0% found this document useful (0 votes)
116 views33 pages

Insulation Coordination

insulation coordination in electrical engineering

Uploaded by

parth gupta
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
116 views33 pages

Insulation Coordination

insulation coordination in electrical engineering

Uploaded by

parth gupta
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 33
CHAPTER 18 INSULATION COORDINATION Original Authors ‘A.C. Monteith and H. R. Vaughan SULATION coordination is the correlation of the in- Tiss of electrical equipment and cireuits with the characteristics of protective devices such that the in- sulation is protected from excessive overvoltages. Thus in ‘substation the insulation of transformers, eireuit break- crs, bus supports, ete. should have insulation strength in ‘excess of the voltage levels that can be provided by pro- tective equipment such as lightning arresters and gaps. The determination of the economic relationship between the impulse strength of equipment insulation and pro- tective voltage level provieled by protective daviees has received and continues to receive a great amount of study. The basic concept of insulation coordination is illustrated in Fig. 1. Curve A is the demonstrated impulse strength of the insulation on a piece of electrical equipment which in operation is exposed to the hazards of lightning surges. Curve B is a protective level afforded by a valve type lightning arrester. Thus any insulation having.a withstand voltage strength in excess of the insulation strength of Ky-cResT fit. 23 € microseconos Fig, 1—Protection of insulation with characteristic of “ protective device with characteristic of “B.” oy Curve A is protected by the protective device of Curve B, ‘Ty protect insulation from excessive vollages the pro toctive devieo must have a lower breakdown voltage. The insulation of electrical equipment in a station or substation is subject from time to time to momentary over- voltages that may be caused by system faults, switching surges or lightning surges. Except for special cases, over voltages caused by system faults or switching do not caus damage to equipment insulation although they may be detrimental to protective devices. Overvoltages caused by lightning are of sufficient magnitude to flashover or break- down equipment insulation and are therefore the most troublesome and of greatest concern to the manufacturers and operators of electrical equipment. Great strides have 610 Revised by: A. A. Johnson been made during the past 20 years in improving the de sign of power systems and equipment with the result that failure of major electrical equipment insulation is rare The problem of providing insulation properly coordie nated with protective devices involves not only guarding the equipment insulation, but also the protection of the devices themselves. To prevent damage to an arrester or a protector tube, each should be applied on a system in such away that it will discharge the excessive vollage safely to ground after which it will couse to earry current to ground, ‘Thus the arrester or tnbe must protect: the equipment insulation and be capable of restoring itself as an insulator against whatever system voltages might exist across it to ground, “Lhe voltage to ground is determined for a system of given voltage largely by the method used for system grounding, the maximum voltage to ground usually being during the existence of a phase-to-ground fault. Rod gaps do not seal off after being flashed over and therefore the circuit must be disconnected from the system to clear gap breakdowns. I. HISTORY Coordination of insulation was not given serious eon sideration until after the first World War, mainly because of lack of information on the nature of lightning surges and ‘the surge strength of apparatus insulation. Since concrete data were lacking on the actual surge strength of insulation or tho dischargo charactoristics of protective equipment, early attempts at coordination were rule-of-thumb methods based on experience and individual idegs. The result was that some parts of the station were over-insulated while others were under-insulated. Also, the gradual increasing of line ingulation in an attempt to prevent linc fashovers subjected the station equipment to more severe surges} and in many eases line flashovers were eliminated at the ex pense of apparatus failures. Growth of power systems, dé mands for improved power service, and more economiesl system operation focused more and more attention on th problems of surge voltages, adequate insulation, and it protection i Thus during the period from about 1918 to 1930 eom siderable work was done by individual investigators laboratories in collecting data on natural lightning and 42, developing insulation testing methods and technique, Although progreae wae seemingly alow, it rosulted in a fal knowledge of the nature of lightning surges and the es lishment of universal eurge producing and measuring 48 vices. Very little correlation between laboratories attempted during that period. \; In 1990, ve NEMA-NELA Joint, Committee 0” Chapter 18 Iation Coordination was formed to consider laboratory testing technique and data, to determine the insulation Jovels in common use, to establisl: die insulation strength, of all classes of equipment, and to establish insulation jovels for various voltage classifications. After ten years of study and collection of data this schedule was fairly well completed. Numerous articles in trade magazines show the results, Tn a report dated January 1941", the com- mittee, now known as the Joint ATEE-EEI-NEMA Com- mittee on Insulation Coordination, rounded out the pro- fram by specifying basie impulse insulation levels for the tiferent voltage classifications. ‘Test specifications for apparatus are prepared on the basis of demonstrating that the insulation strength of the equipment will be equal to or greater Unan he selected basie level and the protective equipment for the station should be chosen to give the insulation meeting thesa levels as good protection as economically justified. Ii, BASIC INSULATION LEVELS Several methods of providing coordination between in- rulation levels in the station and on the line leading into the station * have been offered. The best method is to establish a definite common level for all the insulation in the station and bring all insulation to or above this level. ‘This limits the problem to three fundamental requirements, namely, the selection of a suitable insulation level, Une assurance that the breakdown or flashover strength of all insulation in the station will equal or exceed the selected level, and the application of protective devices that will give the apparatus as good protection as can be justified economically. Data collected from utility systems during the early work on insulation coordination pravided existing insula- tion levels. ‘The data collected (60-cycle wet flashover characteristies measured in terms of equivalent gap spac- tng) fell within well defined limits. ‘The upper limit cor- responded to about ten times By at the upper end of the curve and to about six times By ab the lower end of the curve, E, being the system voltage-to-neutral. ‘The lower limit lay on a curve about four times Eq for systems 46 kv snd below and about three times Ey for systems 69 kv and above. These data together with impulse characteristics of insulation ubtained in thefield and laboratory provided a basis for establishing insulation levels. Impulse test lev- ls, in tarms of inches of gap, wore therefore, selected that ‘epresented a medium between the upper and lower limits defined above and that fell within the scope of available Protective devices. As laboratory technique improved so that different laboratories were in close agreement on test esulte, the toot lovel wore expressed in kilovolts corre- sponding to the test gaps, based on a 1}4X40 microsecond Positive wave, which is illustrated in Fig. 5(a). The basic lbvels were expressed on a 50-50 flashover basis, that is, Yalues in kv erest corresponding to gap spacings giving 80 pereeul Mastiover and 50 percent full wave when sub- lected to 11440 positive impulse. Recognizing