence reaetance 2, is some sort of a mean between 4" and 2,’", and such is the case. According to the ALEE test code," the definition of negative-sequence reactance is equal to “the ratio of the fundamental component of re-160 active armature voltage, due to the fundamental negative: sequence component of armature current, to this compo- nent of armatie current at rated frequency.” A rigorous interpretation of this definition resulls in 2, equal to the arithmetiomean®* £72", However, several diferent dei- nsean be given for». That this is possible is depend- cent largely upon the fact that when a sinusoidal set of negative-sequence voltages is applied to the armature the currents will nol be sinusoidal. Conversely if the currents are siuasoidal the voltages will nt be. In Table 1 are shown expressions" for , based upon different definitions. ‘This table is based on machine ‘without damper windings for which 2,” is equal to-zy, and sd” is equal to x4. In this table Vint vae For each test condition i is possible to establish definitions based on whether fundamental or root-mean-square cur- ronts are specified. For example, inthe first definition if the fundamental component of armature current ie used in caleulating 2 then the expression in the first column should be used, but if the root-mean-square figure of the resultant current is used then the expression in the second column should be used. Tn onler to orient one's self as to the relative importance of the different expressions, figures have been insert in the expressions given ia Table 1 for a typieal machine having the constants xa! =85%, 24=70%, and 24 100%. ‘The magnitudes are tabulated in the righthand columns of Table 1. From the standpoint of practical application, the negative-sequence reactance thal would result in the proper root-meun-oquare curront for method (3) would appear to bo the most important, However, the method of test to determine this quantity involves a sudden short- circuit and from this standpoint proves rather ineonveni- ent. On the other band, the figure for 2. obtained from the use of the root-mean-square values ina sustained Single-phase short-cireuit current [mthod (4)], is nearly equal to this quantity. When the resistance is negligible this negative-sequence reactance is equal to V3E. I 6 4 en Machine Characteristics Chapter 6 where T equals the root-mean-square armature current in the short-cirenited phase; and E equals the rool-mean- square open-cireuit voltage between terminals before the short-circuit is applied or the no-load voltage correspond~ jing to the fiold current at which I is read. Tn general, the same arguments can be applied to other types of machines sueh ae turbine generators and salient- pole machines with damper windings when the parameters xy" and 24" are used. For such machines the difference between 29” and zy is not great. The values for x4” and 1” of amachine with copper dampers are given in Fig. 29. For such machines the difference between 2 based on the Gifferent definitions of Table I will become inconsequential. In addition, for turbine generators, saturation introduces variables of much greater magnitude than those just con- sidered. Por these machines negativeeequence reactaneo ean be taken equal to 2a". Method of Test—In addition to the method implied by the AIEE Code and the ASA whereby 2 is defined as the arithmotie mean for 24" and 24, 2 ean be determined directly from test either by applying negative-sequence voltage or by the method shown in Fig. 30. Jan 60a al = PFS 0 her 1 Ey Eye Bo=-E Pu WOtew 08) fu_ie'~a) fede Bu hee nhs PF ° yatumeter reading ox Zp, where P=wattnetr reading, nob) 00s planets Fig, 30-Determination of the negative-sequence Impedance ‘of eymmetrically-wound machines. ‘TaBu 1—DErINiTions oF NeGative-Sequence REACTANCE, Data a nett (© Applcstion of siouvoidsl uegativese| _2n'zy Root Mean-Square ‘quence voliage achat © suet ‘ 1 vanes es | os tal prnaie cmpren gf my wVii-na vain | | (@) Bustained single-phase short-circuit current! sa (VI=0-2) + Voda VINE 30 av GiSincestd tweens) VERY s | O ATER Ash sotsd 2 |Chapter 6 With the machine driven at rated speed, and with a single- phase short-circuit applied between two of its terminals (neutral excluded) the sustained armature current and the voltage between the terminal of the free phase and either of the shorL-cireuitod phases are incusured. ‘The reading of & single-phase wattmeter with its current coil in the short-cireuited phases and with the ahove mentioned volt~ age across its potential coil is also recorded. The negative sequence impedance equals the ratio of the voltage to the current so measured, divided by 1.79. ‘The negative sequeace reactance equals this impedance multiplied by the ratio of power to the product of voltage and current, 15. Negative-Sequence Resistance ‘The power associated with the nogative-soquence cur- rent can be exprossod as resistance times the square of the current. This resistance is designated the negative sequence resistance. For a machine without damper wind- ings the only source of less is in the armature and field resistances, eddy currents, and iron loss. ‘The copper lose in the armature and field is small as is also the iron and eddy loss in the armature, but the iron and eddy loss in the rotor may be considerable, Copper damper windings provide a lower impedance path for the eddy currents and hinder the penetration of flux into the pole structure, ‘The relatively low resistance of this path results in a smaller negative-sequence resistance than if the fx were per- mitted to penetrate into the rotor. For higher resistance damper windings thenegative-seqiuence resistance increases to « point beyond which the larger resistance diminishes the current in the rotor eireuits sufficiently to decrease the loss Induction-Motor Diagram—The nature of the neg- ativessequence resistance is best visualized by analyzing the phenomena occurring in induction motors. In Fig. 31 fe Fig, 31—Equivatent clrcutt of Induction motor. is given the usual equivalent circuit of an induction motor in which reestator resistance, ,=stator-leakage reactance at rated frequency. resrotor resistance. z,=rotor-leakage reactance at rated frequency. 2,=shunt impedance to include the effect of raagnetiz- ing current and no-load lowes, applied voltage. Tyestator current, Tre rotor current. aeslip. ‘The justification for this diagram ia shown briefly ax follows: The airgap flux created by the currents I, and I, Machine Characteristics 161 induces the voltage Hy in the stator and sE, in the rotor, In the rotor the impedance drop is rd bien, 32) since the reactance varies with the frequency of the cur rents in the rotor. The rotor current is therefore deter mined by the equation ak yerlet jer, or Ben et jes (3) It follows from this equation that the rotor circuit can be completely represented by placing a cireuit of impedance [4 je, aeross the voltage Ey. The total power absorbed by © must be the sum of the rotor losses and the useful shaft % into the resistances r, and power, so that, resolving 1, de power sured by epont the role op per loss. ‘The power absorbed by =r, represents the useful shaft power Neglecting rand the veal part of za, the only real power is that concemed in the rotor cirouit. Assume that the induetion motor drives a direct-current generator. At small slips the electrical input into the stator is equal to the copper loss, ie., the 7.7, of the rotor plus the shaft, load. With the rotor locked, the shaft load is zero, and the total electrical input into the stator is equal to the rotor copperloss. At-200-percent slip, .c., with the rotor turning, at synchronous speed in the reverse direction, the copper Joss is (ry the electrical input into the stator is lire tnd the shaft load 1? or A nogative chaft, functioning as a generator is now a motor. Phy ia just what would be expected, for aa the slip increases, from zoro the shaft power inereases to a maximum and then decreases to 20r0 for 100-percent slip. A further ine crease in slip necessitates motion in the opposite direction, which requires a driving torque. At 200-percent slip the electrical input into the stator is equal to the mechanical .put through the shalt; half of the copper loss is supplied from the stator and half through the shaft. This is the condition obtaining with respect to the negative-sequence in which the rotor is rotating at a slip of 200 percent. rola- tive to the eynehronously rotating negative-sequence field in the stator. Holf of the machine loss associated with the negative-sequence current is supplied from the stator and half by shsft torque through the rotor. ‘The factors of fundamental importance are the power supplied to the stator and the power supplied to the shaft, which can always be determined by eatving the equivalent circuit involving the stator and rotor constants and the smagnetizing-current constants, A more convenient deviee,,162 Beet ike @ ixr, © xelarremeard) Beastie ® rc) Fig. 32—Development of negative-sequence resistance and Feactance from equivalent circuit of Induction mot (a) Negative-sequence diagram for induction motor Aecting armature and no fond losses; (c) simplified network ‘aegative-eoquence reslatance and reactance, singe ¢ is constant and equal to 2 for the negative-sequence, is to reduce the equivalent network to a simple series im- pedance as shown in Fig. 32 (¢). ‘The components of this impedance will be ealled the negative-sequence resistance ry, and the negutivesequence reuctanee #3. The current, flowing through the negative-sequenee impedance is the ‘current flowing through the stator of the machine, and the power loss in ris equal to the loss supplied from the stator ‘of the machine and the equal loss supplied through the shat. ‘The total electrical effect of the nogative-sequence re- sistance in system analysis problems is obtained by in- serting the negative-sequence resistance in the negative- sequence network and solving the network in the usual manner. All three of the sequence currents are thus af- fected to some extent by a change in the negative-sequence resistance. The total electrical output of a generator, not including the shaft torque developed by negative-sequence current, is equal to the total terminal power output plus the loses in the machine. However, the megative- and zero-sequence power outputs are merely the negative of their losses. Tn other words, their losses are supplied by power flowing into the machine from the system. There- fore, the contribution of the negative- and zero-sequences to the electrical output is zero, ‘The total electrical output reduces then to that of the positive-sequence and to include the positiversequence armature-resiatance loss it is neces- sary only to use the positive-sequence internal voltage in the calculations. Or viewed differently, since there are no internal generated voltages of the negutive- or ero sequence, tho corresponding internal power rust be zero. Tm addition to this electrical output, which produces a torque tending to decelerate the rotor, there also exists the negative-cequence shaft power supplied through the rotor. Tt was shown that this power tending to decelerate the rotor is numerically equal to the negative-sequence power Machine Characteristics Chapter 6 supplied to the stator, which, in turn is equal to the loss absorbed by the negative-sequence resistance. Therefore, the total decelerating power is equal to the positive: sequence power output plus the loss in the negative: sequence resistance. ‘The assumption was made that the stator resistance tand the lowes in The magnotizing branch were neglected For greater refinements the stator resistance and the losses jn the anagnetizing branch ean be taken into consideration by substituting them in the equivalent circuit and reducing that circuit to simple series resistance and reactance, ‘wherein the resistance becomes the negative-sequenee re- sistance and the reactance the negative-sequence reactance. ‘The ratio of the negative-sequence shaft power to the loss in the negative-sequence resistance is then equal to the ratio of the power loss in $f for unit, nogative-sequence current inthe sator tory. This ratio can be obtained easily by test by mesturing the shaft torque and the negatives sequence inp when negative-sequence voltages only are applied to the stator. While this analysis has premised induction-motor con- struction, the conchsions can also be applied to eynchro- nous machines ‘Method of Test—Whilersand 2, can be determined by applying negative-sequence voltage from another source of supply to the armature, the following method has the ‘udvantage that the machine supplies its own nogative- Sequence voltage, “Two terminals of the machino under test ate short-cirouited and the machine driven at rated frequency by means of a direet-current motor. The equiv- alent circuit and veotor diagram for this connection are shown in Fig. 33. The postive-sequence power per phase at the terminals ia equal to the product of B and Jy and the cosine ofthe angle ¢. ‘This power ispostive. However, the negative-sequence poster output per phase is equal to ‘the product of By, Ze, and the cosine of the angle between Ez aud Iz, and since Jy=— Ih, and Ey= Es, the negative: e q riving WOTOR Ease + Pha) -3(P—Proe) ao whieh Poashaft input, Pyyvmitition and widage less. Fig, 35Negativesequence, fance of @ synchronousChapter 6 Machine Characteristics sequence power output ix the negative of the pos sequence power output, which, of course, must fellow singe ‘the output of the machine is sero. A negative output is equivalent to a postive input. This input is equal to 703 per phase. ‘Therefore, the pastive-sequence terminal out- put per phase is 73J3, and adding to this the copper loss due to 1, gives the total shaft. power duc to the positive sequence as 3(rul3+nF2). Now from Fig. 32(a), if tq be neglected the negative- sequence input per phase is equal to (etn f on) from which it follows that a) [As shown previously the negative-senuence shaft power per phase is equal to 72, which on substituting $ fromm (84) reduces to (r»—r.1F3. But since ry=n, the expression for the negative-sequence shaft power per phase can also be written (r—nf2. Incidentally, from this the rotor losses are equal to 2(ryy~r))F3. ‘Therefore the total shaft input into, the alternating-current, machine is equal to BfroFitnli-+-(re—n) Ei] and, since I=L, reduces to 6riT}. Including the effect of friction and windage, Posy, and calling Py the tolal input into the allemating-eurrent ruchine from the driving tool, ts (35) ol and, since Th=J where Tis the actual measured phase current, (38) ‘The foregoing noglects the effecte of eaturation, Tests on salient pole machines with and without dampers verify the fact that the loss varies ae the square of the neg: sequence currents. ‘The lose for turbine generators, on the other lund, varies as the 8 power of current. 16. Zero-Sequence Impedance ‘The zero-sequence impedance is the impedance offered to the flow of unit zere-sequence current i.e, the voltage drop across any one phase (star-connested) for unit current in each of the phases. The machine must, of course, be star-connected for otherwise no zero-zequence ourrent can flow. ‘Tho zero-sequence reactance of synchronous machines is quite variable and depends largely upon pitch and breadth factors. In general, however, the figures are much smaller than those of positive and negative sequences. The nature of the reactance is suggested by considering that, if the armature windings were infnitely distributed so that exch phase produced a sinusoidal distribution of the mmf, then the mmfs produced by the equal instantaneous currents of the three phases cancel each other and produce zero field and consequently zero reactance except for slot and 168, cend-connection fluxes. The departure from this ideal con- dition introduced by chording and the breadth of the phase belt determines the sero-sequence reactance. ‘The zero-soquence resistance is equal to, or somewhat, larger than, the positive-sequence resisiance, In general, however, it is neglected in most waleulations. Method of Test—The most. convenient method for test of zoro-sequenee impedance is to connect the three phases together, as shown in Fig. 34, with the field short Rotor at synchronous speed (or blocked) ‘ero-sequence impedence, -£ oy Fig. 4 Connection for measuring 2ero-sequence smped- cireuited. This connection insures equal distribution of current between the three phases. For this reason it is preferable to connecting the three phases in parallel. The is : Ei rerosequence impedance is then equal to Zam gy aa i dicated in the illustration, IV. PER UNIT SYSTEM "The performance of a whole line of appartus, regardless of size, can often be expressed by # single set of constants, when those constants are expressed in percentages. By this is meant that the loss will be a certain percentage of its kilowatt rating, its regulation a certain percentage of its voltage rating, ete. The advantage of this method of representation extends to a better comparison of perform- ance of machines of different rating. A 100-volt drop in & transmission line has no significance until the voltage base is given, whereas, as a percentage drop would have much significance, ‘A disadvantage ofthe percentage system is theconfusion ation of percentage quanti- ties. ‘Thus, a 20-percent current flowing through a 40- percent reactance would by simple mu Which at times is erroneously considered as 900-percent voltage drop, whereas, the correct answer is an 8-percent voltage drop. ‘The per unit system of designation is advanced as poo- seosing all the advantages of the percentage system but ‘avoids this Inst mentioned disadvantage. In this system the rating quantity is regarded as unity. Any other amount of the quantity is expressed ss fraction of the rated amount. It is the same as the percentage system except that unity is used as s base instead of 100. The foregoing,164 ‘multiplication example would in the per unit system be expressed as follows: A 0.20 per unit current flowing through a 0.40 per unit reactance produces an 0.08 per unit voltage drop, which is correct. A further advantage of the percentage and per unit systems lies in the elimination of troublesome coefficients. However, this is not an unmixed blessing as a definite dlnmavaniage ofthe use of the per unit ayer Hen in the V. POWER EXPRESSIONS Ibis frequently necessary to know the manner in whiel tho power output of a machine varios with its excitation, and internal angle. A particular application of this knowl- edge is the stability problem. Several simple cases will bbe considered, 17, Machine Connected through Reactance to In- finite Bus and also Shunt Reactance across its, ‘Terminals, Resistance of Machine Neglected ‘The schematic diagram for this ease is shown in Fig. 35(a), which also shows the significance of the various syinbols to be used in this discussion, ‘The reactances 2s, 4, and the one indicated by the dotted lines represent the branches ofan equivalent x circuit, for which the resistance components aro neglected. For the purposes of determin- ing the power output of the generator the reactance shown dotted ean be neglected. ‘The vector diagram which ap- plies is Fig. 85(b). ‘The total machine current is equal to i, XsOxy 3, xe Fig. 35~—Machine connected to infinite bus through a re which from the internal and external currents one can. obtain And inserting the equivalents of i, and é a jack ‘The symbol caret over « quantity iadioates a phasor quantity. Machine Characteristics Chapter 6 @n) metals jtd Buk cos O—jE sin 8 ‘Upon substituting 2 and 27 in equation (87), there results, ct) etal +IE(E 00 0—JB sin 8} 8) Rquating reals ), ,B sine ate ay 2 sin 6 aaet netstat a ‘And equating imaginaries (40) ‘The power output, P, is equal to the sum of the products of the in-phase components of armature current and terminal voltage, namely Paiglea—zsia)+ialtaie) tale (Sq tail = [eat (ee 20)ta]ig an ‘The power is then obtained by calculating i, and ia from (89) and (40) and inserting into (41). 