that it ¥as not practical to subject most. types of apparatus to § series of flashover tests to demonstrate their insulation levels, @ minus tolerance of five percent was allowed in ‘he definition of basic levels to permit a practical test Insulation Coordination 6ll ‘TasLe 1—Baste IMPULSE INSULATION LEVELS Cola t Colas 2 Column 3 Reference Standard Basic Reduced Clase | Tupulse Level Tnsulation Levels Kv Ky In Use-Kv ie ar 25 oot 5.0 - 5t aT 85t 6 95* not 2 150 35 200 46 | 0 6 350 92 450 5 550 450 1a 0 50 16 750 650 demonstration of accetahility of eqnipment. Finally, in January, 19411, the Joint ATEE-EEILNEMA Com- mittee adopted basic insulation levels (Table 1) in terms of withstand voltages according to the following detimition: “Basic impulse insulation levels are reference levels expressed in impulse crest voltage with a standard wave not longer than 14X40 microsecond wave. Apparatus insulation as demon- strated by suitable tosts chall bo equal to or greatar than the basi insulation level” This requites that apparatus conforming lo Wiese levels shall have a withstand test value not less than the kv magnitude given in the second column of Table 1. It was also understood that apparatus conforming to these re- quirements should be eapablo of withstanding the specified Voltage whether the impulse is positive or negative in polarity. Atmospheric conditions at time of test should be taken into considoration ‘The valuos in Table 1, column 2 were selected initially as the standard basic impulse insulation levels (BILs) to be applied regardless of how the system was grounded. Systems ungrounded or grounded so as to allow full dis- placement of the neutral during line-to-ground faults ro- quire lightning arresters based on the full line-to-line volt age of the system. If the system is grounded solidly or so as to limit the line-to-grotind voltage during ground faults (Xo/X1S3).the so called 80-percent arrester can and has “Thus in some of the voltage clastes of 115 kv am r_of systems have used, with solid groiinding, equipment having insulation with BITS one class lower, as shown in-Table.1,-column 3. ‘On some solidly-grounded systems where the ratio Xo/X1 ‘is equal to about one or less, the one class lower BIL has 612 been used with 75 percent arresters with satisfactory expe rience. As a result of this experience, better overall under standing of the problem, and the economy of redueing BIT, in the higher voltage classes, particularly on trans- formers, the Joint ATEE-FEI-NEMA Committee on In- sulution Coordination is studying the possibility of re- ducing the BIL figures (for Xo/X11.0) to lower values than those shown in Table 1, column 3. Another reason for giving serious consideration to reducing the BIL for solidly-grounded systems is that there are many old formers in service with insulation levels below Usa given in Table 1, column 3 which have given twenty or more ‘years of service without failure. ‘Thus, since the first group of BILs was adopted in 1941, the manner in which the system is grounded has been brought into the picture with the resuit that lower BLL equipment can be protected, thereby enabling systems to be built to do the same job at lea cost: Selection of Basic Impulse Insulation Level The basic impulse insulation level should be selected ‘which ean be protected with a suitable lightning protective device. ‘The best proweetion is provided by modern type lightning arresters.. The spread or margin between the TAIT and the protective act, allowing far maniactring tolerances, is an economic consideration that must balance the chances of insulation failure against the cost of greater insulation strength.- When using lightning arresters the I annnnNEsceeeeeeeeen settee TE than to the equipment insulation, The arrester can be applied so that it will protect the insulation but may under certain extreme conditions, usually unlikely, be subjected to sustained rms overvoltages against which fe eannoto- Cover. Practice has been to apply arrestor so thatthe have an rms voltage rating of a feast Ave poroeut above the maximum possible rms line-to-neutral voltage under any normal or expected fault condition, ‘The BIL of the ‘equipment insulation must therefore be higher than the maximum expected surge voltage across the selected ‘Tollstrate one method fr seestng the BIL of tans: former to bo operated on a 188kv syste, assume the transformer is of large capacity and wye connected on the 138-ky side. ‘The transformer is solidly grounded and the impedance ratios at the transformer terminals are such that Xo/%=2.0; Re/Xv=1.0, Ry/Xi=0.1, Ry= Re and Xr=Xe_ For theso condilons the maximum voltage to ground at the transformer terminals during any type of System foul for any fault resiataneo i 71 poreent af nor mal phase-to-phase voltage as obtained from Fig. 29 (b). Allowing five percent for system overvoltage, the arrester rms voltage rating should be (1.05)(74) or 7.7 porcent which is (77.7)(138) or 107.2 kv. Thus an arrester of 109 hy, whi 8 the closest stands rating, would be ce quired, Curve A in Fig. 2 is the characteristic of a 109-kv Station valvectype arrester for an anmemed TOS mien second wave of 5000 amperes and a plus tolerance of 15 percent on tho average impulse sparkover and a plus tol- Erance of 10 percent above the average drop across the arrester. Assuming a 15 percent margin plus 35 kv between tho 100 ky and the roqused BIL of the transovinc ins Insulation Coordination > Chapter 1g 200; a [8- TRANSFORMER | so B eo Jcannesren 8 | $ 209 201 | 100] ot Te ‘TIME-MICROSECONDS Fig. 2—Courdination of transformer insulation with arrester characteristic. Curve A—100-4v station-type SV arroetor—manimim voltage for 5000-ampere 10 x 20 current wave, Curve B—Transformer insulation withstand characteristic, lation gives 495 ky. Since this value is under the standard ‘of 550 kv, this value can be applied as shown on Curve B of Fig, 2.’ Based on the recommended application valuos for voltage drop across tho 109-kv arrester for a 5000. ampere surge 388 kv instead of 400 kv ean be used, which gives additional margin of protection in 95 pereent of the cases Direct lightning strokes in general have a high rato of vollawe sise (1000 to 10 000 ky per snierusecuad) aad high current values (5000 to 200000 amperes). Such strokes may occur at any point on exposed strnetures whether they are lines or stations. ‘The severity of the surges on station insulation and protective devices largely depends on whether or not adequate shield wires are placed above the structures to intercept the lightning and eonduct it to ground. Without overhead ground wires at stations, dicoet strokes may damage protective devices, thus leaving equip ‘ment insulation without adequate protection. Surges that, originato as diroct strokes on the line and propagate into a station are by far the most common, but are generally easily by-passed lo ground by the ligheiting-provective de- vice. Overhead ground wires above open-wire circuits re Auee the number of strokes that. reach the phase conduc tors as discussed in Chapter 17. ‘The nature of lightning strokes and the propagation of surges are explained in detail in Chaps. 15 and 16. ‘The istics of traveling surges at the station depend upon the nature of tho direct atroke as it originates on the phase conductors, the distance between origin and station, the insulation and electrical characteristics of the line, and the capacitance of the equipment in the station, ‘The surge is attenuated as it travels by corona loss an skin effect, and is distorted by reflection at the station ‘The capacitance of the station equipment charged through the inductance of the line from tho point where the surge originates to the station has the effect of sloping off the front of the surge wave. : The magnitude of the surge voltage that can be im pressed on electrical equipment is not determined by # system operating voltage so there is some argumont Chapter 18 ussociating impulee levels directly with operating voltages, However, low-voltage lines are not as highly insulated as higher voltage lines so that lightning surges coming into the station would normally be much less than in a higher voltage station because the high-voltage surges will flash over the line insulation and not reach the station. Also, the lower operating voltage permits the use of protective dovices with lower discharge characteristics, The insula tion necessary for high operating voltages inherently provides high impulse strength. ‘The impulso levels shown in Table 1, therefore, ean be obtained with the correspond- ing operating voltage class without exceeding reasonable design proportions. IIL. SURGE TESTING ‘The determination of the impulse strength of the various insulations is generally done by an adaptation of the surge generator devised by Dr. Emil Marx in Germany. It con- sists essentially of a group of condensers, spark gaps, and resistors co connested that the condensere aro charged in parallel from a relatively low-voltage source and dis- charged in series to give a high voltage across the test joe. P The only oscillograph available until quite recently for measuring waves of as short duraliva as lightning surges vas the cathode-ray oscillograph devised by Dufour. This, oscillograph was improved hy Norinder thrangh the adie tion of a simple eathode-ray beam control, and today this oscillograph is widely used in this eountry and others. Tn Surge Generator Damping Resistors | Loat Resistor anMAAn Resistor Fotentiol Divider AU Insulation Coordination Deion Gan 613, the Norinder device, the wave shape is revurded on the film in its entirety. Atypical diagram of impulse-testing equipment is shown in Fig. 3. The capacitors, usually rated 100 kv each, making up the surge generator are charged in parallel through resistors. When the charge on each condenser reaches the predetermined breakdown voltage of the sphere gaps separating the condensers, the sphore gaps fash over thereby connecting all the condensers in series. One ter- minal of the capacitor bank is normally grounded. The other terminal must be insulated from ground to with- stand the full magnitude of the discharge voltage. A volt- age impulse of either positive or negative polarity van be obtained by connecting the charging circuit to give the desired polarity. The potential divider shown supplies a reduced voltage to the oseillogeaph proportional to the test voltage. The shape of the impulse wave applied to the test speci- men is determined by the constants (resistance, induct anoo, and capacitance) of the discharge eireuit, some of which are inherent in the eapacitors and leads and some of which are added externally. A typical laboratory in- stallation of impulse-testing equipment is shown in Fig. 4 2. Wave Shape Tt became evident in the early stages of surge testing that. it wonld be necessary to standardize om test wave forms in order to establish insulation levels on a common basis. The accepted designation of defining the impulse fo Test D>— Specimen Electrostatic Potential Divider a 9 i Ennission Tube, wf 5 6OKW Rectifier. 1, | & Timing. System _ =o = Ku Rectifier 3 Relay al 112 Port s Bley |S f 5 | li dsay t | md i 6 Le Gpncent = charging Tx L = | LC Rectifier Wo fw ho i fesording {ban Power-Surge i High fe u 1 roping Phe! re fete Sy fone toa Conrat Sehr * Cathode Ray Panel and Transformer Gseilogroph induction Pegulaton ia. 3—Typical diagram of impulse testing equipment. as Fig. 4—Typical impulse laboratory. Sharon Works of West inghouse Electric Corporation. wave shape is to give the time in microseconds for the impulse to reach crest followed by the time in microseconds for the wave to reach half magnitude. Fig. 5 (a)! For practical reasons a viriual zero time point is estab- lished at 0; and determined by a line drawn through the 0.32 and 0.9E points in the wave front. For example, 14X40 microsecond wave Would have an Or: value of TFinieroseconds and an 0;r1 value of 40 microseconds. In transformer testing where the time to crest is not easily determined, it is taken as two times the interval between he 0.38 aul 0.92 points uu the wave front, that is 22ers. ‘The 1X5 and 1X10 microsecond waves, and other wave shapes, have heen used occasionally in testing insulation \ N v hITIcAL wITHSTANO Insulation Coordination « Chapter 19 However, the 1X40 microsecond wave, either positive or negative, has now been accepted as standard because it simulates the more severe full wave lightning surges anf because it can be obtained readily with the surge genera- tor. ‘The effeet of lightning surges of shorter duration ean bo eimulated with this wave by chopping at short time. 3. Volt-Time Curve The breakdown voltage of insulation or the flashover voltage of gap, particularly the latter, will vary with the length of time voltage 1s apphed. ‘Ihe so-called volt. time curve is ® graph of crest fiashover voltages plotted against time to Aachover for a cories of impulse applica tions of a given wave shape, The construction of the volt-time curve and the terminology associated with im. pulse testing are shown in Fig. 5 (b)‘. The critical of minimum flashover voltage is the crest voltage of the wave that will just cause lashover on the tail of the wave, that, is, it wil eause flashover for 50 percent of the applications, and for the other 50 percent. af the applications there will be a full wave (no flashover). The figure also shows the relation of the critical with- stand voltage. ‘To obtain the magnitude of the voltago, the applied voltage is reduced to just below the disruptive diacharge of the teat opecimen. ‘The rated withstand volte age is the crest value of the impulse wave that the appa ratus will stand without disruptive discharze, 4, Effect of Atmospheric Conditions ‘The flashover charactoristies of insulation In air varies with atmospheric conditions. In general, flashover volt- ages vary inversely with temperature, directly with baro- mettie pressure, and directly with absolute humidity. Test data obtained under various actual weather conditions are usually corrected to the American standard conditions which are: ‘Temporaturo, 77°F. Barometric pressure, 29.92 inches of mereury Humidity, 0.6085 inches of mercury REST FLASHOVER ~-1/ SET, aren cs ——eew 50% OF APPLICATIONS ma or cnest TIME of eRITICAL wave Pon itswoven 4 @ (a) An impulse testing wave illustrating methods of designating sig ilustrating the terminology and defnitions assoelaved with impulse Voltage testing. Wave shape. ae RANGE wo angese— ie Athyn a TIME IN MICROSECONDS o nificant characteristics of the wave, (b) § Chapter 18 Insulation Coordination 615 “AEA Loo ne 7 tel | | I “et ’ E asl le ce i 5 ood i i sod Hf it wo ° 20% ry a6" 3 =” 30a Bets e Fig. 6—Humidity correction factors for fashover voltages of saps, insulators and bushings, based on data from several laboratories ‘Temperature and barometric pressure are usually com- bined into single factor known as relative alr density according to the following relation which is unity for standard atmospherie conditions: 17.95% Bar. Pressure (inches) Relative Air Density’ Se ‘The chart shown in ig. 6° has been accepted, based ‘on an accumulation of test data, aa giving correction fac~ Fig. 7—Impulse flashovar characteristics of standard rod gaps. Long spacings for 1%4 x 40 wave at 77" and 0.60 (8) Fostive waves (b) Negative waves, The ATEN-EEI-NEMA Subcommittee on Correlation of Laboratory Data have published a paper giving a sum- mary of recommended standard definitions and methods applying to high-voltage testing‘. These recommendations aie now generally followed by the industry. 8. Flashover Characteristics of Rod Gaps and In- sulators Because of laboratory differences in test results on ap- paratus insulation in the early days of impulse testing, the rod gap was selected as a yard stick of insulation strength. Because different types of gaps gave different results, a lors for humidity conditions. ‘The measured test voltage is then corrected by dividing by the relative air density “7 tifined above and multiplying by the humidity factor seo obtained from these curves Bod ‘Tanue 2—~TeNtavive AIEE Stanparo ow Insutation Tests | Yon Isoooe Arm Swine, Insvtaron Users anb Bus Surronts — pesol11 Withstand Voltage—Ky 2 soo! menos, 76 FLasnover ta | Low Freq. Impuloe Zaid f yas | Minato Full wave 94] 4 _ 6 Dry (Pos. or Neg. Dry A 25 | 15 ® ACOH 50 19 60 ara Cee 18 26 % i oy aL 3 9% srhewetvents oH 50 uo Fig. 8-Impulse flashover characteristics of standard rod gaps. 23 % 150 Short spacings for 1% x 40 wave at 77°F., 30-inch barometric 5 | 200 1 SR tended tomate he po us om which tha 9-96 a ‘and 0,6u83-Inch vapor pressure, (e) Positive waves. () Nogstive aves. 616 Insulation Coordination ‘TapLe 9—Tentarive AIBE StaNDARD ON INSULATION Tests FoR 2009 Ovrooon Ain Swincubs, InsuLATOR UNITS AND BUS SUPPORTS 1800 ‘Wietstand Voltage Ky veo Voltage | Low Freq. | Low Frea. Impulse ise Rating | 1Min WOSee. 