1f 2 and e are ex- preseod in terms of rma volte to neutral and resctances in ‘ohms per phase, then the above expression gives the power in watts per phase; but if the emis are expressed in terms of the phase-to-phase volts the expression gives total power. ‘On the other hand, if all quantities are expressed in put then the power is leo expressed ia pu. where unity is ‘equal to the kva eating of the machine. If eq! rather than, ‘cis known then ea should be replaced by e,” and za by 14! Wherever they appear in Eqs. (40) and (41) For the special case of a machine with eylindtical rotor in which 2,24, the expression reduces immediataly to a2) Beret Another interesting special ease is that for which the sunt resctance i not present or ==>. Thea ca B cos 6) B sin 9 aed ed «) “Chapter 6 Oe aay chine — O-percent In Fig. 36 is shown a power-angle diagram of a salient pole machine whose excitation is determined by loading at full kva at 80-poroont power factor. ‘An expression frequently used to determine the maxi- ‘mut pull-out of turbine generators is the following in wre (rat in kv) Pull-out in kav (rating of generator in kva) where OC is the field eurrent for the particular operating eandition and OD is the field current: for the rated-current zero-power factor curve for zero terminal voltage (soe Fig, 17). ‘This expression is based upon the maintenance of rated terminal voltage up to the point of pull-out. At pull-out the angle 8 of Fig, 15 is equal to 90 degrees. Since the extant of saturation is measured by the voltage behind, the Potier reactance drop, it can be seen from Fig. 15 that, for 8 equal to 90 degroes this voltage is less than rated voltage, and that therefore little saturation is present. From Fa. (44) since r=, and = 90 degrees, the pull-out, ie ttf, is“, But eis proportional to OC on the air-gap line and 2 is likewise proportional to OD on the airgap line. Examination of Bq, (4) shows that oven if the excia- tion is sero (¢4=0) the power-angle curve is not equal to zor, but equal to 22272 ‘hia rl fom tho fects of ealieny. Note that it diaappoars for uniform airgap machines for which 74. Advantage is some- times taken of this relation in the ease of synchronous oniencers to obtain a somowhetgrenter capability i the Teading (under-excited) kva range. With some excitation systems (oce Chop. 7, Exeitation System) iis posible to shitain negative excitation. The exeation voltage un Eq. is) con be somewhat negative without producing an unstable power-angle diagram. By this device the leading iva range ean be increas as nich as 16 of 20 pereen. 18, Inclusion of Machine Reslstance or External sistance If the machine is connected to an infinite bus through resistance and reactance cireuit, the external resistance and reactance can be lumped with the internal resistance ‘and reactance and the following analysis used. ‘The vector diagram for thie ease is ahown in Fig. 37 for which esi Brig agg 5) Machine Characteristics Pig. 37-—Vector diagram of sallent-pole machine Including ‘effect of serio reelgeance, 64008 Dt rig ie eam a) From (48) for He sin 4rd an Substituting (47) Sato (46) 2.008 04 Ee, sin 04 Mist aate ae ana from whieh ts retamreysin D—zyec 6] 8 Tnglermiaraeene] on and substituting in (47) . : teas [rteee ain ¢—re, 008 ‘] (as) ‘The power output, P, is equal to the sum of the products of the in-phase compunents of 7 and e, oF Pigs c08 946 sin (0) Upon substituting (48) and (49) hie redces to tig [er cove +x sin 25% 20-re] oD ‘The power input into the machine is equal to P plus ‘The expression for this quantity does not simplify and it is better to calculate it through the intermediate step of evalusting ni, which is equal to r++). ‘The foregoing expressions apply to the steady-state con~ ditions. In stability problems it is necessary to determine the average power from instant to instant. In general for this purpose it is permissible to neglect both the uniditec- tional component of etrrenta and the subtransient com- ponentof the alternating current, leaving only the transient component. These latter are determined by the instanta- neous value of ex’. It follows then that the power expres- ions are simply those derived for the steady-state condi with eg replaced by ed’ and 24 by ze’ VI. EFFECT OF CHANGE IN EXCITATION Field forcing in certain industrial applications and con- siderations of system stability require that the voltage fnerease in response to a sudden need. This increase is brought sbout automatically cither by means of the same166 control that produced the increase in load or through the ‘use of a voltage regulator. It is necessary, therefore, to be able to predetermine the effect of an increase in exciter voltage upon the output of the synchronous machine. Tn general, significant changes in exciter voltage never require Jess than about one-tenth of second to bring about the change. By the time thie effect has beon felt through the synchronous machine, which hus a time constant of about ‘4 second, it will be found that the result is always slow when compared to the subtransient end unidirectional components of the transients associated with the change. In other words, variations in exciter voltage are reflected only in the transient components. Asan example, suppose it is desired to calculate the armature current of a machine for a three-phase short-cireuit while it is operating at no load with # voltage regulator set for rated voltage. ‘Immediately after the inception of the short circuit there ie a slight lag in the regulator until its contacts and relays lose. ‘The exciter voltage (and voltage across the field of ‘the main machine) then rises as shown in the upper curve of Fig. 38. ‘The bottom curve refers to the armature eur EXCITER VOLTAGE afte mw nesuLator Tne ARMATURE CURRENT Tne Fig. 38--Illusteation showing relative importance of diferent components of armature short-circuit current and response ‘of transient component to the excicer voltage. rent, the dotted line showing the natuze of transient com- ponent. if there were no regulator, the exciter voltage remaining constant. The line immediately above shows how the transient component changes as a result of the change in exciter voltage. To approximately the same scale, the cross-hatched arca chows the ineroment in cur rent caused by subtransient effects. The blackened arca, shows how the unidirectional component would contribute its effect. This component is quite variable and for 9 short-cirevit on the line might be entirely completed i cycle or less. Tu any event regardless of its magnitude it, can be merely added to the transient and subtransiont, component. It is independent of the exciter voltage. 19, Fundamental Equation ‘Boing restricted to the transient component, the effect of exciter response can then be defined entirely by effects in Machine Characteristics Chapter 6 the field citeuit. ‘The beauty of the per unit system ia exemplified in the analysis of this problem. In p.u. the differential equation for the field eircuit takes the following form eg a In this equation eg represents the exciter voltage or the voltage aeroos the field if there is no externsl field resistor the field circuit. ‘The unit of eis that voltage required to circulate auch field current as to produce rated voltage atmo load on the airgap line of the machine. ‘The term es is the eynchroncus internal voltage necessary to produce the instantaneous valoe of armatnre current for the given armature circuit egardies of what it may be. Ita unit ie rated voltage. It is synonymous with field current when lust field curzont is that field eurrent necessary to produce ated voltage at n0 Toad on the sir-eap line. Tt willbe seen then that the ure of ¢ ie merely @ conveniont way of specifying tho instantaneous field current during the tran- sient conditions; itis the field current necessary to produce the armature eurrent existent at that instant. As shown previously, es proportional fo the lux linkages with the field winding. Tis the quantity that, during the transition peviod from one circuit condition to another, remains con ‘stant. ‘The foregoing equation has its counterpart in the sore familiar forms erent (52) (53) ai eT LG oo) To familiarize the reader with (62), suppose that normal cexciter voltage is suddenly applied to the Beld winding at no load. Since the armature is open-cireuited ed’ and 6 can be written are equal and the equati dea . erat Tage (65) When steady-state conditions are finally attained is equal to zero and ¢,-¢y. This states that since ¢=10, 4 must also equal 1.0, that is, the excitation is equal to the normal no-load voltage. It will attain this value ex- ponentially with a time constant To Another example. Suppose the synchronous machine to be shori-cireuited from no-load and 10 be operating without » regulator. At any instant the armature eur ront, i, is equal to t//ay!. But since «4, which can be regarded as the instantaneous field current required to produce i, is equal to tq 4, then eliminating + between these equations, (56) an Fie ) takes the form ‘Then equation (¢ ded ee TageChapter 6 or if it is to be expressed in terms of armature current ceed 7 Teal Posie (ery "The sustained magnitude of #, is then ‘The initial magnitude of §, since eq’ remains constant during the transition and is initially equal to 1, is 1 ‘Tho homogeneous equation for (67) is sett Omaait Totals omit ri Moi (58) ‘Thus é changes from +, to 1 exponentially with a time constant equal to “1%, Im all problems involving a transition from one eireuit condition to another the one quantity (when subtransient effects are noglected and the time constant in the quadra- ‘ture axis is zero) that remains constant within the machine is the flux linkages with the feld winding, which in turn is ceflected in the quantity eg!. It is necessary, therefore, to caloulate ex’ for the eitenit condition preceding the transition. All the discussion of the following eases as ssumes that this point is understood and that ¢ is known for the beginning of the transient period. Several eases will now be discussed, 20. Three-phase Short Circuit of Unsaturated Ma- chine ‘The problem is to determine the transient component of shori-sireuit current in response to the exciter voltage given in Fig. 39. ‘This is most quickly and conveniently found by a graphical method, which, for want of better Fig. 39—IMlustration of “Follow-up Method." Machine Characteristics 367 ‘name, has been called by the author “the follow-up meth- od.” Tt in a method that can be applied to any problem involving a simple time constant, "To demonstrate the method, consider a simple resist- ance-inductance cireuit to which the voltaye, B, is applied. Jet the differential equation for this eireuit be ere ERI +h (59) where the symbols have the customary significance, Divitisg through by there resulta E_,,ba RO Ra (0) resins teat ine eats tha rian il be fant by, og E if eG (61) ta teen # ete septate crt that approaches for the instantaneous value of B. T is the instantaneous magnitude of current. If the current at any instant is platted by the point @ (Hig. 40) and the corresponding value of 3 for that instant is plotted ax the point b (Fig. 40) displaced hotizontally by a time , Toe Fig. 40—Construction derivation of “Follow-up Method.” then the vertical distance between @ and b gives the mu- ‘erator of (61) and the horizontal distance the denomina- a tor, ‘The slope of the line between o and b is equal to Sr. If an interval of time Af is chosen following the instant under consideration and £ is assumed constant over the interval then the change in J during the interval, AT, is aunt ae Te nal la of arent ote terval fs then given hy the point 12 at an instant Al later i ‘thon plotted and the line drawn from ¢ then the value for AI for the second interval ig obtained. Following such Procedure it is possible to construct the complete curve for . The construction ig illustrated in Fig. 39, in which168 a sve markt i nantes magne of tin tne 0. Bot lcd he ight tine 2 Tat etl alot ating, Dieta uneins Kea ing Seber oe critic emi rs ge cent ah Br pad oT othe tance by Wich he steady-state curve which J tends to approach, is offset ovine poo in and, Theda extn eg fae ey 6 the ar wae ee ln nal Sete ops Boca ESERIES eae Et Sepa os anl cas then by sons using 7 eenzaites Tit 2) Dividing through by za fener (3) ‘The construction dictated by this equation and the follow-up method ia shown in Fig. 4 $s plouad ago Tine st component of short-clrcult current, a8 ‘influenced by excitation. nt ae eng pled an neva 2 enee zero. ‘The initial value of # is determined through 62’ wih an 10a ¢=0. ‘This makes te inti snot ‘bs from refer Machine Characteristics Chapter 6 i=, Starting {rom this value the actual magnitude of 1s Sheined as funeton of tine 21. Unsaturated Machine Connected toInfinite Bus As stated previously the subtransient and unidirectional components of current are not of importanee in the sta- bility problem. For this application it is desirable to de- termine how ec’ varies as this influences the power output Of the machine and in turn dictates the degree of accelera- tion or deceleration of the rotor. ‘The circuit shown in Fig. 35(a) ie typical of a setup that might be used for an analytical study to determine the effect of exciter response im increasing stability limits. Another caso of considerable importance is the action of a generator when # heavy load, such aa a large induction motor, is connected suddenly sueroas ite terminals or across the line to which is is con- nected. In starting the motor the line voltage may drop fan excessive amount, The problem might be to determine the amount to which this condition could be ameliorated by an appropriate excitation aystem. Since reactive kva is more important than the real power in determining regulation, the motor can be represented as a reautor and the cireuit in Fig. 35(a) utilized. Having determined the manner in which é’ varies, the power in the case of the stability problem and the terminal voltage (cq'—34 the case of the voltage problem, can be ealeulated easily. Equation (62) must be used again to determine the manner in which ey’ varies in response to changes in exeiter voltage ‘and phase position of the rotor with respect to the infinite bbus, ‘The instantaneous armature current can be found in terms of the rotor angle @ and e' by replacing eg and za of Bay, (40) by ea! and 24, respectively, giving t2Jes!— 28 008 8 fattest eee ‘The synchronous internal voltage, ea, is equal at any instant to wn eam ea' +t (tae ha (65) and upon substituting (64) cused (2424)! (66) wie 08 Om ea! + aie 7) yn teet (6s) ‘The time constant Ty’ constant. the short-clreuit transient timeChapter 6 If @ were constant or if its motion as a function of time were known then the whole left-hand side could be plotted (displaced by the time 7.) and treated by the follow-up method as the quantity that ea’ tends to approach. Un- fortunately @ is not in general known beforehand, and itis necessary to calculate @simultancously in small increments in a simultaneous solution of eg’ and 8. ‘The magnitude of @ is determined by the electro-mechanical considerations discussed in the chapter dealing with System Stability. In solving for ey’ a progressive plot of the left-hand side ean be made or (67) can be transformed to the following form ded AP Te, | edee~2i) ee tna tetas ore] ‘nd the increment ealeulated from the equation yates, ed Ht ae 70) A shunt resistance-reactance load such as an induetion motor is not much more difficult to solve numerically but the expressions become too involved for analytical solution. It is necessary only to calculate ig in terme of ee’ and @ juat ‘as was dono before and then follow the same steps as used for the resetanee load. 22, Unsaturated Machine Connected to Resistance Reactance Lond A case not too laborious to carry through analytically is that for which a resistance-reactance load is suddenly applied to a synchronous machine. Let rex and ex be the external resistance and resetance. The addition of a sub- script (to machine constants indieates the addition of rex (F z.as to the respective quantity. The equations of Seo. 17 then apply to this ease, ie. in the equations is made equal to zero and zq replaced by za, ete. Following the same procedure as previously, there re- wults from Bq, (48) when ey and xq are replaced by eu’ and zis and ¢, is equal to zero, NOTE F tide ‘The field current or ite equivalent, the synchronous in- ternal voltage, is then eam ed + (Qu 2isdia am me eum aho) ged ESR om Substituting this expression in (52) there results that 7) (4) in which Tee 7) Machine Characteristics 169 From this point the follow-up method can be used as before. After e,’ is determined sa a function of time any ‘other quantity such as terminal voltage ean be obtained readily. 23. Saturation In analysing transient phenomenon of machines in the unsaturated condition, the theory was built around the concept of the transient internal voltage, ey’, a quantity evaluated by using the transient reactance, ‘zi’. In the presence of saturation it was found that for steady-stato conditions by the introduction of the Potier reactance, 2, (seo See. %) the proper regulation was obtained at full Toad zero power-factor. The use of z» and e, also resulted in satisfactory regulation for other power-factors. In extend- ing the analysis into the realm of transient phenomenon, @ will continue to be used as a base from which to intro duce additional mmf into the field eireuit to take care of saturation effects. The treatment will follow quite closely the same assumptions as were used in determining the steady-state regulation according to the Two-Reaction Potier Voltage method of Sec. 3(4), With this assumption the fundamental Eq. (52) for the field circuit, becomes eemecb(o due toe) +758 (7) As before ea represents, neglecting saturation, the voage behind the synchronous reactance of the machine or what is equivalent the field current roquired to produce the fnstantancous ¢¢, including the demagnetizing fect of the instantaneous armature current. The total eld eur rent ie obtained by adding ¢ toe. Income cases it iefound simpler to convert all of the right hand side to the single variable ¢, but in others it is simpler to retain the variable in the form of ea. Two applications of this equation will be dscused. ‘Machine Connected to Infinite Bus—The cirevit shown in Fig. 35(a) is the one under discussion and for which Eq (6) applies for the unsaturated condition, ‘This ‘equation can be expanded to include saturation, in accord- fice with Eq, (9), tothe following _fetrerstte, —_alza~as) OC gaetiintete ateetale mF Hle due oop) +N ™ ‘This can be converted to ded _ex~(sduetoes) , lz ag) a The (@etet rete taar) Td Th in which defined from Eq. (68). Before (78) can be used it will be necessary to determine ein terms of ed. ‘The components of current, fy and i, can be determined from (0) and (10) by replacing e by a and zy 24 2B sin 8 fra teatige ™ 80)170 ‘The dircet-exis component of ¢y is equal to (ea —2))ia Crab tet gata b tie and the quadratureaxis component of e is frg™ q— Fig sb ing wactavtsteaz “The amplitude of eis then equal to eevee (82) (83) While this quantity does not simplify grestly, it does nat. -xppear so formidable after numerical values are inserted. ¢p,can thus be caleulated for any instantaneous value of eq! and the s corresponding thereto substituted in Bq. (78), juatiou (78) provides x means for computing increments ‘of change in efor use in step-by-step solution. Thue (et) As s becomes small and saturation effects disappear, the solution relapses into the sare type as used when satura tion is negligible (Eq. 66), for which the follow-up method is frequently applicable. ‘The relations just developed are useful in estimating the extent, to which ea’ varies in system stability problems, Fig. 42 shows the results of ealeulations on a system in which a generator is comected to a lange network, repre sented as an infinite bus, through a reactance equal to 0.6. $ east s ue, ge 3 10 0 wei 10 008 ue 108 008 lo Mai ae 8s Ba Fig, 42—Bflect of rate of response upon ey’ ae a Tine-to-line {fanic represented by the three-phase shunt load j0.2 ie applied to generator which had been operating at 90 percent power- factor. 20 percent of air-gap mmf required far ison st rated voltage. A line-to-tine fault is assumed applied to the connecting ‘transmission line on the high tension bus at the generating. end which is equivalent to a three-phase short eireuit through a reactance of 70.2 ohms, ‘The curves justily the Machine Characteristics Chapter 6 asoumption that is usually made in stability studies that where quick responze excitation ia installed, ¢4! may be regarded us constant. Machine Connected to Resistance-Reactance Load—This ease is the same as that considered in See. 22, except that saturation effects are to be included. Upon incliding the saturation term into Eq. (74) there results that (85) [ecto due to. free +2 in wien patitact tin» Faint Te is wall to reel! agin that this analysis neglets sub- sransint effects and assumes that the ime constant in the Gqudratte wuss aero. Trin Fas. (48) snd (4) sade aul toate, ex i Teplacel by ean the coreesponding thhanges in eactatceasoviate with ei are made, andi tviition the subseripts are changed to indicate total re aotances, Then (30) tel (1) Bat ° cs rei (88) Laat ‘ ‘The total current is then he (69) ‘The voltage & entVveTR VGA oy ‘Upon substituting ¢’ from (90) into (83) and using (86) also there results that VETOED, VGA, et (a due to e,) dey at Tal oD As can be seen from Fig. 48 the solution of this equation lends itself well to the follow-up method. On the right hand side the assumed exciter response curve, ex is plotted. fas a function of time. Multiplying this quantity by the ‘coefficient of ¢, the term éyx is obtained. This is the value tends to attain if there were no saturation effects. As the follow-up method, the zero of time from which the instantaneous curve of ¢, is drawn, is displaced te the left an amount Ty’ minus bslf the interval of time chosen in the step-by-step solution. Along the ordinate of ¢ a curve 4s; equal to the second term is plotted in which s is obtained from the ne-load saturation curve shown in (b). For any. instantaneous value of ¢, ais plotted downward from tu. 1 the construction progresses. So starting from the initial value of ¢,, of which more will be said later, a construction, line is dravwn to 4 point for which «, was the value corre- sponding Co the initial value of ¢,, Bor the second intervalChapter 6 bel i i ‘SATURATION CURVE (OF MACHINE phat) ras seo oboe Farge INTERVAL? 5| ee sarge A THE Fig. 42—Graphical determination of terminal voltage as polyphase series resistances, pled. ‘in taken for the value of ¢, at the end of the first interval or, to be slightly more accurate for an estimated average value of ¢ for the second interval. And s0 the construction proceeds By the same reasoning whereby ¢ was obtained in Eq. (90) the terminal voltage e. can likewise be obtained, giving en iV Batra VEC at rad, ada tand substituting ey’ from (90) oye? @ ‘This permits of the calculation of e from e, after the conti ar bes competed ‘During the nnston fro ona operating condition to ‘the next, only ¢s' remains constant; ¢, changes. It is es sential therefore that ea’ be computed for the initial operat Soe condom,‘ conventional conetton shown a ip Ste) con tw unl Ths dre the lal vate 615 forthe nw operating codon fa wih thea value of e, ean be computed by Eq. (90). Common cases for which these calculations apply are the determination organ for ota suddenly apd a gure Vnttoe in wc he can fe ae the Machine Characteristics +H erro (ervecTon pacKan FOR MACHINE ALONE. m™. As rom nrenvan #2 i 1 “a FOR WTERVAL OL sudden disconnection of a loaded generator from the bus throwing ite load upon the remaining units or the starting of an induction motor by direct connection to a generator. For the latter case, if the expacity of the induction motor isa significant fraction of the kva of the generator, a severe drop in voltage results. Thus » 500-hp motor throxtn on 8 3300-kva generator produces an instantaneous drop in voltage of the order of 13 percent. The effective impedance of the induction motor varies with slip and to be rigorous this variation should be taken into consideration. It ie ‘usually sufficiently accurate to use the blocked rotor re- aactance for the motor impedance up to the speed corres- ponding tomaximum torque in calculating the factor which determines ¢,. in terms of ¢. Beyond the slip correspond ing to maximum torque, the effective impedance varies rapidly to the running impodance. Simultaneously with the increase in impedance the lagging kva likewise drops off which results in a considerable rise in voltage, ‘This effect is clearly shown in Fig. 44 taken from some tests made by Anderson and Montelth® As running speed is approached the generator voltage rise, the excitation be- ing too high for the particular loading. To form s better idea of the magnitudes involved in such calculations, Fig, 45 shows curves of terminal voltage as an induction motor equal in horse power to 20 percent of the kva of a generator issuddenly throwa upon an unloaded generator for differ-a2 3800 1200, 2402 3100 ee] oo ccacoaron woxtaet]igg gg + g F 2300 £00] Spee Bog noo Ser erie is to za 30 ao OE ig. 4—Performance of 3888 kva, 0.6 power-factor, 3600 rpm, LT short-circuit ratio generator as a single 800-hp induction motor pump is started. Induction-motor starting torque equal to full-load torque and pull-out torque equal to 28 fll-load torque. Full lines represent operation with fixed ‘excitation and dotted lines under regulator conteol. 2 | TT foft tt >| Z RESPONSE big —t | i eT Ba —s 2 pS | a0] —|— 7 Of OF 08 Os 10 ‘Time ~seconos Fig. 45—Terminal voltage of 2 500 ka, s0-percent power- factor engine-type generator (24=1.16, x= 0.39, 1/030, 413 peteent scarution) ss % 100-np induction motor Is con: nected, cent rates of response of the exciter. Ordinarily one is primarily interested in the minimum voltage attained dur- ‘ing the aceclorating period and co the ealculations have been carried out to only 1.0 second. The curves show con- ditions for constant excitation and for exciters with 0.5 and 1.0 ratios, respectively. 24. Drop in Terminal Voltage with Suddenly- Applied Loads ‘When a relatively large motor is eonneeted to 8 gener ator, the terminal voltage may decrease to suck an extent as to cause undervoltage release devices to operate or to stall the motor. ‘This situation arises particulssly in eon nection with the starting of large motors on power-house auxiliary generators, ‘The best single eritetion to describe this effect when the generator is equipped rvith a regulator to control the oscitation is the maxiraum drop. ‘The pre- vious section describes a method whereby this quantity, ‘can be ealeulated. However, the problem arises eo fre- quently that Harder and Cheek#? havo analyzed the Machine Characteristice Chapter 6 problem generally and have plotted the results ia curve form, ‘The analysis has been earried out for both self-exeited ‘and separately-excited exciters. The results for the former are plotted in Fig. 46, and for the latter in Fig. 47. These ‘curves are plotted in terms of the four parameters: (1) magnitude of load change (2) Nr (3) Tle, and (4) rate cof exciter response, R.. ‘The response is defined in the chapter on Excitation Systems. It is shown by Harder and Cheek® that variations in 24, saturation factor of the generator and power factor between zero and 6l) percent, have little effect upon the maximum drop. ‘The assumed valuo of 24 for theso calculations was 120 poreent. An accurate figure for maximum voltage drop can be obtained for values of zg other than 120 percent by first expressing reactances and the applied Iced on a new kva base, such ‘that zy on the new base is 120 percent, and then applying. the curves. For example, suppose a load of 1500 kva (ox pressed at full voltage) of low power factor is to be applied. to a 3000-kva generator having 30-pereent transient re- faclance and 1i(l-percent synchronous reactance. Suppose ‘that the generator time constant is 4.0 seconds and the cexciter has nominal response of 1.0. ‘To determine the drop, express the transient reactanee and the applied load ‘on the kilovoli~ampere hase npon which za is 120 percent. ‘The base in this case will be 3000 120/150— 2400 kya. (On this base the transient reactance 24’ is 30% 2100/3000 ~ 24 percent, and the applied load is 1500/2400=62.5 per- ent. If the exeitor ie eof excited then from the curvos of Fig. 46, the maximum voltage drop is 15 pereont for 62. percont load applied to generator having 24-porvent tran- sient reactance, a time constant of 4.0 seconds, and an exciter of 1.0 nominal response. This same maximum drop would be obtained with the machine and load under consideration, ‘The initial load on 2 generator influences the voltage drop when additional load is suddenly applied. As shown, in Fig. 48, a static or constsnt-impedance initial toad re= duces the voltage drop caused by suddenly applied loud. However, a load that draws additional current as voltage decreases may increase the voltage drop. Such loads will, be referred to.as “dynamic” loads. For example, a running. induction motor may drop slightly in speed during the voltage dip s0 that it aetually draws an ineroased current and thereby incrosses the maximum voltage drop. The dynamie initial load curvo of Fig. 48 is based on an initial load that draws constant kilowatts and power factor as the voltage varies. VII. CONSTANTS FOR USE IN STABILITY PROBLEMS ‘The stability problem involves the study of the electno- mechanical oscillations inherent in power systems. A fundamental factor in this problem is the munner in whic the power output of the generator varice as the position of its rotor changes with respeet to some reference voltage, ‘The natural period of power systems is about one second, Because of the series resistance external to the machine, ‘the time constant of the unidirvetional component of arm auuze currout is usually so small as to be ncgligible inChapter 6 Machine Characteristics 173. Fig. 46—Maximum voltage drop of a synchronous machine WITH SELF-EXCITED EXCITER as affected by (a) magnl« tude of toad change, (b) xs as, (6) T'ss and (d) rate of exciter response. x«' on curves refer to saturated or rated-roleage (1.07 Wai 111.20; 1,70.75; no-load saturation curve/air gap Tine normal ‘second; added load ls constant Impedance of 0.39 pl initial load zero.Mackine Characteristics Fig. 47—Maximum voltage drop of a synchronous machine WITH SEPARATELY-EXCITED EXCITER as affected by (a) magnitude of load change, (b) x's m, (¢) Ty) and 1d) rate of exciter response.Chapter 6 : } | got | | oma 8 £ Tome g i, os INITIAL, LORD-PER CENT ig. Effect of type (whether dynamic or static) and inital, oad, assumed at 0.80 power factor, wpon the mazimum volt= ‘ge drop when 100-percent low-power-factor load is suddenly ‘applied to an a-e generator. comparison with this natural period. The subtransient component is likewise eo small that its effects can be noglected. ‘There remains then only the transient com- ponents, those components associated with the time cou- stants of the field winding, that are important, 25. Representation of Machine ‘The transient stability problem is primarily concerned ith the power-angle relations during aystem swings fol- lowing a disturbance. Because of the dissymmetry of the ‘two axos, it is nccoasary theoretically to take this diaaym- metry into consideration, However, in most cases an impedance is in series externally to the machine 60 that, the difference in reactances in the two axes becomes a smaller proportion of the total reactance. The results of calculations presented in Chap. 13 show that for most practical purposes it is sufficiently accurate to represent ‘he unsymmetrical machine with a symmetrical machine hhaving the some ze! In spite of the close agreement of salient-pole with evlindrical-rotor results, a few cases arise for which it is necessary to use salient:pole theory. Relations for calcu- lating the power output have been given in Sees, 16 and 17 and for computing the change in internal voltages in Sec. 2a). It is shown in the latter section that if the exeiter is of the quick-response type, the eollage ¢,' ean, for all, practical purposes, be regarded as constant, Methods for ‘he inclusion of these factors into the stability ealeulations Ihave also been treated in Chap. 13. A knowledge of the inertia constant, H, is a requisite for the determination of the acceleration and deceleration of ‘the rotor. Tt represents the stored energy per kva and ex be computed from the moment of inertia and speed hy the following expression 0.2317 RY(epm)*10-+ eva Where H~Tnertis constant in kw-soe. por kva, WR*— Moment of inertia in Ib-ft, (93) Machine Characteristics We Further consideration of this constant is given in Part ‘XIII of this chapter. 