1,540 Full Wave $1200 Kv | (Dey) | (Wet), (Pos. or Neg.) #10005 75 | a | a % Eee 5 0 6 no geo | 0 1150 400} 0 200) | | 100 Al 1 | 145 ene 7 aa Impulse flashover voltages of i, 20 standard rod gape, averaged by the AIEE, EI, NEMA Gust | as 2% ‘committee from resulta of tests from several laboratories, 316 306 | el 680 1300 = sto | 085 1550 ss standard gap was established, ‘Tho following defines the, '*°*| standard rod gap* B add i “The rod gap shall consist of two, 0.5 inch square-comered & | square-eut rods spaced co-axially and overhanging their sup- 2 ports at least one-half the gap spacing. ‘The rode shall be 3"102} mounted on standard apparatus insulators giving a height of 3 fH ‘gap above the ground plane of 1.3 times the gap spacing plus ¥ so0)—| four Inches, with a tolerance of +10 pereent.” Laboratory technique is now developed to the point that 70 apparatus insulation levels can be expressed it lexis of { voltage, However, the rod gap is still sometimes used sod Va 1309 T COI rr re t i Owrs ours 1209 ‘a a t { Fig, 11—Impulse lashover characteristics of suspension inau~ i 1 | latora for 104 x 40 waves at 7PR., Aicinch barometric and ot ‘0.6085-inch vapor pressure. Relative air density= 1.0. $ Att t (a) Positive waves. Broo9 - i sd () Negative waves. g 120 Z 20 t | T COCCI a a ony 5 1 ae vot I ri 00} - dood | fet I Seo} le 8045s oe ee Peri tt & i» Z 00) } [ Fig, 9—Impulse fashover characteristics of 40 C 7 and vertical rod gaps, wet and dry, for Ii 240 waves, water 0, i 4 2 in, per min., 1350 ohms per cubic inch, ! ti if (a) Posie waves A {) Negative waves ‘unre A—Average curve for standard rod gape. Fig. 12—Siaty-cyele fashover of suspension insulators show? B—Test curve for 40-inch standard gap. in Fig. 11. ‘C—Standard gap, wet test DVartioal gap. E-Vertical gap, wot test, Dry: relative air density=1.0, _humidity=0.6085. - Wott provipitation 0.3 in. per min. ; resistance of water=7000 ohms per cubie inch. Chapter 18 Insulation Coordination 617 300 409 Seo0} S300 é e — 3 3 tiv forme] get Ree a z Neate | Cole = S poate ‘Ot 3 4 3 6 To 0% eS 8 Tes 4s 6 7 8 3 ON WOROSECOHOS TO'FLASHOVER WCROSECONDS TO" FL ASHOVER WucRoséconos’ ro’ rLasnoveR (a) oo) o 609 ve sod Cy & $500 $009 1 4 S400 g 2 LI & oo] #700) 4 Fe 3 seal Z 2 LT al 5 g 3 7 ShS"4 § Soa Seca eomve| | US") 200} >> al | Jif TT r 1 “Seat Tres 45 678 %% 33-45 e753 (8 12 34 8 & 7 8 8 0 WT Muchoséconus ru tussmuvan —ieRosEcotos To FussHveR TucRosteoos To FLASHOVER a a) wo Fig. 12 Impulse Rashover characteristics of pa ‘at etan (a) 7.5 ky class (b) 15 ky class. thet information on its fashover characteristics is useful Suspension and apparatus insulators play an important part in the coordination of station equipment, not only in ‘stablishing the insulation lovel but also in determining the magnitude of surges entering the station. Suspension insulators are generally made up of at least three ten- inch wnits in a strug, spuced 5} inches apart, Apparatus insulators can be either the pedestal type or the so-called Dost tye. \ complete résumé of impulse and 60-cycle flashover chatacteristies of rod gaps and insulators was published by P. HL. MeAuleyt. For convenience some of these data ure reproduced in Figs. 7 to 17. Figure 9 is of particular terest in that it shows Uke effevt uf mounting and at- ‘nospherie conditions on the flashover characteristics of rod gaps. ‘The voltage distribution across strings of standard sus- 200 T 1 ™ | bwade uber \ [| 8 iculas types uf apparatus Insulators on positive and negative 1¥4 x 40 waves ied air condition: (c) 28 ky class. (4) 84.5 ky class (e) 46 ky class 69 ky elas pension insulators of various lengths is given in Hig. 18, ‘The data for 10 to 18 insulators were obtained by labora- tory tects by Soroncen**, ‘Table 2 gives data from a Tentative AIEE standard on the 60-cycle and impulse withstand characteristic for indoor air switches, insulator units and bus supports. Table 3 gives similar data for corresponding outdoor in- sulation. |_Iwcnosec:ro'rpAsnoyen we ey Sng $ d sol >t Scoot 3 CO erie 8 500 ills seq nore q sive are oa i a a 20 22 MIGROSECONDS TO FLASHOVER \cterlatics for 88 kv class, igative 14g 4 40 waves wi ‘standard air conditions. 15—Impulse flashover characteristics of two to seven apparatus insulators, for 1¥j x 40 waves at standard ‘ir conditions. (2) Positive waves, (©) Negative waves. 2 3 618 Insulation Coordination Chapter 18 3 309, se 200 - 5 . | 5 Beco Bes tty groot + & wi ; 3 F | Si bead pes bso a 3 [| nse 3 +) 1 8 Seo | IS 3 Lt 2 eo} | 20 3209 WET Pe [OsiTve Tt TI “| a nC Ce Cea MIGROSECONDS TO'FLASHOVER MICROSECONDS TO FLASKOVER MICROSEGONDS TO FLASHOVER to) °) e) | 7 ¥ s00— + 300] pe : C 4 seal 3 fe bes TREGATIVE 2 600] E tot 5 g . 3 ex cI Seed] | Seop NE 2 t : t z Ly }{ | iN} ft 4 eosmnive | wos — | 180; " 2005 I | 35 Be 55S MIGROSEGONDS To FLASHOVER WGROSECONDS TO FLASHOVER MicROSECONOS To FLASHOVER @ es) too 100 Tey ey eo | 5 Seal Soop} | 5 -S 00] 3 C le] 2 8 5 Ares) § i i we) L ‘ Peo 2 7 7 | ds r00 3 k - 3 [TT] 3 Sao { 8 Ltt TY Ssod - Lnesarve|_[_| - TIT try * = i positive] festive [| SEE ieanve 20 sof {1 Tee ee Se 0 25 wicROSECONDS TO FLASHOVER ibnosecbnos 19 FLASHovER MicROSECoNDS To FLASHOVER @ Fig. 16—Impulse flashover characteristics of particular sizes of pin type ingulators for posi Since the voltage-time curves for various types of insu- of the insulators shown do not meet the required impulse Intnrs are not available from the standards, the eurves in withetand voltago. Figs. 13, 14, 15, 16, and 17 are given even though some (ny a ive and negative 114 x 40 waves at ‘standard air conditions. 6. Impulse Characteristics of Transformer Insu- lation Because a power transformer is usually the most expe sive equipment in a station and Lecuuse its failure may mean a lengthy and costly outage, it is investigated most critically from an insulation standpoint. SGATIVE aa by the breakdown voltage of the major internal insulation ‘The impulse level of a transformer can be determined (insulation to ground), the breakdown voltage of the minor CI 5 +5820} > F600 | g J Tecearve| 2 1 § 20 ee S400} —teesiTIvE} insulation (insulation between turns and windings), 9 {egsmve | ation (insulation between turns and windings), af MIGROSECONDS To FLASHOVER 7 t Fig. 17—Impulae flashover characteristics of line-post inaula- lation in a transformer differs from flashover in air im tors for positive and negative 14 x40 waves at standard alr main respects. First of all, the impulse ratio (the F™l ‘conditions. EE Te the ashovor voltage of tho bushings, or a combination of these. ‘The impulse characteristic of the internal inst f 60, of minimum breakdown on impulse to breakdown 0” Chapter 18 Fig. 18. Power frequency yoleage across eacls wii it ass insu lator string starting with insulator No. 1on the grounded end of the string. Insulator No. 5 in a string of 10 has 7 percent of che total sting voltage. (Without grading rings except ‘where noted.) cycle peak) is higher, being from 21 to 22 for trans former insulation, whereas, it is 1.5 or lees for rod gap, insulators, bushings, ete. 