26. Network Calculator Studies For most problems the synchronous machine ean be represented by its transient reactance and a voltage equal ‘to that behind transient reaetanee. For the rare euse for which calient-pole theory is required, the following pro- cedure can be followed. It is impossible to set up the two reactances in the two axes by a single reactor, but if the reactance, Zq is usod and a nev voltage, equ introduced as representing the internal voltage, both position of the rotor and the variations in ey’ ean be carried through quite ply Fig. 49 shows a vector diagram similar to Fig. 14 in which eqa is included. ‘This voltage is Inid off along es and ea’ and terminates at the point a. ‘The reading of power tat eqa is the same as the actual output of the machine, As tho ‘exciter voltage changes e’ and ¢ga likewise change. Fig. 49_-Construction of au: for network calculator studies, ‘The incremental changes in ¢gs can be obtained as follows. From Fig. 49 itis evident that at any instant eats + (a oy From Eq, (62) det ot lane) and bed = = Fh (eget (05) ae 7g where Acs’ is the inerement of «,’ in the increment of time ‘4 From Fig. 49 there results also that cam tqat (aaa (96) so that bode gt maf nee nde] ae In network calculator studies of system stability, ¢, qa, and ig are known at.any instant. From Bq. (94) it is, evident that the increment of ¢g is equal to the increment ined. Thos dia]at (97) (98)176 ‘This method can be applied regardless of the number of ‘machines involved in the study. ‘To obtain the initial value of eqs, caleutate ey! from the steady-state conditions before the disturbance. is, the quantity whieh remains constant during the instant representing the change from one operating condition to nother. The proper és is obtained by changing the magnitude of ge until Ea. (04) is satisfied. ‘To include the effect of saturation, break the reactance Zq which represents the machine, into two components zp and (z,—Z,), the latter being next to the voltage &a. ‘The voltage at the junction of these two reactances is gp, the voltage tehind ,. The effect of saturation will bo included by adding the saturation factor ¢ taken from the no-load saturation curve (see Fig. 17) for ép, to the exeita- tion obtained by neglecting saturation. This corresponds. to method (d) of See. 8 for steady-state conditions. Eq. (98) then becomes (99), 27, Armature Resistance For most stability studies the loss associated with the resistance of the armature is 90 email aa to be neglixible. ‘The exception to this rule is the case for which a faule occurs near the terminals of a generator. “The loses in an ae generator during » three-phase short circuit can be large enough to affect significantly the rate at which the rotor changes angular position. ‘This is of particular importance for stability studies. ‘Two of the most important factors determining this effet are the locs- tion of the fault and the value of the negative-sequence resistance. The latter is dificult of evaluation particulanly for turbo-generators—the type of machine io which the effect is greatest. One must rely almost entirely upon eal- culations, whieh are extremely complicated. For a board studies of system stability i is convenient to represent the zachine losses by means of a resistance placed in series in thearmature. The value of this resistance should be chosen 0 that its loss, with the reactaneo of the machine repre- sented by 2, be equivelent to that of the machine under sctual conditions, An approsimate evaluation of this equivalent resistance will be developed for a turbo- generator Lt the initial value of the subtransient component of short-circuit current be designated, *”. The components of the unidirectional curzent have n’maxinum value 723" and ure related in the three phases in a manner as discussed jnSeo. 8. ‘Tho cum of the unidirectional components in all throe phases produce an eswentilly sinusoidal wave of minh that is stationary with respect to the armature. This sta- tionary mmf develops a flux that in turn generates curcents having a froqueney of 60 eps in the rotor. This effect is similar to that produced by negative-sequonce currents ia the armature except that the latter produce a sinusoidal anmf wave thal rotates at a speed corresponding to 60 cps in a direction opposite to the rotation of the shaft and ultimately generates citulating eurents in the rotor hay ing a frequeney of 120 eps. ‘The magnitudes of the mint? ‘waves in the two eases are equal for the same eret values Machine Characteristies Chapter 6 of unidirectional and negative-sequence currents, The crest ‘alu of ho mgaivessuete suet t/a he coat valu fois 2 In the case of negative-sequence currents, part of the loss is supplied by the shaft and part is supplied through the armature, The loss associated with the circulating ccurronte in tho rotor as developed in Soetion 18 is apprez- ‘imately equal to 2(r—r,)i8, Assuming for the moment that the loss varies as the square of the current and neglecting the differences due to the frequencies in the two cases, the lass for the unidirectional components of current is Actually, however, the loss varies more nearly as the L8 power of the current so that the expression becomes 2Wr=n) Gay Now considering the effect of frequencies Since the depth of current penetration varies inversely as. the ‘square root of the frequency, the resistance varies directly as the square root of the frequency. ‘The loss for the uni- directional component is then Anon) VAn=n) vated ye Cer aon 36, 7 18 sq 1 16 2 { 14 A 2 i t i, ia Be, Tos a t AG Loss 05 0s OS TIME IN SECONDS Big. 80—Development of fa of turbo-tenerator for the dition of @ three-phase short circuit serosa the termi ‘the machine for various duration of the short clrcult.Chapter 6 aN 5 sak L 1 Necnoensen ES é Ato Cr = S SAE Foe — stn oauens=-—] > — sauenr rue TT d I ‘ne m secowns 2 8 : ae Bod eer pak t : Fos} WITH DAMPERS. 8 2 poe: ( a er a rr ‘TIME. N SECONDS o Fig, 51—Typleal equivalent resistance, fay for different types of machines (6) for three-phase short eireult across the terminals used (©) for three-phase short cirruit crews the terminals of & rerios: connected transformer of 10 poroint impedance. ‘Since the unidirectional current decreases exponentially with a time constant 7, the loss as «function of ime Volo nye = ee mn addition to the losses assoriated with the unidirec tional curent, the load losses as reflected by mean also be ignifeant for a three-phase fault weross the terminals Neglecting the sub-transient component, the a-c com- ponent of short-cireuit for a three-phase short circuit from nce (102) (103) Mackine Characteristics uit ‘Tho loss associated with this current is [G3 To form an idea of the order of magnitudes of these (108) T,=0.08. f0.8, ‘The reaults of the caleulations are shown in Big, 60. ‘The upper dashed curve is the loss associated with the uni directional component and the lower dashed curve the load Josses. The full line represents the total losses. The cut- rent flowing in the generator as represented on the board ig constant and equal wo ‘The equivalent resistance, ‘rm, to be inserted in sexes with 4! must be such that the Jiograted lass over any interval must be the same as that in Fig. 50. The dotedash curve in Fig. 50 gives the values Of fq Obtained by this method. Figure 51 gives similar values of ra for other types of machines. ‘The ourves in Fig, 51(a) wero caleulated for short cireuite at, the terminals of tho machines, those in Fig. 51(b) are for three-phase short circuits across the tertainals of a transformer connested in series with the machine ‘YIIL. UNBALANCED SHORT CIRCUITS ON MACHINES WITHOUT DAMPER WINDINGS Because of the dissymmetry of salient-pole machines without dumper windings, the armature currents at times of three-phase short-eirenits, as shown in See. 12, contain second-harmonic components. For unsymmetrical short- cireuits, such as from terminal-te-terminal, the wave forms of currents and voltages become even more complex. Both odd and even harmonies are present. 28. Terminal-to-Terminal Short Circuit In particular consider a colent-pole machine in which saturation is neglected and which ie operating at no load fo whieh a short-cireuit is suddenly applied across two terminals. ‘The shortcieuit current in these phases is then V3n sin 1+ Ge in which fy indicates the phase position during the eyele at which the short-circuit occurred Tt will he observed that this ean be resolved into two components (105) VI sin (2th) GPa)+ (=n) cos 2 afro) V3 Losin by BITE Hn) cos 2 Balto) ‘The first component is shown in Fig. 52(9) for a typical 1achine and consists of odd harmonies only. ‘The second First (200) Second: «207)mee AT 4 Fig, 52—Armature current and field current in a synchronous machine when terminal-to-terminal there cictuit io oud” deny applied. 30 ett R075 x 4-90" component is shown in Fig. 52(b) for ¢o= +90" and con- sists of even harmonies only. ‘The latter component is de- pendent upon the instant during the eyele at which the short-circuit oecurs and may vary anywhere between the values given and the nogative of thosevalues in accordance with the coefficient, sin de, Figure 52(¢) gives the total current, the sum of Figs. 52(a) and 52(b) ‘The units chosen are the p-u. in which for the machine operating at no-load at rated cireuit voltage J, would be equal to 1.0 and in this ease the current jis given in terms, of crest magnitude of rated phase current ‘Tho components of armature current shown in Figs ‘52{a) and 52(b) have seeociated with them the field cur= rents shown in Figs. 22(d) and 52(e), respectively, the for- ‘mer comisting only of even harmonies and the latter uly of odd harmonies. In Fig. 52(f) is shown the total field current. The average magnitude of this current is equal to ee Machine Characteristics Chapter 6 sitvEgd 18 Famepae™ VBL AES, 7 atv Eg __ sin (2rft+o) (leyFae) + (qa) cos 3 aft OY} With the assistance of Fig. 52 it will be seen from Ba, (105) that the maximum amount of the edd harmonie vale the total current ia dependent upon the inetant during tho eyelo at which shortcireuit occurs and reaches evi Assuming no decrement for either the odd or even harmonics (108) ‘eomponent is equal to ‘The maximum value of (109) od immo) Sem vin lo edniat qu) o © * Fig, 53—Wave form of voltage across terminals of a water- wheel generator without damper windings for a terminat-to- ferminal short circate from no-load, xy/t.°=2-5. (a) Initial value of od harmonic nrmporent (deeay slowly) (©) intial value of even harmonic component fr sin d= 1 ces rapidly). Ts magnitude varke evweon that given and ite segative depending upos the point durisg the eyelet which hort eieuit occurs, Te may be zero. (6) Total initial vale for sia ge=Chapter 6 Phe rms total current is equal to the square root of the sum of the squares of those components. Tt must be re- membered that the unit of current is the crest of rated terminal current. Whon expressed in terms of the rated ‘rmg current the above figures must be multiplied by V2. ‘The voltage from the short-circuited terminals to the free terminal, neglecting decrements, is equal to oe BLK [sin(2xft-+4o)+3h sin B(2af-+ $s) 5p sin 5C2nft-+go)-+.-) + 81s sin gi [2b €08 2(2e/0+ du) FAB cos LOnfl+ B+. J ag) in whieh a3) ‘and b has its previous significance. Like the short-cireuit current this voltage ean likewise bo reselved into two components that together with the total voltage are plotted in Fig. 53. Tho maximum possible voltage, that which ocours when sin @p is equal to unity, Sn(e%-1) ain When sin d0, the even harmonic component ia equal to oro and for this ease the maximum voltage is, ais) avon aaa ae tne ‘The corresponding line-to-neutral voltages for the termi- nal-to-terminal short-circuit are 24 of the above fignees. In all of these expressions the eres! value of rated line-to- neutral voltage has been used as a base. When the rms figure is used, the above quantity must be multiplied by V2. For a terminal-to-neutral short cireuit, noglecting de erements, the short-circuit current is ey (118) 29. Unsymmetrical Short Circuits Under Capa- citive Loading ‘When a sulient-pole machine without damper windings is loaded by a highly capacitive load," there is danger, 3. S4-Schomatle dlagram of a vhree-phase, salient-pole alternator to which a three-phase bank of eapacttors and @ terminalcto-terminal short eireuit are applied. simultane- ‘ously, Machine Characteristics 179 i VRGKatleeb conto sate COBO e=-314K[ sino +3bsins0+Sbtsinse+—3 ‘manmatancrzerseterser-3 7 x rel 3B weed. ig. 85 —Equivatent circutt to which Fig, 54 may be veo VaverHt at times of unbalanced short circuit, that resonaneo occur between the reactance of the machine and the load with the possibility that dangerously high voltages might result. Considering a purely capaeitive load such as an unloaded transmission tine, theschematie diagram isshown in Fig. 54 and the equivalent eireuit in Fig. 55 for the condition of a terminal-to-terminal short cireult.. The emf ‘applied to the eireuit is equal to the open-circuit voltage for the same short-cireuit condition. ‘The oselllographie results of tests made on a particular machine as terminal to-terminal short cireuits are applied for differont amounts of connected eapacitanee are shown in Fig, 56. Resonance wil ecu nar pints for which he quantiy 25 =m where m represents the integers 1, 2, 3 ete., and slso the order of the harmonic. The nature of this reousnee phe- nomenon ip illustrated moro elearly by tho curve of Rig. 57, fn which is plotted the maximum voltage during short cireuit in per unit. To orient one’s self with regard to the length of line involved in these considerations, the figure in miles whieh appears below each oscillogram of Fig. 56 represents ap- proximately the Iongth of single-cireuit 66+ or 220-kv transmission line that, with @ generator having the char- rcteristies of the one used in the test is required to satisty the given value of ra//7e"F. These figures were arrived _at by assuming a generator capacity of 25 000, 75.000, and 200 000 kevs for 66-, 132-, and 220-kv lines, respectively. Forsmailer machines the length will decrease in proportion. ‘The possibility of the existence of such resonant condi- tions van be determined for other types of loads and other types of fauits by setting up the network for the system and replacing the machine by the reactance v/2e0- Thit circuit should be set up for the positives, negative, and aero-sequenco networks and the networks connected in ‘eevordance with the rules of symmetrical components. Any condition for which the impedanes as viewed from the ing is sero or very stnall shoud by avoided. ‘Since the dangor of these high voltages arises from the dissymmetry of the machines, i¢ can be eliminated effec- tively by the installation of damper windings. Fig. 58 presents oscillographic evidence of the voltages existing for machines equipped with different types of dampers as180 Machine Characteristics Chapter 6 Xo LENGTH OF SINGLE LENGTH OF SINGLE CIRCUIT LINE IN MILES GIRGUIT LINE IN MILES = /xgXq 7A 9) 199) mi (ay ° ey ig, 8¢—Eflect upon the terminal voitago of varying the shunt capacitive reactance when a termi cult Is applied to a machine without damper windings.imal { Be A6 20 ke) Wig? Fig. 57—Experimental values of crest voltages (ewelfth cyclo) itch In Fig. 516 closed. Unit of ro-terminal voltage before short cireult, 3,/34'~2.2. Machine without damper winding. terminal-to terminal short circuits and capacitive react~ tives are applet simultapeeusly. While a continuous Gr conncetol damper winding i» most fictive ton connected damper winding having rato of 25 equal to at least 1.35 will be found adequate for practically all purpose IX. DAMPER WINDINGS ‘The addition of copper damper windings to machines effectively simplifies the characteristics of the machines as viewed externally in that harmonic effects are largely eliminated. However, the addition of other possible cir- nite for current flow complicates the internal calculations. ‘The influence of dampers ean in most eases be evaluated jn torms of their effect** upon the subtransient reactances in the two axes. 30. Types of Damper Windings Damper windings are of several general types. Connected Dampers—These consist essentially of windings similar to a squirrel-cage or an induction motor. ‘They ate continuous between poles as shown in Fig. 59 in which (a) shows the connection betwoon poles for a slov*- speed machine and (b) shows the additions! bracing re- quired in the form of an end ring for higher speed ma- chines, In this type of damper, 2” and 24" have nearly the same magnitudes. ‘Non-connected Dampers—The dampers in each pole ace are independent from those in adjacent poles, as shown Machine Characteristics 181 (n) Fig. 58-—Bflect of damper windings. ‘Terminat-to-terminal short elreults (a) No dampers () Connected vopper damper (6) Connected high resistance damper. () Non-Connected copper dazaper. ‘Terminal-to-neutral short clrcult: (@) No dannper ©) Connected copper damper. G Connected high reaistance damper, @) Non-connectad copper damper. in Fig. 60. They are somewhat cheaper than connected dampers but at the expense of no longer being sble to make xy” and 24" equalFig. 