'Secondiy, the impulse break: down of transformer insulation does not vary as much with time as seen from a typical volt-time curve, shown in Fig. 19%, After three microseconds the breakdown volt- age is substantially constant ‘The insulation stress between turns or between coils in a transformer is deponclent lnzgly upon the stcopness of the surge wave front, It may be further aggravated by cecillations within the transformer or by a “piling up” of the surge voltage in a small portion of the winding, (See Fig. 20%.) Modern transformers are designed, however, 80 180 ol so] Shs I Ho} 4 — oo} FR _ PCP micRoseconos Fig. 19—Volt-time curve of typical major insulation in trans- formers. Insulation Coordination 613 | on 3 [23 wenosecowe rrour | |_| 2" [jis wenosecona| Fawr : a4 + = ANNI god WARE # LAX {tf oes HH ol 4-4 { — _t c 4 LJ al J aaa PERCENT wiNoING Fig. 20Effect of wave front on initial voltage distribution in. ‘some types of transformer windings. that the minor insulation meets all the requirements af applied impulse tests. To demonstrate this, modern trans formers usually must be capable of passing a chopped wave est of a higher voltage erest than the full wave test. This ‘hopped wave is produced by flashover of a gap or bushing in parallol with tho transformer insulation, ‘Phe standard impulse tests for transformers, regulators, and reactors for the different voltage classifications as standardized by the American Standards Association C 57 are as follows: Standard impulse tests consist of two applications of aw chopped wave followed by one applteation of a full wave, Either positive or negative waves may be used. (a) Chopped-Wave Test (1) For this test, the applied voltage wave shall have a crest voltage and time to flashover in accordance with Table 4 (2) The chopped wave shall be obtained by flash over of a suitable air gap. kiLoveuts BUSHING ! l i WicROSECONDS 21—Typical volt-time curve of transformer winding and bushing (heavy solid line represents the overall volt-time curve of transformer, co be used when protecting against lightning surges NLS MEEDAT 1S, ‘Tabun 4—Sinnuany Inrusse Testo Fon Taanaruanens, Reu’- ‘Tablis d—Graxpano Wrrhsranw Taos Vousauss um ArraMatlg LATORS, AND REACTORS "BUSHINGS 7 es Impulse Test MS Kv second Full Wave | O1-Type ‘Transtormers 900 | Cil-Type Transformers over RAS Kv 2) Great Ra) Kvaor Les 500 Kva — Oil-Type Reg- ; aoe strument lating Transformers — Oil-T'ype Constant-Cur-| Oil-Type Current Limit. Ingula. |_ Outdoor Bushings | rent Transformers—Step ing Reactors—Step and tion maou Inaulae| Resulators 250 Kva or) lators Over 250/Kva Sin- ation | Apparatus | Apparatus | In nzuls-) TeaeSingle-Phaseand 750 gle-Phase and Over 750 (1) | awed door tien | Kiva or Less dhree-Vhase) Kiva Three-V’base BY a Cha ine 5, ] 1 Min! Large Small| jngy” i 6 Full oy | a8 nen Chopped Wave Chopped Ware | shave 1 atin] oe afin ose] 42, | ©) | t——— |———| Dry | Wet Dey] Wet reales Pa xin | a Time Time os oe es Kv |toPlash-| Ke | Xv |toFlahe| Ky 12 | 6 | 2 Creat | overin | Great | Creat | overin 25 | ai | a0] 15) as] 20 | o} a5] 4 Micro | | Meroe 50 s| alo second seconds 87 3 | 7% t— 6 5 12 [a | 10 wu [is | a tee os | oc | aas o | 15 | 6 5 to | 150 50 | | 15 s | 16 | 35 sist ena [tase se) a | 16 } uo | vs | 95 fa ao 3 | uw | as mw | 20 uo es Es 0 | ms | 30 | 150 | 1 | 30 fa a5 | 230} 30 | 200 | 2 | 30 50 ao | 20 | 30 | 250 | 20 | 30 | 250 : os eo.0 | to | 20 | 350 | 100 | 20 | a0 fa a we | 50 | 30 | 450 | 520 | 30 ey ellen ey us | ev | so | sv | ow | so oo | oe | ae 1080 133 | 70 | 30 | 0 | 70 | 30 cortaialfeso Pane yic50 1 | a6 | 30 | 750 | 85 | 30 245 | aio | 665 | [1550 re | 1035 | 3.0 | 900 | 1095 | 3.0 | 900 1) Bunbingpots even ipulatonclaicasin aren general recommended ow amo | 20 | so | ia | 8.0 | 2060 Ce ered yi | 1500 | 30 | 1900 | 1500 | 30 | 100 ieee rari ean ais | ates | 3.0, | asso | ames | 3.0 | 1550 (b) FulleWave Test For this test, the applied voltage wave shall have a erest Value in accordance with ‘Table 4 (c) Excitation During Impulse ‘During the impulee teat if the transformer is excited at normal voltage and frequency, the impulse shall be timed within 30 electrical de- agrees of the crest of the normal frequeney volt- age of opposite polarity. The test. vulues for the different vollage classifications are shown in Table 4. Since the bushing represents a vital portion of the trans- former insulation, its impulse fashover must be carefully considered in establishing the transformer insulation lev- els, The standard withstand voltage tests for apparatus bushings as given in ASA C 76 Standard 1943 are listed in Table 5. rd ering ere (7) Burhings for te fa lodoor appara oie i indoor ye Ste ai aay cat eter ara (©) Bashing tor anal indoor appartan may ve ape to withstand » fw fe ‘The volt-time characteristics of the bushings on a trans former differ from the volt-time characteristics of the transformer internal insulation, In general, the bushing will have a higher fushover at short. time lags than the transformer internal insulation. At long time lags its flash- over may be slightly more or slightly less than the winding breakdown. ‘The impulse strength of the winding is essea- tially the same for positive or negative waves; whereas the bushing eritial flashover may be higher for one polas” ity than for the other. The manufacturer takes the over jimpulee characteristics of a transformer into account WI Chapter 18 Hg. 22--Power transformer undergoing Impulse generator is in building in background. xiving its withstand voltage characteristic. vndergoing an impulae t A transformer st ia illustrated in Fig. 22, 7. Impulse Characteristics of Other Station Appa- rats In addition to power transformers, the outdoor station senerally has instrument. transformers, circuit breakers, iigeonneet switches, and bus insulators exposed to light= ning eurgos. Some stations will also include reactors and relating equipment. All of this equipment now meets the basie impulse insulation levels listed in Table 1 ‘The standard withstand impulse tests for instrument transformers, regulators, and reactors are shown in Table 4, referred to above for transformers. The withstand impulse tests for outdoor circuit breakers, disconnect stitches, and bus insulatore ara the same ae thoce listed in Table 5 for outdoor bushings IV. CHARACTERISTICS OF PROTECTIVE DEVICES ‘The purpose of a protective dovice is to limit the surge voltage that may be applied to the apparatus it protects and by-pass the surge to ground. Tt must withstand con- tinuously the rated power voltage for which it is designed. ‘The ratio of the maximum surge voltage it will permit on lischarge to the maximum crest power voltage it will withstand following discharge, called the protective ratio, is, therefore, measure of its protective ability. Of great importance ‘also is its ability to discharge severe surge Currents, either of high magnitude or long duration, with- out injury ‘There are three general types of lightning-protective ‘levies, each having its field of application; namely, the Insulation Coordination ea1 rod gap, the protector tube, and the conventional type lightning arrester. Rod Gap— Although tha rn gun hos the avantngn of being extremely simple and rugged, it has two impor dlauvantages froma protective standpoint. Tirst i does hot fulfill one of the requirements of a true protective device in that it will not valve off power voltage after it hhao once been flashed over by a strge. valve ‘The circuit must be decnergized to clear the fashover are each time the gap operates. Second, its breakdown voltage rises more at short time lags than most insulation, whieh means that a relatively short gap is required to provide protection against surges having steep wave fronts. It would thns have a low flashover at long time lags that would result in numerous flashovers with consequent outages resulting from minor lightning surges or severe switching surges. ‘The rod gap is, therefore, generally used only for back-up protection or on circuits where the outages with short gaps cun be tolerated or compensated for by high-speed reclosing of the eireuit energising breaker. Modifications of the rod gap, such as the fused gap and control gap, havo boon uced occasionally. The fuce gap imply a rod gap with a fuse in series with it to inter- rupt the power follow current eaused by the Aashover It, therefore, has the same surge protective characteristic 909; oe ee 8 MicROSECONS Fig. 23—Impulse characteristics of transmission type pro- tector tubes for grounded-neutral circuit 622, Insulation Coordination ‘Tape 6—InvusTaY PERFORMANCE CHARACTERISTICS OF DISTRIBUTION EXPULSION-TYPE LIGHTNING ARRESTERS Front of Wave navet |___ tape Sparano ver Ct Artester | Rate of Rise ker | oxo Mi a pte il ve Mierowre. | ytin | avg | Max k 6 m | m|% | 2| >| & [slale 2} ow feta fo | sini eh pully ite bis sprkover voltage as the plain rod gap and, although it prevents a cireuit outage upon flashover, it requires the replacement and maintenance of fuses. Alko, to be effective it requires proper coordination between the fuse blowing time and adjacent relay timing ‘The control gap’, consisting of a double gap arrange- ment to approach a sphere gap characteristic, has a some- what belter volt-time characteristic than the rod gap. Tt can be used with or without fuses, and although it is ap- plicable for back-up or secondary protection, it is usually considered in the same class as the rod gaps, as far as major protection is concerned. Protector Tube—The transmission type protector tube has a volt-time characteristic, Fig. 28, somewhat better than the rod gap and has the ability to interrupt power voltage after flashover. It is, therefore, used extensively to prevent flashover of transmission line insulators, dis. “connector FoR Line conductor WATERTIGHT JOINT L—_senies cars wer process: sovoen-seaeo— ~\ Porous eLocKs-—~ VALVE ELEMENT CONNECTION (IN BACK Nor visiate) Fig. 24—Sectional view of station-type lightning arrester. Chapter 1g re ay (b) rrr aft 5—Typical oscillogcams of current diacharae est Tighening arvesters ree a (a) Standard 10%20 mioroaveond current surgr applied 19 ky station type arreste (b) Arrester voltage during discharge of current sure® (©) Curtont and voltage of 3 kv station type arrester ing a 5310 microsecond current surge having & ‘excess of 100 000 amperes, Chapter 18 Insulation Coordination 623 ‘TABLE 7—PERFORMANCE CHARACTERISTICS OF VALVE-TyP& LIGHTNING ARRESTERS eoner Front of Wave Impulse Sparkovir | Discharge Voltage-Kv on 10,220 Micrneeond Currnt Wave® Type and ~ 5 Mele] peo ia ky 5,000 Amperes 20,000 Amowres voagenace | pune Bee aa eal | Ave. | Max. ave | tex | t moe | r 3 w| 2 uf oa] oa | | 2 é 5 a | 46 | | a miu) a 8 |e x] a | at a | a | os ” a) wo | a2 | oe a | 2 | 7 B |] a al} a| 7 a | *| By 38 208 ve | au wor | nie | ait } iss | iat | 130 » 20 up | is fase | ag | iss | j uo | ao | a2 Fd octet Ricoh cca | ie | um | ise | {tar | ist | ao ‘0 Se] sr | j 3 | ss | ar | ros | 96 | 225 50 a7 | 8 | a0 | 208 | 209 | 299 | 98. as | 21s | 280 wo soo | 220 | 204 | 250 | aie | are | 267 am | 208 | 33 | oon | ar | sm | am | aor | ee | om | del | 00 | a | sain | \ } | | | ef ow fl wf] | mf nf don male » | a 8 | | | a] a | | 2) m2] fe) om] oa] a > | % | a] x] a | oo] a) | a 3% | a | 3 | a 2 m | af of | o} af oa | a fel ela) 5 wm | ela) eo) se] a} aw) @) |) | | | { | 2 wr | | 83 | 0 | or | m| 2] 2 | | a | ww] a6 | st : x | we | ae | nit | 100 | a0 | sor | aos | nie | ne | ata | aso | deo B 30s rat | vor | rae | aa | tar | tas | age | awe | at | as | too | a5 © x3 vas | ar | aso | ase | ans | asa | as | aso | das | asa | doo | aoe x0 a wre | 205 | 196 | sor | use | azo | azo | sor | ist | aot | an | 205 nr) ata | 266 | a6 | a00 | 29 | at | arr | ag | sar | sat | asa | aco 73 | aos 2er | 300 | 288 | 25 | 270 | 252 | 202 | 288 | 270 | 288 | aia | sos a | ee a5 | aor | as0 | axa | 355 | 305 | ai | ase | aa | av7 | ais | oo wy foe] sa fase | sar | a6 | son | ase | aoe | ce | a0 | set | aor | as st | ume | a | ans | are | ame | ate | sae | ase | ase | ane | oo | ar | us| ime sis | sa | 500 | as7 | sas | som | sa | ozs | se | 500 | oon wo | aos | orm | ons | ass | sos | anc | aos | co | ot | os | axe | sas 190 1033 sor | two | geo | oar | cis | ot | ws | vos | iat | oo | gy | 242 2017 1 860 | 988 | 945 | 806 | 887 | 860 | S72 | 960 | 931 _| 940 | to35 | 1004 “Pat paar ede cher specLover loge tonnect switches, and bus insulators. It is also used on bransmission-line towers adjacent to a station to reduce the magnitude of surges coming 1n on the line and thus ‘lieve the duty on the station arresters. ‘The tube is not A the present time considered adequate protection for unsformer insulation, except for distribution type ratings 15 ky and below. Its application on cireuits ahave 18 kv ‘auives certain limitations in system short-circuit currents ‘nd recovery voltage rates ‘The protector-tube principle is used extensively for ex- sion type arresters in the distribution classifieations ‘2 ofthe arate wanvfactuted wil uve earaginticg wot oxen the lieth ola. For dsbaton sere the main alse 15 kv and below. Industry performance characteristies of distribution expulsion-type lightning arresters are given in ‘Table 6. Valve-Type Arresters—The conventional valve- type lightning atiester, typical example of which is shown, in Fig. 24, provides the highest degree of protection of all protective devices. Its essentially, fat, valtstime. charac teristic makes it ideally suited for the protection of trans- forme® insulation in the higher voltage classes where the margin between operating voltage and surgo strength is relatively low. If properly applied, its discharge voltage 624, ‘Tante 8—IxsuLation Tests For LichesinG ARRESTERS (Withstand Test Voltages) Station-type Arrestors pe Att ratings ne and Ditton Ar- | Line and Distribution- | type Arresters rated. tus [rte igpe area rated | ow 9 Ke aoe eer | fen | at Tmpuise | o-cyle | Impulse ier | He [Tenover | 1230 ys [ ow eh | 0 0) en ave je Fu ae 1 Min) 10See) Crest Kv 3) | Dey | Wet | 23) ® [sya | 6 5 2 | 6 % a | | no | 0 % | wo | 200 ts | 50/9 | io | 30 «9 | 2| 30 2 | * } 50 | us | az 50 ae | us 0 at | te | 385 | a5 | too | | 19 | 195 | 365 | 385 | 900 2a0 | _a4a | 545 | 115 1050 (0 Shere apt ig tobe made of a ete Nalng ower ola ing ‘hum one step lower than the rated careuit voltage ay MATINEE SEE at coltagen without neato tleranen ser fotvraemtive plant wanes nay Bets oheiever evs the romains below the breakdown strength of the transfarmer insulation, even at short time lags. Experience with actual lightning discharges and laboratory tests have demon- strated the ability of the modern lightning arrester to discharge surges commensurate with direct strokes of lightning, Lightning arresters for a-e power circuits are rated ac- cording to the maximum line-to-ground circuit, voltage they will withstand. There are three classes available; namely, the station type with voltage ratings ranging from 3 to 242 ky, the line type, for 20 to 73 kv, and distribu- tion type, 3 to 15 kv. ‘The characteristies of these arresters, ara given in Table 7. Station-type arresters, as distinguished by their heavier construction, better protective charaeteristies, and higher discharge-current capacity are used for the protection of substation and power transformers. Line-type arresters are used for the protection of distribution transformers, small power transformers, and sometimes small substa- tions. Distribution type arresters are intended primarily for pole mounting in distribution circuits for the protec- tion of distribution transformers up to and including the 18.0-Ky classification. ‘Modern station-type arresters are designed to discharge ‘not lose than 100 000 amperes; line and distribution typos not Jess than 65000 amperes, each with a 5X10 micro- Insulation Coordination ‘ Chapter 1g second test wave. In addition, they are given an insula, tion test in accordance with Table 8, which is from ASq C 62 Standard dated April 1944. ‘The valve-type lightning arrester is usually made yp of iva elements, a gap element capable of withetandiy power voltage and a valve element capable of suppressing the current following the discharge of the surge. The breakdown of the gap, which is affected somowhat by the rate of voltage rise, determines the initial discharge voltage of the arrester. ‘The voltage drop across the valvg element, which depends upon the rate of rise and magnie tude of surge current discharged determines the arrester voltage during discharge. ‘Typical oscillogeams of arrester current discharge tests are shown in Fig. 25. ‘The protective ratio of a modern lightning arrestor ig substantially constant through its range of voltago ratings which means that the gap break-down voltage and the maximum surge discharge voltage for a given surge con. dition are approximately proportional to the voltage rating of the arrester. The curves of Fig, 26 show how the gap breakdown varies with rate of voltage rise, and the curves of Fig. 27 show