39—Connected dumper winding (0) Slow-speod machine (©) Highspeed machine Fig. 60—Two types of non-connected damper windings. Special Dampers—In this classification fall such dampers as double-deck windings, which are in effect a ouble winding, one of high resistance and low reactance and the other of low resisianee and high reactance. ‘The principal uses of this type are in motors where the com- ‘bination provides tetier starting characteristics. At low speeds the high reactanee of the low-tesistance winding forees the current to flow through the high-resistance wind- ing, which produces a high torque. At higher speeds the low-resistance winding hecomes effective, Another type of special winding is one that is insulated from the iron and Machine Characteristics Chapter 6 connected in series to slip rings, By connecting a variable resistor externally to the slip rings the starting charaeter~ isties ean he varied at will. ‘The general characteristics of damper windings will be discussed under the following heads, 31, Balancing Action and Elimination of Voltage Distortion One of the earliest needs for damper windings arose from the use of single-phase generators and, later, phase bal- neers. Such machines if unequipped with damper wind- ings have characteristies which resemble closely those of a three-phase machine without damper windings when a single-phase load is drawn from it. Voltage distortion similar to that discussed under unbalanced short-cireuits occurs. In addition, if this condition persists the currents that flow in the body of the pole pieces, produce excessive Jheating. The addition of damper windings provides 4 low- resistance path for the flow of these currents and prevents, both wave distortion and excessive heating. Because of the steady character of the load, damper windings in single-phase machines and phase balaneers must be heavier than those in three-phase machines, The best criteria of « polyphase machine to earry un- balanced load are its negative-soquence reactance and resistance. The former relieets its ability to prevent un- balancing of the voltage snd the latter its ability to carry. the negative-sequence current, without undue heating of the rotor. ‘These properties are particularly important for such fluctuating loads as electrie furnaces. Not only do the dampers reduce voltage unbalance but also reduce ‘wave form distortion, 432, Negative-Sequence Reactance and Resistance As discussed previously the negative-sequence reactance and resistance of amachineare both affected by the damper ‘windings. Table 2 shows the effect of different types of ‘windings upon # 100-kva generator and Table 3 upon a 8000-kva synchronous condenser.** Both of these tables represent teat results, "Tauuse 2— Constants OP A SvcHRONOUS GENERATOR A ‘Arrocte by ‘hype oF Daupen WiNDIKG (100 KVA, 2300 Vous, 25.2 AuPEnes) Type No damper 0.146 | 0.181 | 0.98 ter | 0.164 | 0.02 | 0.062 | 0.213 0.300 | 0.246 0.087 | 0.118 33. Damping Effect In the early days when prime movers consiated mostly of reciprocating engines the pulsating character of theChapter 6 ‘Tass —Constanns oF & SYNCHRONOUS CONDENSER 48 ‘Arroore BY ‘vee or Dax WINDING (9000 KYA, ‘4000 Vouts, 721 Aurenes) * med adta) ‘Type boos ee ee Test [Calculated | Test _| Calculated No damper 0.01 | 0.0 075 | 0.60 Connected copper. | 0.028 | 0.020. | 0. 021 Connected brass. | 0.085 | 9.018 | 0.195 | 0.215 Connected Everdar | 012 | 0.125 | 0.20 | 0.21 torque made patallel operation difficult, ‘This waa success fully solved by damper windings in that the damper wind ing absorbed the energy of oscillation between machines and prevented the oscillations from becoming cumulative. More recently in consideration of the stability problem Jow-resistance damper windings have been advocated for the samo reason. While a low-resistanee damper winding will deerease the number of electro-mechanical ascillations following a disturhanee this effeot in itself is not impor- tant! in increasing the amount of power that ean be transmitted over the system. ‘The general influence of damper windings, their negative- sequence resistance and reactance, and also their purely damping action, upon the stability problena, is discussed jin more detail in Chap. 13. 34. Other Considerations Affecting Damper Wind- ings Synchronous generators feeding loads through trans ‘mission lines having a high ratio of resistance to reactance tend to set up spontancous hunting.» This tendency is steater at light loads than at heavy loads, the eriterion at ‘which it tends to disappear being ishen the angle between the transient internal voltage and the load voltage equals, the impedance angle of the connecting impedance. There need not be any periodic impulse, such as the pulsating torque of a compressor, to initiate this phenomenon but it may very well aggravate the condition. Damper wind- ings are very effective in suppressing such inkerent hunt- ing conditions and also alleviate hunting produced by periodic impulses, although the latter phenomenon is ust ally eliminated by altering the natural frequency of the system by changing the fly wheel effect of the generator ‘or motor o both. Synchronous motors connected through hhigh resistance lines or cables also develop spontancous hunting but not so frequently as they are always provided with a damper winding. Series eapacicors in decreasing the effective series react- ance inerease the ratio of resistance to reactance and thus tend to inerease the likelihood of spontaneous hunting. Tn general, where beneficial effects ean accrue with the use of damper windings, the benefits are greater for con nected or continuous dampers than for non-connected dampers. Mechanically there is no choice as both types can be made equally relisble. ‘The non-conneeted winding lends itself somewhut easier lu the removal of a pole but not to sufficient extent to constitute a consideration in the choice of type to install. A ratio of 2," to 24" as low as about 1.35 can be obtained with non-connected and 1.1 with connected dampers, Damper windings for which this, ‘Machine Characteristies 188 ratio is greater than 1.35 and less than 1 percent, respectively, to the price of the machine. In con- sideration of the many complicated problems involved in the selection of a damper winding it wonld appear, in view of the lov increase in cost of the connected damper, that iff any damper winding is thought necessary, the connected type should be used. X. SELF-EXCITATION OF SYNCHRONOUS MACHINES ‘When a synchronous machine is used to charge an un- loaded transmission line whose eharging kva is equal ap- proximately to the kva of the machine, the machine may beoome self-excited and the voltage rise beyond control ‘The conditions that must be satisfied for this phenomenon to oveur are made manifest by determining the machine characteristics for a constant induetive reactive load. No Loa SATURATION Tounve: TERMINAL VOLTAGE cn a FIELD ce Fig. 61—Gonstruction of regulation curves for Induction loading. In Fig. 61 the line OG represents the no-load saturation ‘curve, Suppose the machine is loaded with « three-phase reactor equal to z, ohms per phase. To determine the regulation curve for this impedance, that is, a curve of terminal voltage plotted against field eurrent, proceed as follows: Choose an armature eurrent such that 2", the terminal voltage, is approximately rated voltage. ‘This voltage is given by the distance BD in Fig. 61. By adding “In thisdisoasion, the terminal voltage ie ogneded asthe termine to-neutral value. When terminal-Hnterminal voltage is used the vot ‘age dtops considered will have to be multiplied by V/3.184 to this distance the z»i drop, DA, the voltage behind Potier reactance denoted by the point A is obtained. The mag- netizing current to produce this voltage is given by the distance OB, In addition to this, however, the field current. i is required to overcome the demagnetizing effect of the armature current. For normal current, Ai is the distanee EN in the Potier triangle, OMN. Tn conclusion, to produce the terminal voltage F, the field current OC is necessary. ‘The tangle BAC is a sort of Potier triangle, in which the Potier reactance is replaced by a reaetanee equal to (z,+2.). Thus by drawing any line HG parallel to CA and GJ parallel to AP, the intersection with the vertical from H determines the terminal voltage for the excitation H. When the load consists of balanced capacitors having a reactance ze in which ze ia grester than z,, the impedance fs viewed from the valtoge behind Potier reactanee is ea pavitive and the armature current is magnetizing instead tf demagnetizing. This case can be treated in a manner similar to that for an induetive-reaetance load with some modifications as is shown in Tig. 62. Tn this figure the distance CF represents the terminal voltage produced by the external drop zi. Since the current Jeads the terminal voltage by ninety degrees the voltage behind Potier react tance for the assumed armature current is found by sub- TERMINAL VOLTAGE. ° ue FIELD CURRENT Fig. 62 Construction of regulation curves for capacitive leading. fal Machine Characteristics Chapter 6 “TERMMAL VOLTAGE. o FIELD CURRENT Fig, (Regulation curves for constant capacitive load of ‘such valuen on t0 give the loada ae rated voltage Indicated on the curves. HG parallel to 0q, Point q represents excitation and internal voltage, neglecting saturation, to produce rated ‘terminal voltage with 100-pereent capacitive current. tracting the drop xyé giving the distance CD or BA. To produce this voltage the magnetizing current OB is re- quired but since the armature current is magnetizing to the extent of Ai, the actual field current necessary is only OC. This determines Fas a point in the regulation curve. For other field currents such as the point H, draw HG parallel to CA until it intercepts the no-load saturation, curve at G. Then draw GJ parallel to AF. ‘The intersection, with the vertical from HT determines the point J. Fig. 63 depicts the regulation curves for different sizes of capacitors. The number assigned to each curve represents ‘the percent kva delivered at rated voltage. Fig. 62 is equal to tan > (rma) Ai sero excitation it ean be seen that if this angle is sufficiently small, intersection with the no-load saturation curve is, possible, but as « increases a point is finally reached at Which intersection is impossible and the solution fails, ‘This signifies that when this point is reached self-excitation does not occur. This eritieal condition occurs when the (ae—a9)i ‘The angle a i At slope equals the slope of the no-load saturation ai curve. Tn discussing the significance of sy use was made of Big. 10, where it was pointed out that DA is the current necessary to overcome the demagnetizing effect, AZ, of the armature current. ‘The distance AB is the syuchronousChapter 6 Machine Characteristics 185, reactance drop rai and DC the Potier reactance drop. Thus the slope of the no-load saturation curve is equal to Ai 1909 RPM ‘The condition for self excitation is then that @ ao Gi ra dt aun Stated otherwivo, tho machine will become aolf-oxcited if the ka of th achie ws nil by is less than the IW wicnoraRaos PER PHASE daring he of thee nn i exept for ee cases, of the order of 120 percent, danger may threaten when the charging kva requirements of the line exceed approximately 80 percent of the kva of the machine. XI. CAPACITANCE OF MACHINE WINDINGS A knowledge of the capacitance to ground of machine windings is necessary for several reasons, among which are: (a) Grounding of Generators. This is discussed in con- siderable detail in the chapter on Grounding. ‘The ca- pacitance to ground of the windings must be known so that the associated resistance ean be selected. (b) Systom Grounding. ‘The capacitance must be known so that the contribution of this clement to the ground current can be determined for single lino-to- 7 ground faults. The contribution to the fault current for (00) this condition is equal to v3 2xjCoEX 10 where jis the system frequency, B the line-to-line voltogo and C, the Fig. 64-Capacitance to ground of TURBINE-CENERATOR capacitance per phaso in microfarads, ‘windlags for 13 200-vole machines in microfarads per Phase, (c) System Recovery Voltage. The capacitance of the For other voltages multiply by factor Kin inserts rotating machines may be an important element in the determination of the system recovery voltage. Tt is cum os os} || g300 verre 900] ours /. 3800 vours ave on Ty CIN MICROFARADS PER PHASE S € € € §€e ¢ € we FF va Fig. 65—Capacitance to ground of SALTENT-POLE GENERATORS AND MOTORS in microfarads per phase.186 71 Ca sl bs —t 0 00 $00 300405 S00 HORSE POWER Pig. 66—Capacitance to ground of 2360-volt SYNCHRONOUS MOTORS In microfarads per phase to ground. For voltages between 2300 and 6600, the capacitance Will not vary more ‘chan £15 percent from the values for 2300 volt. tomary to represent. the machino eapacitance in this work Ly placing one-half of the total capacitance to ‘ground at the machine terminal. For details of this ype of calculation refer to the chapter on PowerSystem Voltages and Currents During Abnormal Conditions. (@) Charging Kva. In testing the insulation of ma- chiues, particulary in the Beld, is sometimes nevesuy to know the approximate charging kva of the windings so thal a transformer of sufficiently high rating ean be provided beforehand to do the job. "This is required either for normal routine testing, for testing at time of installation or for testing after rewinding. The cherging eva per phase is equal to 2x/CyB"X 10" where Co is the 004 ——— Fr] 108 B06 306 HORSEPOWER 00 F00 Fig. 67—Capacttance to ground of 2300-volt INDUCTION, MOTORS ia microfarads per phase. For voltages between, 3300 and 6600, the capacitance will not vary more than, “£15 porcent from the values for 2500 voles, Machine Characteristics Chapter 6 capacitance per phase to ground in microfarads and Eis, ‘the applied voltage from winding to ground. Figures 64 to 67 provide basic data calculated for West- inghouse turbine generators and salient-pole generators land motors. ‘The generator data was obtained from refer ‘ence 28 and the motor deta from como unpublished max terial of Dr. E, L. Harder. This information should be typieal of other machines to within about + 50 percent. Tn general, it should be borne in mind that these character- istics vary greatly between machines of different designs Fortunately, however, not very great accuracy is required for the applications cited a ‘XII. NATURAL FREQUENCY OF SYNCHRONOUS MACHINE CONNECTED TO INFINITE BUS A synchronous machine connested to an infinite bus possesses a natural period of oscillation which is given in the ASA C50—193 Rotating Electrical Machinery Standards as 35.200 IPxf a Gomi N Wine Fe er mini (113) the synchronizing power in kw per electrical in displacement, the system frequency. ‘When given an angular displacement, the machine ascil- Iates with this frequency and finally eubsides unless sub- jected to periadie impulse of proper magnitude. It is not within the scopo of this work to discuss this subject in its entirety, but merely to derive the above expression, If an ineremental displacement 49 be assumed, the cor- responding synchronizing power is AP=P.A8 in kw (as) and AP is in degrees. From the Stability Chapter it ean be seen that the asceleration of the rotor is 180 f am iS AP in dog/nct GAP in rad/sect (ats) where the kva refers to the rating of the machine and H the inertia constant, Substituting £ from Eq. (93) 108 om Fat TW ame? 8 d/seet (118) and substituting AP from Eq. (114) 10° fP. “Oza WR rom an = Kae, (us) m0! GP, 0281 (WRI irpm)* a “The sign of Pn netully negative an an iezement in 49 produtcsa torque which gends to return the mache tothe Speating angle. ‘Thos, Kis postive. Now 21a) ge 7 KO9. (120)Chapter 6 Further, let AO=A sin 2af.t then substituting this relation into Eq. (120) (121) —(af.)* A sin Qxfst= —KA sin Qnfet from which he ‘Substituting K from Eq. (119) 587 Se mV WHE VRS Ber 80 (122) which converts to Hq, (11). ‘XIII. TYPICAL CONSTANTS AND COSTS Both the voltage and the current at which a machine ‘operates affect certain of the principal constants through the variability of the permeability of the iron. In this ‘sense, these so-called constants are not in reality constant. Consider the transient reactance, x4/. If three-phase short circuits are applied to a machine from no load, the reaet= ances so obtained vary with the excitation, Two of these ‘quantities have been given special designations. Thus the reactance obtained when the excitation is such 98 to pro- duce rated voltage at no load before the short-circuit i called the “rated-voltage reactance” and the roactance ‘obtained when the excitation is reduced s0 as to produce from no load a transient component of the short-circuit TURBINE GENERATORS 3 304 - _ z B80 aK 5 - eaLievr-pove cewenarons Bod ed it toro go i? 3 : sauienr-pove 5, EricrmonoU CONDENSERS 3 SQ a TRANSIENT COMPONENT OF ARMATURE CURRENT IN PER UNIT Fig. 68—Saturation factors for transient reactance, ‘Three- phase short circuits from rated voltage 0 lond. Current Tmnited by series reuctance. Machine Characteristics 187 enrrent equal to rated value is called the “rated-current, reactance.” A knowledge of these two values of zu’ is not sufficient, for all applications for which za’ is required. The rated- current 2’, because of lower excitation, lends itself more readily for determination from test. The rated-voltage 24" is that required for chort-circuit studies. Saturation within, the machine is a minimum for the former and a maximum, for the latter. ‘The rated voltage value is sometimes called. the “saturated value” and is the value usually given by the designer. Certain applications, such as stability studios, demand a quantity determined under conditions for which, the terminal voltage is near rated voltage and the armature current is likewise near its rated current. Fig. 68 obtained, from data presented by Kilgore! shows how the reaetanees of typical inachines of different elusses vary if three-phase shorCeireuits were applied from rated voltage no load, the currant being altered by introducing different external reaetances in the armature circuits. ‘The rated-current figure is used as a base for all the curves. The particular reactance on the curves for rated current is the one that. would have greatest utility for stability and regulation, problems. No specific name has been assigned to this quantity. Similar considerations apply to the subtransient react- ‘ances, with this difference, that the rated-current reactance zz" is obtained from the same test as that for which the rated-current reactance of 2,’ was obtained. In this ease rated current refers to the transient component and not, the subtransient component of current. Fig. 60 shows how ut aopoLe ol ‘TURBINE GENERATOR of { oa} — on 7 ° | ronan cenerator 0, SALIENT-POLE MACHINES |wirhouT CAMPER WINOINGS. i TRANSIENT COMPONENT OF ARMATURE CURRENT IN’ PER UNIT SATURATION FACTORS 1M TERMS OF RATEO-CURREWT VALUE. Walue used ag base. All reactances from three-phase short circults without erternal reactance. Saru- ration factors for ealient-pole machine with damper winding, “a equal to unity.188 24!" varies with the transient component of current, all points being obtained from three-phase short-circuits with no external reactance, the current being altered by the excitation before the short-circuit. Tn general, itis unnecessary to make this distinetion for tho nogative-sequence roneiance, ‘The A/E code" sug- ‘gests determination of #2 by means of the method diseussed °0 906846006" eG0G0 vA Fig. 70—Normal unsaturated transient reactance (x1) for waterwheel generators "30080, 160669 120000 under Negative Sequence Reactance, the current during the sustained terminsl-to-terminal’ short-circuit being limited to the rated current. ‘The normal value of 2g, designed into waterwheel gen- erators varies with the kva capacity and speed. ‘These values are plotted in curve form in Fig. 70. To obtai Jower values than those indicated usually involves an increased ¢ “The angular relations within the machine are determined to-s large extent by x4. ‘The variation, by test, of xq for several salient-pole machines" is shown in Fig. 7 Machine Characteristics Chapter 6 ong, 1, [28.000 Kya g oe Zoro Boos| g 5006] ~ s 3750 4A 08! J OE 0406 08 101 PER UNIT ARMATURE CURRENT ig, 72-Variation of x, for turbine generators, 1% (e CONSTANT IN SECONDS ocmunerueed OT ES 4S 8 8040 BOO RATING IN THOUSANDS OF VA Fig. 73—Open-cirevit transient time constants of a-c gea- ferators and motaee, <0) Ee, 27 256 i a 5 ARMATURE CURRENT IN PER UNIT “32 q| Fig. 71—x, for salient-pole machines. w A» 7500 kva generator without damper winding. 