how the discharge voltage varies with the magnitude and rate of rise of surge current, for typical line type and station type arresters. From these curves, expressing the gap breakdown and discharge voltage, each in terms of kv per kv of arrester rating, it is possible to determine readily the protective characteristies of any rating arrester for an expected surge condition. STANDARD ALEE TEST WAVE 100 KV/MS PER 12 KV [ARRESTER RATING PTT t ‘GAP BREAKDOWN, ARRESTER RMS YOLTAGE RATING BREAKOOAN IN Ki PER KV OF v z 3 + % Fig, 26—Average impulse gap breakdown of station- and late (a) Represents rate of rise of 5 ky per microsecond per Kv @ arrester rating. apd (b) Reprosents rat of rise of 10 ky per mieroseco arrester rating (0) Representa rate of rae of 90 ley por misrorecond pet Et arrester rating Chapter 18 a Insulation Coordination 625 130; Et 420000 | Ltt | es tit T+ $900 Eo 79% eee TT “TH Jstanoao aicel nr L [lox 20 wo Tear wave] TOO: LITT | ‘stanoano atce|_|_ ox 20 ms test wavel | [| 1000 2000 3000 5000 SURGE CURRENT IN AMPERES (b) 7060 20000 46000 Fig. 27—Average discharge voltage characteristics af typical lightning arresters. (Numbers on curves represent rate of rise of current in amperes per microsecond.) (4) Line type. () Station type. For example, suppose it is desired to determine the gap breakdown voltage and maximum discharge voltage of a 73 kv, station-type arrester for a surge wave front rising ata rate of 500 kv per microsecond, and a discharge cur- 2000 amperes with a maximum rate of rise of 2600 amperes per microsecond. A voltage rise of 500 kv per microsecond with a 73 kv arrester, represents a voltage rise 300 6.85 kv per microsecond per ky of arrester rating, From Fig. 26, the gap breakdown voltage would be 3.X73= 263 ky, at 0.5 microseconds. The maximum dis- harge voltage frum Fig. 27 (b) would be 8.273 =254 kv, Y. APPLICATION OF PROTECTIVE DEVICES 8. Selection of Arrester Rating A valve type lightning arrester begins to dischargé at 4 definite value of overvoltage in accordance with the carve of Fig. 20, aul valves off at a lower voltage, corre~ Sponding to the maximum permissible voltage rating of the arrester. If tha power voltage is shove the valve-off Voltage, the arrester may continue to discharge power Current until it is destroyed. Although modern lightning arresters will discharge with- \z-10-GROWND voLTAGE Loo 2 waxy x VALUE GIVING MAXIMUM VOLTAGE [TTI | ~ao =e0 ag —=25 9 Xo/X Fig, 28—Maximum line-to-ground voltage at fautt location for laolated-neutral systems during fas Values shown are maximum values for single line-to-ground fault For double line-to-ground fault the voltages are less for ratios of %o/Xi between —= and —2, X=zero-sequence espacitive reactance and 2X, positiveseyuense aubtiausient reactance, ogi out injury any lightning surge except the most severe direct strokes originating close to the arrester, it is not practical or economical to design them to discharge power current for any appreciable time. A lightning discharie’ may reach thousands of amperes, but the time is short, being measured in microseconds, so that the energy that, is absorbed by the arrester is small compared to the energy that would have to be absorbed with a few amperes power fiow for even a few cycles. ‘The first consideration in applying an arrester should, therefore, be the maximum line-to-ground dynamic voltage to which the arrester may be subjected for any condition of system operation or fault. High operating voltage may exist on the far end of long, high voltage, unloaded transmission lines because of charg ing current ‘ilowing through the line reactance. It ean also be caused by the sudden loss of load on water-whee! generators. It is sometimes posible to rearrange the switching scheme to eliminate or at least minimize the possibility of overvoltages from these sources. ‘These fac- tors, however, must be taken into account in the appli- cation of lightning arresters. The maximum rms line-to-ground voltage during a sys tem fault ean be calculated by the methods of Chap 14, taking into account tho constants of the system, the type of fault, and the fault resistance. ‘The selection of the arrester rating should, where possible, be based on such calculation. Where the fault voltages are not determined more accurately by calculation, Fig. 28° and Fig. 29", ccan be used as guides in selecting the arrester rating. ‘The curves of Fig. 28 show the maximum line-to-ground voltage during fault for isolated-noutral systems as a fune- tion of the ratio of zero-sequence capacitive reactance, X,, to positive-sequence inductive reactance, X1. In Fig, 29, applying to grounded neutral systems, the ratio of 626 Tnoudution Covritinution ~ Chapter 1g 1 | : d |__|0g f ( | s V| { | Cn | TSA ak 1 > Sy iy 9 @ s ¢ . Xo/% Kalk (a) Voltage conditions neglecting positive- and nogative- _(¢) Voltage eonditions for Ry= 2 a cae oa) Fig, 20_Maximum line-to-geound voltage at fault location r > for grounded-neutral system under any fault condition, Note: Numbers on curves indicate mavimam line-torgrosnd fault voltage of any phase for any type of fault in percent of unfaulted line voltage for area bounded by eurve and axes, When using the curves all impedance values must by it die stane kv base oi ohms on samme voltage bse. For all curves Rymzeresanquence resistance, = zeto-sequenes inductive reactance, postive-sequence subteassient reactance, “Xy-noyativesequenee reactance, Keke ‘The effect of fault rositance was taken into account, ‘The fault r= sistance which gives the maximuia voltage to ground on any phase ‘was the value used. ‘The discontinuity of the eurvesis eaused mainly by the eftect of fauit resistance, ‘quence resistance, Ry to Xi, is plolted against Xo/X, for several different values of maximum line-to- ground fault voltages, ranging from 65 to 100 percent of line voltage and for three values of ,/Xy, namely 0, 0.1 and 0.2. The area below each curve represents the regio it 'which Che maximum fault voltage is below the value indicated on the curve. ‘The curves represent, the max 1175 mum voltage at the fanlt location. A:fault at the arrestet 2 ill generally subject the arrester to a higher voltage thal tT Y ‘a fault at some other point in the system. However, st condition does not always exist, and should be checked oO t 2 s 4 For example a fault near the source of a radial feadet Horm eireuit grounded at the source only through © Re op Xt Oe ne eee resistor or reactor might have a larger value of 51 Chapter 18 and therefore produce a higher voltage on an arrester Jocated at the end of tho feeder than a fault at tho ar rester location, In applying arresters, it is customary to make an allow ance for operation at a voltage in excess of that usually considered as normal. ‘This is usually five percent. For example, an arrester rated at 105 pereent of normal line- turline voltage is used where the line-to-ground voltage tr expected to reach normal Ting-to-line voltage during fault. Likewise, on a solidly grounded neutral system ‘where the maximum Tine-to-ground voltage during fault is expected not to exceed 80 percent of line voltage, an arrester rated at about 84 percent. of normal line-to-line voltage has been used generally. The line-to-ground volt age for fault conditions where X4/X, ratio is near 1.0 of iesimuy allow tho ueo of arroctors with loo than $1 percent of line-to-line voltage. “There are a number of isolated-neutral systems on which arresters rated at 105 percent of nominal line-to-line volt sige have proven satisfactory over a period of years. How= there are also systems of this type where it has been necessary to use arresters of a higher rating to prevent txevssivo failures. This is indicated in Fig. 28, which shows that if fanlt resistance is inchuded, the fault voltage may exeeet! 