30 ao 8080100 Bra Wego witht saps wna RATING IN THOUSANDS OF KVA Sm mor with. damper winding roared, Fig. 14-Openccircuit transient time constante of turbine Benerators. ‘The sero-sequence reactance, as evidenced by Fig. 72 the wide variation of T'. with the size of the unit the taken from Wright's paper,” is not affeeted to any great extent in the region for which it. hae greatest use. For practical purposes the effect. of saturation upon the open-circuit. transient time constant Ty's and the sub- ‘transient short-circuit time constant 7" can be neglected. In general, 7.’ varies ” in the same manner as 24, 50 that the relation 7y/=747,', is etill maintained. Because of curves of Figs. 73 and 74 teken from a paper by Hahn and Wagner,! are also ineluded. ‘Table 4 gives both the range of typical constants that. are characteristic of normal designs and also an average ‘that can be used for general purposes when the specific, value of a particular machine is not known. ‘The negative- sequence resistance is that obtained at a negative-sequence current equal to rated current. 1t must be kept in mindChapter 6 Machine Characteristics 189 THOUSANDS OF KYA is} Fig. 75—Inertia constante, (a) Large turbine generators, tuchine included, () Large vertical type waterwheel generators, including allon- ‘ance of 15 peroot for waterbed. that the fos assoclated therewith varies as the second 3 ale! & les le power of iefor salient-pole machines cither with or without ale] = [e2}fs) B | 22 ¢ damper windings and as the 1,8 power of i for turbine 5 geype) ge | ge F generators. Column (9) in Table 4 refers to the a-c re & _|— sistanee, ri, (which includes the effect of load losses) and 5 Ee we ‘column (20) the de resistance, rs. Blsl es (32 5 ‘The inertia constant, which i discussed ia Chap. 13 5 & in kowise given in Table 4. The general variation of I] of 3 = ; Cubuenets tad he onesie seus eens =|) Z heel generators are given in Fig, 75. The effect upon g|*| & [58 2 2 = - 6 3 E a 8 [het — z| 2 [~~ @ x lob s| ap a] S Me = 8 Brblaes G | af Lace aru conorwtna é He Df age eeli}2] 372 Ege 4 pS a Ae RG oI =e ee g sige tals else Sle] 2 seq] 3 ; i ya oe i e/a) 2 8 rece Toons or E pays els gerectill ghd , f z 7 z & 4 . ws 3 ro 2 FH : 5 $ § ‘ ze 0 i g E 4 : é of increasing the short-cut. ratio and changing the powerdaetor ie given in Fig. 76. The WA? represented by the curves of Figs. 75 and 76 are those obtained from a normally designed machine in which ao particular effort has been made to obtain abnormally high H. When magnitudes of WR* in exeess of these are desired a more expensive machine results. ‘The additional cost of the additional WR? is ahout proportional as shown in Fig, 77. The cost per kve of waterwheel generators depends upon its ka and speed. The extent of this variation is shown in Fig. 78, Machines of higher short-circuit ratio ‘F povrer-factor aremore expensive in the proportion shown re Reactances are per urit, time constants are in seconds, Values below the line give the norinal range of values, whi Grains tao190 Fi. Fig. Machine Charocteristies Chapter 6 PERCENTAGE CREASE IN H a SHORT CIRCUIT RATIO Fig, 76—Bifect of short-circuit ratio 0 8S a5 40 00 PERCENT INCREASE WH {77—Bifect of increasing H above the normal values by Fig. 75. INCREASED MACHINE ‘cosT=rERCENT 4 T (COST PER KVA IN DOLLARS, 230 ‘THOUSANDS OF KA 30 0 78Cost of waterwheel generators including direct-con- ‘nected esclters only. (©. power factor and 1.0 short circuit ratio) (0.9 powerJester and 11 short eet ratio) (0.0 power actor and 1.25 short cit ratio) PERCENTAGE INCREASE IN COST 22 SHORT CIRCUIT RATIO 2630 Fig. 79—Rftect of short-circult ratio upon cost (Normal 1.0 short-circuit ratio and 0.8 power-factor used as bate). RVOROGEN COOLED ® “OUTDOOR, A aip|ZooLeD !N000R ‘COST PER «VAIN DOLLARS: a THOUSANDS OF kVA Fig. #0-Gost of synchronous condensers including exciter ‘and autotransformer. in Fig. 70, Naturally these figures will vary from year to ‘year with the cost of materials and labor, ‘The condenser cost por kva including the exeiter, pilot exciter, and auto-transformer is plotted in Fig, 80. The cexciter kw varies with the size of the unit, ranging from 1.2, 0.7, and 0.82 percent for a 1000, 6000, and 50 000-kva, unit, respectively. ‘The cost of normal exciters for water-theel generators varies from 7 to 18 percent of the cost of the generator alono for slow speeds, and from 2.3 to 6 pereent for high spews. ‘The larger figures apply for units of about 3000 va and the smaller figures for machines of about 50 000 kva, Direet-connected pilot, exciters cost approximately 30 pereent of that of the exeiter XIV. INDUCTION MOTORS ‘The equivalent eireuit of the induction motor is shown. in Fig. 31 “The loss in the resistor *=Sr, represents the shaft power and sinee the circuit is on a per phase basis, the total shaft power is thusChapter 6 ‘Total shaft power = in watts (123) Sani) im bp. (128) ‘The rotor copper loss is (Breih. ‘other losses, the efficiency is ‘Therefore, neglecting total shaft power ‘otal shalt power-+ rotor eopper loss les Efficiency. or (025) S41 ‘Thus, the efficienoy decranses with ineressing slip. For 10 percent alip the efficioney is 90 pereent, for 90 percent, slip the efficiency is 10 pereent. Similarly, the rotor eopper logs is directly proportional ta slip; being 10 peveent for 20 percent stip wud 90 percent for 90 percent ‘The total shaft power ean also be expressed in terms of torques. Thus, Total shaft horse power = 21 )(pm) aya. (18). (026) Bquating (124) and (126), the torque ie T7041 — Gein en 27 inom: te ps ai ‘The equivalent circuit of Fig. 31 can be simplified con- siderably by shifting the magnetizing branch to disoctly vross the terminals, The resultant approximate cireuit shown in Fig. 81. This approximation permits of ky Ket Fig, 81~Approsimate equivalent circuit of induction motor. relatively simple determination of ig, so that Eq. (127) becomes 7.045" a " —e Sint. a2) (a) + eetal ‘Most transients involving induetion motors fall within ‘one of two categories; first, those in which the machine is disconnected from the source of power and, second, those in which the machine remains connected to the sourve of power. In the first case the transient is determined largely by changes in magnetization and may be quite long. An Machine Characteristics 191 example of this case is the phenomena that occurs during the interval between the transfer of power-house auxiliaries from one source to another. In the sevond case, the tran- sient is determined by reactions involving both te stator and rotor and the duration is quite short, Examples, of this easo, are the sudden energization of an induetion motor ‘or sudden short eirenit seross its terminals, 35 Contribution to System Short-Circuit Current In the calculation of aystem short cirenits only eyncheo- nous machines are usually eonsidered but in special eases where induetion maehines eonstitnte # large proportion of the load, their contribution to the short-circuit current even if it duration is only a few eycles may be large enough +0 influence the choiee of the breaker fram the standpoint ofits short-time rating, that is, che maximum rms current the breaker ean earry for any time, however emall ‘AS a. first approximation ihe shart-ireuit current sup- plied by an induction motor ean be resolved into an alte nating and a unidirectional component much lle that for asynchronous machine, ‘The initial rms magnitude of the alternating component: is equal to the terminal voltage to neutral divided by the blocked rotor impedance per phase. The time constants ure namely, for the alternating component, (blocks phase in ohms) suse bn ohms) cycles, for the unidirectional component, (blocked rotor reactance per phase in ohms) 2x (stator resistance per phase in olims) Fig. 82 shows the short-circuit current of a 25-horse-power, 550-valt squicrel-eage motor. The dotted line in the upper in cycles, ig, 82—Short-circwie currents in armature of equirrel-cage induction motor.192 curve indicates the computed value of the envelope of the alternating component of short-cireuit current. The am- plitude shows a substantial check but the computed time constant waslow, This ean probably be attributed to using the a-e resistance of the rotor rather than the d-e resistance. ‘The dotted line in the lower curve is the computed value of the unidirectional component, which checks quite wel. ‘Wound-rolor motors, operated witha substantialamount, of external resistance, will have such small time constants ‘that their contribution to the short-circuit ean be neglected. 36. Electro-Mechanical Starting Transient Fig. 31 shows the conventional diagram of an induction rotor. In the present discussion the per unit system of ‘units will continue to be used, in whieh unit current is the jourrent necessary to develop the rated power at the rated ‘voltage. The unit of both power and reactive volt-amperes vill be the rated kva of the motor and not the rated power either in Kilowatts or horse power. This convention is consistent with the choice of units for the impedances. At tated slip the volt-amperes input into the stator must be ‘equal tunity but the power absorbed inthe resistor —'r, ‘will be Jess than unity and will be equal numerically to ‘the ratio of the rated power of the motor to the rated kva. ‘The unit of shaft torque requires special comment. The shaft power can be expressed as Ine Shaft Power in kw=kvtrws TS*. (129) Ta terms of torque the shaft power is equal to Shaft Power in kw morse! nth st 2a(enm) wen.) (99) “33 000, Equé 38, there results thet 38.000 i FOTO pInoen eet By IBD If unit torque be defined as that torque required to pro- duce a shaft power equal to rated kva at synchronous speed, then from (130), the unit of torque is Tin tbs (132) For the purpose of determining the nature of electro- mechanical (rausients upon alartiag a motor from rest, the first step involves the calculation of the shaft torque as a funetion of the speed. Either the conventional method of the circle diagram or expression (132) can be used. Tn. using the Istter method it is only nevessary to solve the network of Fig. (31) and substitute the solution of I, there from into Eq. (132). A solution of a typical motoris chown, in Fig. 88. For most motors the magnetizing branch ean be neglected, for which case the torque expression becomes Machine Characteristics Le 5 = oe : ° ee pe fea z um god ee 5. caccutareo Tomoue| | ° | POINTS NEGLECTING 5 12 1. Fig. £3—Illustrating calculation of speed-time curve of an induction motor upon application of full voltage. ‘The crosses close to the torque curve in Fig, 83 were computed by this expression Tu Fig. 83 is also shown the torque requirements of a particular load such as 2 blower. Upon applying voltage to the motor the difference between the toraue developed by the motor and that required by the load is the torque available for seceleration of the rotor. ‘To convert to accel- eration it is convenient to introduce a constant, H, which 4s equal to the stored energy in kw-ecc. per kva of rating ‘at synchronaus speed. Hf may be computed by means of Eq. (93). WA? must, of course, include the HK? of the con- nected load. ‘Suppose that one per unit torque is applied to the motor hick meaus that at synchronous speed the power input, into acceleration of the rotor is equal to rated kva, and, euppose further that the rotor is brought to synchronous, speed in one second, During this intervai the acceleration, is constant (1 per unit) and the power input increases, Tinearly with dme so that at the end of one second the stored energy of rotation is (16 kva) in kw-sec. Thus 1 per unit of torque produces 1 per unit of acceleration if the inertia is auch that 34 kva of stored energy is produced in ‘one second. From this it can be seen that if the inertia is such that at synchronous speed tho stored energy is H, then to develop this energy in ono second, the same acceler ation but a torque 2H times aagreat is required. Therefore, ‘there results that Tin per unit~ (138) (134)Chapter 6 csc cn epost as. Thus feom (134) oe, 135) ‘This function is likewise plotted in Fig. 83. ‘The utility of this form of the expression may be seen at once from the fact that ie known asa function of # and the time to reach any value of @ can be determined by a simple etl. Hs “oe By summing up areas (such as indiosted by the shaded portion) in a vertieal direction, the time to reach any speed is obiained. ‘The curve of time so obtained is plotted in Fig. 83. ‘he following formula ean be used to form an approxi- mate idea of the time required to accelerate a motor, ‘hose load varies as the square or cube of the speed, to half speed ‘Time to half speed: (138) Het, seconds re (137) All of the above units must be expressed in per unit. Re- member also that. z, should include any external react- ance in the stator back to the point where the voltage may be regarded as constant and e, should be that con stant voltage. 37, Residual Voltage If an induction motor is disconnected from ite supply, it rotates for some time, the rate of deceleration being de- termined by the inertia of its own rotor and the inertia of the load and also by the nature of the load. Becauso of the inductanee of the rotor, fux is entrapped and voltage 58 PERCENT OF BASE PRICE tool -208=220-440-580 2 Machine Characteristics 198 ano Bone 8 ne 370 Fig. 8&-Decay of residual voltage” of group of power house, auxiliary motors. 210) 8 3 OPEN CIRCUIT TIME CONSTANT=CYCLES 500) ow, S00 Fig. 85 308 409s foo HORSEPOWER Typical time constants for 2300-volt aquierel cage ‘induction motors, ooo 1500 ‘appears at the open terminals of the machine. If the volt~ age souree is reapplied when the source voltage and residual, voltage of the motor are out of phase, eurrents exceeding. starting values may be obtained. 4 1591-14000 7001-11800 901-7009 2300- 051-4209] MOTOR SIZE IN HORSEPOWER Fig. 86—Approximate variation of price with voltage and horsepower of squirrel-cage Induction motors, These values apply approsimately {or ® pul "or less for 6o-cyele ‘motors. Most economical used as base price.194 Figure 84 shows the decay of a group of power-house auxiliary motors. The group had a total rating of 2500 Jew of which the langest was 1250 hp. This eurve includes not only the offect of magnetie docay but the reduction in voltage due to decrease in speed. ‘The open-circuit time constant for individual 2300-volt machines is given in Fig. 85. ‘There is great variance in this constant between different designs but these curves give an idea of the mag- nitude for squirrel-eage induction machines. 38. Cost of Induction Motors ‘The price of induetion motors of a given rating varies with the voltage. As the rating increases the most eco nomical voltage also increases. ‘To form a basis of judg- ment of the effect of voltage upon size the curve in Fig, 86 ‘was prepared. REFERENCES A. Power System Transients, by V. Bush and R. D. Booth, AILE.E. Tronscions, Vol. 44, February 1925, pp. 80-07. 2, Further Satis of Transition Stability, by Re D. Evans and C.F. Wagner, A.ILB.B. Tronsasions, Vol. 5, 182, pp. 51-80. 1% Synchronous Machines —T ard TI—An extension of Blondes ‘Two Resctiow-Theoty—Seady-Btate Power Angle Character istics, by RB. Doherty and C. A. Nickle, ADELE. Tronsac- tions, Vol $5, 192, pp. 912-042. 4 Synchronous’ Machinee—IIT. Torque Angle Charactristios Under Transient Conditions, by Il. B. Doherty and C. A. Nickle, AEE, Transactions, Vel 6, 1927, pp. 1-14, 1, Synchronous Machines, IV, by R.E. Doherty and C. A. Niskda, ALLE, Tronsackons, Vol, 47, No. 2, April 1928, p 487. 6. Synchronous Machines, V. Theee-Phate Short Cioult Sym chronous Machines, by R. E. Doberty and ©. A. Nickle ALLELE, Tronaactions, Vol. 49, April 1880, p. 70. 7. Detnition of an Tdeal Sypehronos Dfaehine aad Formula for the Armature Plax Linkages, ly R, H. Park, Goneral Hlocric Review, June 1928, pp. 382-33. 8, Two-Reaction Theory of Synchronous Machines—t, by R. HL. Park, AV.6.8, Tranencsions, Vol. 48, No.2, Saly 1920, p. 716 Machine Characteristics 10, un 2, Chapter 6 ‘Two-Resction Theory of Synchronous Mashines, 1, by RH Park, ALT-E.E. Transneions, Vol, 52, June 1958, p- 882. AIBE, Test Cade for Synchvonous Machines, ALEE, Publi ation No, 603, June 1948, Discussion, by C.F. Wagner, A.1-E, Transactions, July 1987, p. 904 Uneyrametrical Short-Citenite in Water-Wheol Generators Under Capacitive Lonting, by C.F. Wagner, ALE. Transao tions, November 1997, pp. 1985-1986. Overveltages on Woter Wheel Generators, by C. F, Wager, The Electric Journal, August 1988, p. 821 sn September 1988, p. 351. ‘Damper Windings for Water-Whoal Generators, by C.F. Wag- ter, ALE.E. Trarsoctions, Vol. 50, March 198, pp. 140-I51- EMlect of Armature Resistance Upon Hunting of Synchrosous Machines, by C.F Wagner, ALEK, Transactions, Vol. 4, aly 1980, pp. 1011-1024. ects of Saturation on Machine Reactances, by L. A. Kilgore, ALLB.E, Transactions, Vol. 54, 1985, pp. 548-850, Determination of Synchronous Machine Constante by Test, by SH Wright, ALE.E, Transactions, Vel. 50, 1931, pp. 1331-1300, Standard Decrement Curves, by W. C. Haho and O. P. Wagner, ALLELE. Transactions, 1982, pp. 359-981 Approximating Potioe ‘Keabtaios, by’ Storing Beckwith, ATER. Tosnsactions, July 1997, 2 813 Aculiary Power at Richmond Station, by J. W. Anderson and A. C. Monteith, A.LE.B. Trawertons, 121, p. 827 . Proferted Standards for Large 3400-RPM 3-Phase 60-Cyslo (Condensing Steam Turbiae-Generators, ATEE Standards Nos. 601 and €02, Mey 1919, ‘Regulation of A-C Generators With Suddenly Applied Loads, by BI, Handes and R. C. Cheek, ADEE. Transactions, Vol. 63, 1044, pp. 310-318, Regulation of A-C Gansratora with Suddenly Applied Losds— Uyby E.L, Harderand R.C, Cheek, 4.].E.2, Troweactons 1950. Practical Caleulation of Cireuit Transient Ranovery Voltages, by JA. Adams, W, F, Skeats, KC. Van Sickle und TGA, Sites, AE, Tronsoetions, Vol Gl, 1082, pp. 771-778, 5. Bus Tranter Tests on 2300-Volt Station Auaiiary Syatem, by ALA. Johnson and H. A. Thompson, presented before AIEE Winter Meeting, Jon, 1950,CHAPTER 7 EXCITATION SYSTEMS Author: J. E. Barkle, Jr RIOR to 1020 relatively little difficulty was en countered in the operation of elevtrieal systems, and ‘operating engineers ad little concern about system stability, As the loads grew and systems expanded, it be- ‘caine necessary Lo operate synchronous machines in paral lel, and difficulties encountered were not well understood In certain aress it beeame necessary to locate generating stations some distance from the Toad centers, whieh ine ‘volved the transmission of power over long distances. It soon became apparent that system stability was of vital importance in these cases and also in the operation of large interconnected systema. In 1022, a group of engineers undertook solution of the stability problem to determine the factors involved that most affected tho ability of asystem to transfer power from ‘one point to another. ‘The results of these studies were presented before the ATEE in a group of papers* in 1924, and it was pointed out that the synchronous machine ex: citation systems are an important factor in the problem of determining the time variation of angle, voltage, and power ‘quantities during transient disturbances, E.’ B. Shand stressed the theoretical possibility of increasing the steady: state power that could be transmitted over transmission lines through the use of a generator voltage regulator snd ‘an excitation eystem with a high degree of response so that ‘operation in the region of dynamic stability would be pos- sible, Ttvas not recommended that this region of dynamic stability be considered for normal operation, but.that it be considered additional margin in determining permissible power transfer. Improvement of the excitation systems, therefore, ap- peared to be at least one method of increasing the stability limits of aystems and preventing the separations occurring, during transient conditions. Greater interest in the design of excitation ayatemsand their component parts developed, and exciters with higher speeds of response and faster, more accurate generator voltage regulators were soon introduced to the industey. Early excitation aystemé were of many different forms depending principally upon whether the main generators ‘were small or large in rating snd whether the installation was a steam or hydroelectric station. The two broad classifications were those using a eommon excitation bus and those using an individual exeiter for each main gen- erator. ‘The common exciter bus was generally energized by several exciters driven by motors, turbines, steam en- kines, waterwheels, or combinations of these to provide & main and emergency drive. Standby exciter capacity was provided in the eommon-bus syatem by a battery floated *ALL.B.B, Traneacion, Vol. 43, 198, pp. 1-108, fon the bus. Tt usually had eufficient capacity to carry the excitation requirements of the entire station for at least an. hour Motor or turbine drive was also used in the individual exciter system, but it was not long before it was realized ‘that direet-connection of the exciter to the generator shaft ‘offered an excellent answer to the many problems encoun- tered with separately = « © ” }oo380 FieLO CURRENT PER CIRCUIT 80880 Fig. 7—Esample for calculation of response of exciter. 167 kw, 126 volts, 1200 rpm, 6 poles Oe oto} — Bepasniy exited a= 128 voll ‘ies ciate pes per sitet “ Galng valtags 168 volta “ent celing voltagy “B16 amperes per sireit 01 0015 Tevitance por cree 18. sin ‘To field windings 68 ots Extra eistaneo por set 8.5 oh Teale eaece Sa 8 ood | {er polo at 125 vote det wel Gux=18 a a a ‘Total per pole at 125 volts 20.5 1x? _, a O3~ 1802 iyo BI? no288 ky 1.05, Fig. 8—Ausiliary curves for calculation of response for exam- ‘ple given in Fig. 7.202 Excitation Systems Chapter 7 TABLE 1 Oe O_] ao, aa i 7 ee | kes ra Meat) a i _|_* ” a ine. baci sca] Fa | 8. 5 ‘rom @) | from) |__| ZAD aw | aa) me | 35) ma | ws | ws [ ois ° no | 37 | az | 39 | we | 556 | ose | oom | a5 | oo | oom |oose w|i | acs | ta | ass | wa | oie | oom | 32 | oor | oom | o.oss ro | too | a7a | so | aa | oma | ss7 | oso | so | oor | oom | oss wo | sat | aoa | 5a | ase | siz | aoa | oom Po as | cum | ors | 0.20 a0 | 6.25 | a2 | oo | aus | 058 | oo | osm | x0 | oor | os fos iss | oso | aie | ta | att | sro | ato | os || 19 | oo | oo ow wo | 7a | os | 7s | 9 | mao | vo | oom $22 | ooo | ois | ost siving the curve shown in Fig. 8 Time ean then be de- termined by integrating this eurve. One method of doing, this is by means of the table constituting the insert of this figure. ‘This is found by dividing Y into increments of ‘unit width, except for the first element for which Ay. is only 08. ‘This is done to obtain convenient divisions, Tnerements of time Atare enumerated in the first column. ‘The second column represents time, the summation of the Ateolumn. On the other hand, the same integration can be accomplished in tabular form. Continuing in Table 1, column (9), the difference of successive values of y from column (5), constitutes the base of inerements of area of a ig: 8 Likewise, column (10), the mean of suceessive values of column (8), constitutes the mean of clementary areas. The product of these two values tabu- luted in column (11) is the increment of time. Column, (12) is merely @ progressive summation of (11) and gives actual time. By plotting column (1) against eoluma (12), the response curvo is obtained For higher speeds of response, the eddy currents pro~ duced in the solid yokes ean retard the buildup of the flux. ‘The extent to which this is effective is given by the curve r z > + 5 CALCULATED NOMINAL RESPONSE-UNITS Fig. 9- Correction factor to be applled ro calculated response to include eflect of eddy currents, according to W. A. Lewis! in Fig. 9 by W. A. Lewis'. ‘This curve supplies a correction to be applied to calculated responses Separately-excited exciters are usually, but not neces- sarily, faster in response than self-excited exeiters. ‘They do, however, have other advantages, such as being mare stable at low voltages, voltages at which self-excited ex citers may have a tendency to creep. Improvement in speed of response can be obtained by two general methods; (1) decreasing the time coustant of the field eizeuit, and (2) increasing the pilotsexciter voltagein the case of separately- excited oxciters or the ceiling voltage in the case of solf- exeited exciters. The former is usually accomplished by paralleling the fleld cireuits placing at the sane Give re sistors in series to limit the eurront. ‘Thus, if the parallels, are donbled, the number of poles and likewise ¥ per cireuit are halved. It is necessary to add more resistance to the external eireuit so that the resistance per circuit. remains the same. In Eq. (7) the only change is that ¥ is one- half and, therefore, the terminal voltage rises twiee as fast. 4, Calculation of Response Under Loaded Con- ditions ‘Most of the cases for which the exciter response is desired are concerned with sudden changes, such ns short eizeuits, in the armature cireuit of the synchronous machine. Asso- ciated with these changes one usually finds that the field current, of the alternator has inereased a considerable amount, perhaps in excess of the armature eurrent, rating, of the exciter. Because of the high inductance of the field circuit of the synchronous machine, the armature current of the exeiter can usually be regarded as remaining sub- stantially constant at this increased value during the period for which the response is desired. ‘When current flows in the armature, the phenomenon of armature reaction must be taken into consideration except for those machines that have a compensating winding. The function ofthe compensating winding, which is wound into the pole face of the field winding, is to annul the effect of the cross-magnetizing mmf of armature reaction. However, for machines without compensating windings, the mmf of armature reaction produces an mmf that varies linearly from the eenter of the pole piece, one side being positive and the other side negative. This effect is shown in Fig. 10 (a} im which MN represents the msximum magnetizing mmf at one pole edge and PQ represents the maximum demagnetizing mmf at the other pole edge. Fig. 10 (b) represents s section of the no-load saturation eurve in which 0 represents the generated voltage on the vertical co-ordinate and the field mmf on the horizontal eo-ordinate Tf A and C are so laid off that OA and OC equal MA and PQ, reapectively, from Fig. 10 (a), then because of the linearity of QW of Fig. 10 (a), the abscissa of Fig. 10 (b) between C4 represents the mnt distribution along the pole face. Further, sineo the generated voltages are propor-Chapter 7 3 Pune oF ARMATURE, REACTION SECTION TAKEN FROM NO-LOAD SATURATION CURVE vi UE To ARMATURE. REACTION AT EDGE OF POLE. fun FROM (0}) (o) NO-LOAD MME Fig. 10—Eflect of armature reaction im reducing tocal flux ‘across gap. (a) Shows distribution of armature mmf (b) Sec- ton of no-Iead saturation curve. tional to the airgap fluxes, the section of no-load satura tion eurve chows the effect of the superposed armature mmf upon the density of air-gap flux across the pole. "The higher romf does not increase the flux on the right-hand Sido as much as the lower mmf deereases the flux on the left-hand side, As. result, the total flux and consequently Che generated voltageare docreased from the valueindicated by CA to thet indicated by FG, whieh is obtained by integrating the area under the curve DOB and drawing FG so that the two triangular areas are equal. The extent to which the average flux or voltage is decressed ean be in= dicated by a “distortion curve,” such as shown by the dotted curve of Fig. 11. This effect is mast: pronounced in the region of the knee of the saturation curve as at both higher and lower feld enrrents, there is a tendency to add fn the one side of the pole just as much flux as is subtracted ‘on the other. The terminal voltago is reduced still further by the armature resistance and brush drops, resulting ia a load saturation curve for constant current, such as shown in Fig. 11 From this same curve it ean he seen that for s given fold resistance line, the forcing voltage (e«—rei) for avelf-excited Excitation Systems SATURATION CURVE. (OSTORTION CURVE FOR CONSTANT [ARUATURE CORRENT. VOLTAGE TERMINAL VOLTAGE FOR (CONSTANT ARMATURE CURRENT FIELD RESISTANCE LINE Fig. 11 Load saturation curves for exciter asguming constant ‘machine is very much smaller under load than under no load, In caleulating the ux linkages in accordance with Ea. (5), the distortion curve should be used for a. Except for these two changes, the load response can be ealeulated in the same manner as the no load response. For separately-excited exeiters, the forcing voltage re maine unaltered by the Toading on the machine ae it is, independent of the terminal voltage. ‘The armature resist ance can be regarded as part of that of the main field wind- ing. There remains only the distortion effect to consider ‘which amounts to only several pervent. For machines with, compensating windings, this olfect is negligible. 5. Effect of Differential Fields on Response Differential windings are provided to reduve the exciter voltage to residual magnitude or below. They consist of a small number of turns wound on each pole, so connected that the mmf produced thereby is opposite to that of the main windings. Fig. 12 (2) shows schematically such an arrangement. If the differential windings are not opened when the regulator contacts clase to produce field forcing, the differential cizruit reduces the response of the exciter. ‘The extent to which this is effective may be calculated as follows: Let a=nurber of parallel paths in the main winding. b= number of parallel paths in the differential winding, ‘c= number of turns per pole of the main winding. d= number of turns per pole of the differential winding, Ne=total number of poles of exciter. i, =current per cireuit of main winding. ig= current per circuit of differential winding, ‘The resistors Rg and Ra in series with the combined main and differential windings, respectively, may be ineluded in ‘the ealeulation by inereasing the actual resistances in each, of the main and differential circuits by alt, and bits, re~ spectively. With these inereases the resistances of each of the main and differential circuits will be designated by theMAIN winomes. (a) in Bt Bh 24/24 Hey; i iY Bi BI ai Bi or 14 12-beamate rn fr nn dierent symbols rm and ra, respectively. Referring to Fig, 12 (b) the following equations ean be written ae ® ex Rata tia) + ra ® jn which ¥ and ¥4 are the flux linkages in each of the ‘two respective circuits If all the field flux cute all tums, then = Rain bis) rata YE ee= Rain t bie) praia y=Nen pr plein 10" i) tax Bax fox per pole in 10° linea) vt, cio) Tf it be assumed that the two windings be replaced by another winding having the same number of tarns and circuit connections as the main windings, then the in- ‘stantaneous minf of this winding is the same as that of the combination if ite eurremt, ¢, is from which a Excitation Systems Chapter 7 If (10) and (11) are inserted in (8) and (9), then en Ratner dt Rice ay ecnRati tig Rotenivet % aay te we a by mating (8) ty #5, can be enna ty subtracting from (12). ‘The current é4 can then be solved terms of i. Upon substituting the expression for ts into (12) there is finally obtained that aa) in which as) be 7 Ani Plea Equation Mshows that the ordinary fiuctinkage cueve for the exciter and conventional meth of ealeulation can teusedifthecoeficient of beusad asthe resistance of exch cirout,i be thecarrent read from thesataration curve, and the voltage across each circuit be multiplied by the eo: offcient of In other words, the calculations should be tarred out as though the differential winding were not present, except that instead of using the expression {e, reid) todetermine the forcing voltage, ¢ should be mul- tiplied by (1-44) |, and re by [ren(g+8) JJ. 6, Three-Field Main Exciter ‘The three-field main exciter shown schematically in Fig 13 i of conventional construction so far as mechanical details and armature winding are concerned, but it is built with three electrically independent shunt fields. Field 1 is connected in sorias with a variable resistance across the ain ‘SHUNT, FIELOS iN 2} 35 Reversiae vanineve voutace source stapwizne asus, SeLF-ExcrTeD FIELD RHEO b wrleter ‘eee pre source IWTERLooK HED ARMS Fig. 15 Schematic diagram of three-fleld main exciter. Fleld 1 fa selfexcited and provides bass excitation, field 2 lea ‘separately-ercited controlling field, and feld 3 is smatl- ‘capacity Dattery-excited stabillaing eld.Chapter 7 main torminals of the exciter and operates it of the selfexcited field discussed in Sec. 1. Field 1 provides, ‘the baso excitation for the machine. Field 3 is a small, separately-excited shunt fleld that obtains its energy from fa station battery or any other sourge of substantially constant dee voliage. It is eapable of supplying 5 to 10 percent of the normal total excitation requirements of the main exciter, and its purpose is to provide exciter stability at low voligge output under hand control. eld 3 is used ‘only when the exciter speed of response or range of voltage ‘output maker it desirable, Field 2 is a shunt field that is excited from a reversible variable-voltage d-e eource under control of a voltage regulator. This field also provides for stability of the oxeiter when the voltage regulation is under control of the voltage regulator Fields 1 and 3 have theostats in their enen cuits. These are usually motor-operated under mantal control. The rheostat arms are mechanically connected together go that resistance is added in one field circuit as it is removed from the other. ‘Thus, when the self-enengized, shunt fied is carrying a high excitation current, the sepa- rately-excited field carvies a negligible current. ‘The combined effect of fields 1 and 2 is shown in Fig, 14 and ‘can be explained by assuming that the current in field 2 is zero, When the field rhcostat is adjusted to give a voltage output greater than that reprosonted by the distance Ue, all excitation is supplied by field I, and the relation between ‘tho exciter torminal voltage and the total fickd ampere turns is represented by the line ab. Operation in this region is the same as a self-excited exciter. If the resistance in the cirouit of field 1 were inereased to give a value of ampere ‘urns less han Od in Fig. 14, and if field 1 were the only. field excited, Ube machine would be uustuble us pointed, ‘out in Soe. 1. ‘To obtain a terminal voltage less than Oe, such as Of, the resistance in the solfexcited field circuit would be in creased to reduce the ampere-turs produced by that field 40.0}. These ampere-turns would cause # generated volt age equal to Ok. However, at the some time the current in fiold 1 is reduced, the current in field 3 is increased, and the generated voltage duo to field 3 being energized, is represented by Aj. ‘The ampere-tums of the two Bclds and the gonerated voltages add so that the distance Of is, the total Lerininal voltage. Since the current in field 3 is controlled by the amount of current in field 1 through the mechanical coupling of the field-rheostat arms, the total terminal voltage can be plotted asa function of the ampere- turns in field 1 alone and is represented hy the eurve ekab in Fig. 14. If the field-resistance eharaeteristic of the self excited field is plotted on the same curve, there will always, be a positive point of intersection hetiveen the reeista line and the saturation enrve ekab and stable operation ean be obtainod for any voltage greater than Qc. The voltage represented by Oe is usually loss than 10 percent of the rated voltage of the exciter. Operation at smaller values ‘would not ordinarily be necessary exeept in the ease of &. synchronous-condonser exciter. Smaller terminal voltages are obtained by holding the eurrent in the selfseited field to zero and reducing the current in separutely-exeited fild, 3. Exciter polarity ean be reversed by reversing both, field circuits when the eurrents are zero and building up the manner Excitation Systems 205 MAIN EXCITER TERMINAL VOLTAGE i 1 { a] CURRENT IW SELF-EXCITED FIELD Fig. 14—Equivatent no-load saturation curve of three-field main excieer showing effect of stabilizing fold 3. Field 2 te open-cireulted. in the opposite direction, ‘Thus, manual control of voltage is possille over the complete range necessry When the voltage of the main oxeiter is undor the control of a valtage regulator that varies the magnitude and polar ity of voltage applied to the separately-excited feld 2, ‘the manually-operated field rhecstat in field 1 eircuit is set to provide same hase amount of excitation. This setting i determined by the operator, butis yenerally high enous to supply sufficient field current to the eve generator field to maintain steady-state stability. ‘The current in field 8 is usually negligible with such asettingof therheostatwhen ‘the generatorisearrying any load. The polarity and magni tude of the voltage applied to field 2are then regulated s0 ‘that the flux prociueed by field 2 either aids ar opposes the flux produced by the lace exeitation in fold 1, thas, either ereasing or decroasing the oxeiter terminal voltage. Since ‘the effect of field 1 is that of a conventional self-excited machine, @ small amount of energy input to field 2 ean ‘control the output voltage over a wide range. The opera- tion of tho threo-ficld main exciter is made stable by separ rate means for the twa conditions of operation: by a separately-excited stabilizing field under manual eoptrol, and by the voltage regulator controlling the input to field under regulator control ‘The three-field main exciter has an advantage over the single-icld acparately-escited main exeiter described jn Sec. I in that control of the exciter terminal voltage is, not completely lost if any trouble should occur in the eeparately-excited fied circuit. The trouble might invalve the variable-voltage source for field 2 or the voltage regu- lator that controls it, but even though the eurreatt in the field ehauldl become 2oro, the esciter will continue operating At a terminal voltage determined by the setting of the theostat in the self-onergiaed fold cirvuit. The only effect fon the sve gencrator would be a change in its internal voltage which would cause a change in reactive loading of206 the machine, Under similar circumstances of failure with the single-field exciter, the source of excitation for the a-c generator field would be lost and s shut-down of the unit, ‘would be necessary. 7. Calculation of Response of Three-Field Main Exciter ‘A method of calculating the response of a single-field cexciter is given in See. 2. The method uses step-by-step, intogration to take into account the saturated condition of the exciter. If additional fields are present, damping. ‘currents flow in those fields during voltage changes. Their cifect is to reduce the rate of change of flux in the exciter iron paths. ‘The following analysis presents a means of replacing the assembly of several elds with one equivalent, field so that the response oan be ealeulated. ‘The specific fields involved in the three-field main exciter are the self-excited field 1, the battory-exeited field 3, and the separately-excited field 2 as shown in Fig. 13. “The Unree fields are wound to forin a single element to be mounted on the field pole, so that the mutual coupling is high and can be assumed to be 100 pereent with small ‘error. Also, the same leakage coefficient can be applied to each of the fields. Tn the following symbols the subscript indicatee the particular field to which the symbol applies, Thus, N; io the tune per polo of field 1, Ny tho turne per pole of feld 2, eto. Number of poles, assumed to be connected in series. ‘umber of tums per pole in the field winding. ‘otal useful flux per pole in Maxwells times 108 initial useful flux per pole ia Maxwells times 10% fhange in flux per pole 4’ ~ dn. ‘Total amperes in field circuit, initial amperes in ficld eireuit, Thange in amperes in field winding=i’—ty L=Induetance of field winding in Henrys. K= Flux proportionality constant __ Maxwells 10" per pole ‘Ampere turns per pole = Flux leakage factor= 1+ N. e be ¢ Jeakago fu ‘useful ox ¢= Voltage proportionality constant terminal volts R=Resistance of the complete field circuit, ohms. t= Time constant of complete field circuit, seconds, E\=Torminal voltago applied to field 1. Eqg= Initial value of terminal voltage. ‘Ey= Change in terminal voltege= oltage applied to field 2. E=Initial value of voltage applied to field 2. ‘B,=Change in voltage appliel to field 2. By =Fixed voltage applied to field 3. ‘ a = Differential o ? tal operator $ pate ntl or saadystate vale of total uel ux pee do K(Nitiort Nato+ Nain). (ao) ‘When the feld currents are changed to force an increase in Excitation Systems Chapter 7 terminal voltage, the totel useful fax at any later instant of time is gle KUM Nei +N). an ‘Tho change in total flux per pole is the difference between, these two values, O= 6 d= K (Mit Matt Nat). (1s) ‘The basie formula for the self-inductance of any of the field circuits is wou henrys, and sine the xis expresied as aswell per pole ime 10, the self-induetance of the eireuit of fekd 1 becomes, POND. rr) Iya POND. PRN. ‘The time constant of the field cireuit is the total self inductance divided by the total resistance, whi PKNOA RR Equations similar to Eq. (19) can be written for selfin- ductances Jy and Ly and similar to Eq. (20) for time con- stants é and 4 "The voltage applied to each of the field circuite is stored in drop the ot once a 8p (20) in the circuit inductance. The vollage equations at any instant of time are Bi =e! = Bui +NAPpp! (a) By = Ryid +N Pod" (22) EY =Ris +N Pod (23) During the initial steady-state conditions, when the total useful fx is constant and pdy=0, Ea=cha= Rei tN XPpbe (2) Big Rein + NMP pe (25) Ea Rain} NP obo. (25) Subtracting the two sets of voltage equations, a set in terms of changes from steady-state conditions is obtained. Since the voltage Bi is supplied from @ constant-potential source, By — b= Rei NPS 7) Eas Rain NAP pb (8) Om Raivt NMP Pd (23) It Be (27), (28), and (29) are multiplied by AX, and “2, respectively, and added, the result obtained after suttltating froma Bg (18) and (20) ia Bao? pes pee et (atht+h)ed. (30) ‘earranging the terms in Faq. (50)5 St [| Tpit) Hit ibis @nChapter 7 ‘When solved, Eq. (31) expresses $ and hence the terminal, voltage as & function of time if saturation and the con- sequent change in constants ure neglected. ‘The three fields on the excitor can be assumed to be replaced with a single equivalent self-exvited field as shown, in Fig. 15. The quantities referring to the eq —— te REGULATOR ou ‘ret Teninas, welts VOLTAGE oui aes, ig. 15 Sel-excited atngle-field equivalent of three-feld mat, ‘exciter. field are designated by the subscript ¢ The field has applied to it a voliage equal to the terminal voltage of" plus an equivalent voltage Ey’ supplied by the regulator. During steady-state conditions, Bart bo Reis NAP bw 2) Ab any instant of timo, ee (33) Subtraating Bq, (2) from (83) Bet0b= Reinet NP ob. (4) Using the relations o=KNGy (35) 1,-POND. PRN @5) be on Eq. (34) reduces to. bp aaxe-[0 8) Equation (88) is of the eame form as Eq. (31), and by comparing similar terms, it is derived that, het bleble G9) (40) ‘The selfinductance of the equivalent ficld is given by Eq. (86), and the resistance is ay ‘The applied regulator voltage is, by i Excitation Systems Eliminating 9 by wing Fa. (40) (42) Equations (88) and (1) ean be solved only if saturation is neglected. However, fora small interval of time, it ean ‘be assumed that the machine constants do not change, and. the change in fux ealeulated by either equation will be the same. Ifat the end of the first time interval, the machine constants are appropriately adjusted to’ new values applicable to the next small interval of time, the flux change can be calculated for the second interval and will bbe the same by either equation, ‘Thus, the flux rise cal- culated from the equation for the single equivalent field by using the normal step-by-step methods that take into account saturation will be the same as the aetual flux rise with the assembly of several fields. The various time constants for the machine in the unsaturated condition may be used to determine the constants of the equivalent field, ‘The above equations ean be generalized to the ease of machine having any number of the three types of fields considered. Letting fs, Z- ancl Ny refer to all coils to which, regulator voltages are applied, and t and Ny refer to all coils which are self excited, Eq. (31) in the general form, becomes, (43) where Bf sum of time constants of coils of all types. The sum of the time constants should slso include a value for the frame slab, which acts az a short-circuited turn, and eddy currents in the s For d-c machines of the size used as main exciters, the frame-slab time constant may sppreach 0.2 sceond. The constants of the equivalent eelf-excited field are ‘determined from the following: tad ww) Men 5) 1p is determined by Eq. (80) 8) a cited fields are present in the machine, the only requirements Lo be satisfied are given by Fas. (44) and (4D. Any value of 1. can be used provided the appro- priate value of 2 is calculated from Kgs. (36) and (48). When no selfexcited ficlds are present, the equivalent field is not ecif-exeitod and has applied to it only the rogulator voltage. Hf no rogulator-controlled filds are present, the roquire- ‘ments to be met are given by Eqs. (44), (45), (6), and208 (47), and the equivalent fild is a self-exetted field with no regulator voltage applied Using this equivalent single-feld representation of the multiple-field main exciter, the voltage response can be cealeulated by the step-by-step method of See, 2. The volt= age B is determined by the source of voltage under regu lator control. For example, if the regulated field is a sell excited fiold, the voltage E becomes equal to the exeiter terminal voltage at each instant of time. 8, Main-Exciter Rototrol ‘The most recent development in the field of rotating main exciters is the adaptation af the Rototrol rotating amplifier a2 main exeiter. Any generator is in f rotating amplifier” in that a small amount of eneras to the field is amplified to a large energy outpnt st the generator terminals, However, thename rotating amplifier has been specifically applied to a form of rotating machine possessing an unusually large amplification factor. In such ‘machines, the change in input energy to the field is « small, fraction of the resulting change in energy output of the amature. In the ordinary de generator, the change in, field energy required to produce 100-pereent change in ‘output energy is usually within the range of I pereent to B percent of the machine rating. ‘Thuis, the amplification factor might be between 30 and 100. In the ease of the Rototrol, the amplification factor can exceed 10 depending, upon the design of the machine. ‘The main-exciter Rototrol is not adaptable at present to use With generators operating at Jess than 1200 rpm. ‘The principal field of application is with 3600-rpm turbine gen- erators, The two-stage main-oxciter Rototrol can be built with sufficient eapacity to supply the excitation require- ments of the largest 3600-xpm generator, but when used with 1800- or 1200rpm generators, the maximum rating of generator is restricted. In any event, the Rototrol is, direct-connected to the generator shaft ‘The slower the speed of a generator, the larger the physi- cal size, For a given voltage output, the reduction in speed is compensated by an inerease in the total fux, requiring a larger volume of iron to maintain the same flux density. 2. 16—-A 210-ks, 280-volt, 4-pole maln-exciter Rototrol fot ‘irect-connection to generator shaft at $600 :pA. Excitation Systema Chapter 7 ‘Tho excitation requirements, therefore, are greater for slow-spoed generators. The main-oxeiter Rotatrol bas nat Deen built in expacities large enous to supply the excita- tion requirements of large slow-speed ae generators. Fur- thermore, as the Rototrol rated speed is decreased, its excitation requirements also increase and a lnxger control- Jing energy is quired. ‘The combination of theae factors has largely restricted the use of the main-exeiter Ratotral to direct-conneetion with 8600-rom turbino generators. A 210-kw, 250-volt, 3600-rpm main-oxciter Rototrol is illustrated in Fig, 16, and to all outward appearances it is ‘conventional type of dee machine, ‘The mechanical de {ails euch as the enclosure, brush holders, eommutator, ee, are of conventional 3600-rpm exciter consizuction, but’ the electrical connections are quite different. ‘The armature winding is of the lap form but. has no crost connestions, and there are a number of specally-comnected field windings to provide the high amplification factor. ‘A detailed discussion of the theory of operation of the Rototeot is beyond the scope of thie chapter, and can be fount in the References. ‘The discussion here will be con- fined toa description of the operating principle os it applies to use of the Rototrol in exeitation systems. A schematic diagram of the mait-exciter Rototral is shown in Fig. 17 (a), and the equivalentechematie diagram is shown in Hig. 17 (b). The Rotatrol ean be built with one or more stages of amplifieation, and the main exsiter Roto- trol isof the two-stage type. ‘The field connected between terminals F3-F4 is called the control field, and windings appear on only the two south poles, !and'8. The eireult botwoen terminals F5-FB energizes a field similar to the control fild, and it alzo appears on only the to eouth, poles. This field operates in the same manuer as the con- trol field in controlling the Rototrel terminal voltage but, it is called the limits field. ‘The control field is energized by the voltage regulator and normally has control of the vollage output. However, the Hits field is energie by devices that restrit the maximum or miniaum voltage ‘antpn, s0 that the Finis feld ean, under certain cond tions, overcome the effect of the control feld. ‘The output terminals are L1-L.2, and it should be noted that the circuit between the brushes of like polarity enorgizes additional fold windings that are compensating and forcing feds and also serve a8 series fields. ‘The windings energized by the cireuit hetwcen terminals FI-F2 are shunt-feld windings used for tuning purposes as discussed later. AS far as ex- ternal eicuits are concerned, the main-exsiter Rototrol ean be represented as shown in Fig. 17 (c): the eontrol field is energized by some exciter-voliage controlling device, the Timitsfetd is energized by a device for limiting the max jmue or minimum output or both, and the line terminals supply voltage to the lad in series with the series field ‘The operation of a conventional self-excited dc gener- atorisunstable when the feld-resstance line coinoides with ‘the airgap line of the saturation eurve as shown in Sec. 1. Although this characteristic is undesirable in the self excited generator, itis an important part of the Rototrol principle. Reasoning identical to that in See. 1 ean be applied to a seriesexcited generator where the self-oxeited Winding is in series with the load and both the load and the field can be considered asa shunt aeross the artatare,