105 percent of normal line-to-line voltage. lation should be made to determine the maximum fault voltages. ‘The overvoltages encountered on systems grounded, Uhrough ground-fault neutralizers are less than on isolated nouteal ystems, provided the ground-fault neutralizer is properly tuned.” Arresters rated to withstand maximum, line voltage, usually 105 percent of normal eireuit voltage, are, therefore, considered safo for application on these systems, ‘There is some risk of damage to the arrester if the grotind-tault neutratizer 1s not properly tuned. Switch- ing operations on these systems may also produce high voltages to ground. However, it ia generally not feasible to select arrestors of sufficiently high rating to eliminate all risk of arrester damage from these eauses. It has been common praetiee to apply arresters rated at 15 percent of cireuit voltage to systems grounded through, impedance, and arresters rated at 84 percent of circuit voltage (80 percent of 105 percent rating) to systems con- Sidered solidly grounded. Experience has shown that such applications are generally safe against over-voltage at time of fault. However, as indiented by Fig. 29, the possible Jine-to-ground voltages during faults on systems vary through a wide range, depending upon the ratio of system constants. Arresters’rated at some voltage between 75 and 105 pereont of cireuit voltage may, therefore, be better suited from an overall standpoint. As indicated by Fig. 29 (a, b, ©), the maximum voltage to ground varies with the ration af Ro/X1, Xo/%1 and RNs. Thus the voltage to ground can be determined for a given system if the impedance constants are accu- ‘ately known, In some cases, particularly for the higher Voltage systems, where the Xu/X; and Ro/Xy ratios are 10 of tess, auventers Tess tun 64 pereenl of line-to-line ‘Voltage ean be used, thus allowing the application of trans- former insulation with a minimum. acceptable impulse ‘insulation level. Insulation Coordination 27 ‘The 8t-percent arrester can be applied safely on systems whose constants are within the range indicated by the 80-percent curve of Fig. 29 provided the impulse insulation level of the equipment is protected. As a general guide to arrester application, with full insulation on the pro tected equipment, the 84 percent arrester rating is satis factory if the following conditions exist. 1. ‘The ratio of the zero-sequence resistance, Ro, to tho positive coquence eubtransient reactance, Xi, a8 viewed from the point of arrester location is one ot Jess, 2 The ratio of zero-sequence reactance, Xo, to positive sequence reactance, X;, as viewed from the point of arrester loeation does not exceed three under any condition of operation. 8 The arrester cannot remain cnergizcd from un- grounded sources of power after the grounded new tral sources of power have been disconnected to clear a fault. 4, The system neutral is grounded at every souree of supply ot short-circuit current. i the fault is to the arrester ground, then the resistance of tho arrester ground should be included aa part of the zero-sequence resistance of the system. When this is done, the curves of Fig. 29 also apply to the arrester at, that fault location, In addition to high arrester voltages resulting from system faults, high momentary or peak voltages may also be caused by any of the following: 1, Switching surges may reach several times normal line-to-ground voltage with certain combinations of system constants, 2. High harmonic voltages to ground may exist during fault conditions on lightly loaded lines energized from generators with damper windings for whieh X('"/X4" is too great, See Chap. 6, 3. Ateing grounds or the accumulation uf stabie charges from dust particles in the air on ungrounded sy tems may eanse repetitive discharges thrangh the arrester that exceed its thermal capacity. Tt is not considered feasible to apply arresters rated. sufficiently high to withstand the overvoltages that might be produced by any of the above. However, the possi- Dility of damaue to the wsrester front these causes should bbe considered in making the application. It is sometimes possihla ta make minor medifieations to the system equip. ent or operation that will greatly alleviate these sources of trouble. Where there is doubt as to the arrester rating, the maximum line-to-ground voltages should be ealeulated by ke methods uf Chap. 1, 9, Coordination of Protective Devices with Appa- ratus Insulation ‘The margin that should exist between the BIL of the insulation to be protected and the maximum voltage that ‘can appear across a lightning arrester is a much-diseussed ‘question, The answer is difficult because it depends on many factors. The breakdown vollage of the arrester is alfeeted by the rate of voltage riso and tho discharge vollage by the rate of rise of the surge current and the magnitude 08 of the surge eurrent. ‘The distance between the arrester location and the protected insulation affects the voltage imposed on insulation due to reflections. The severity of the surge depends upon how well the station is shielded, the insulation level of the station structure, and the in- coming line insulation. A typical problem is reviewed later to give one way of applying a suitable margin. Direct strokes to an arrestor should be eliminated, where possible, by proper shielding because the current in a direct stroke may be in excess of that for which the arrester is Uesigued. Wheie shielding is impractical, the arrester should protect tho insulation within the range of direct~ stroke surge currents within the capability of the arrester. Currents in excess of the arrester rating may damage or ruin the arrester. For a traveling wave coming into a dead-end station, the discharge current in the arrester is determined by the maximum voltage that the line insulation ean pass, hy the surge impedance of the line, and the voltage character- istic of the arrester, according to the following relation: 28 I where [,—arrester current, E —magnitude of incoming surge voltage E,—arrester terminal voltage Z —surge impedance of the line Z = TT 10 UNITS: oI MicROSECONDS 17 16-1820 Fig. 30—Coordination of insulation in a 138-kv substation for 134 40 microsecond positive wave. (1) Transformer with 850 kv BIL, (2) Line insulation of 9 suspension units. (3) Disconnect switches on 4 apparatus insulators (4) Bus insulation of 10 suspension units () Manimusn 114 x-40 wave permitted by ling inzulation. (6) Discharge of 121-kv arrester for maximum 114 x 40 full Tnsulation Coordination Chanter 1g = Suppose it is desired to protect the 138-kv substation shown schematically in Fig. 30 against traveling waves The system at this point is grounded so as to allow a basig impulse insulation level of 550 kv. ‘The major equipment consists of a power transformer, circuit breakers, disco neet switches mounted on four apparatns insilator uni and bus insulation consisting of 12 suspension insulators, The line insulation consists of nine suspension insulators, The arrester is located elose to the Lranstormer. Adequatp shielding is provided over the substation and the incoming tranamiceion lines, “ ‘The line insulation of nine insulators permits a travel wave of 860-kv crest (14X40) and rate of rise of 1000 k¥ per microsecond to enter the station. This rate of rise represents 8.25 kv per microsecond per kv of arrester rating for due required 121-ky arrester. Prom Fig. 20 the average arrester-gap breakdown is 3.6X121 or 435 ky at 0.5 microsecond, which, with a 15 percent. plus tolerance, becomes 500 kv, Assuming a line surge impedance of 409 ohms, the magnitude of the arrester current is about 3200 amperes determined as follows: 2 (860 000) —435 000_ . 1 a '=3900 amperes The rate of rise of current would be approximately 2 (1.000.000) 5000 amps/microsecond. 400 From Fig. 27(b) the discharge voltage for a current of 3200 amperes and a rate of rise of 5000 amperes is 3.45X 121 or 418 kv. Adding the manufacturing tolerance of plus 10 percent gives 460 kv as the discharge voltage provided by the 121-kv arrestor for the assumed cuudi- tions. Since the rate of rise has been taken into consid eration in establishing this protective level of 460 ky, no additional margin need be added. There is, however, @ difference of 550 minus 460 or 90 kv between the pro- tective level and the BIL of 550 kv of the transformer insulation. Suppose a direct stroke at the atation diocharges through the arrester a current of 50 000 amperes, rising to crest in three microseconds, with a nominal rate of rise of 20000 amperes per microsecond. The discharge voltage from Fig. 27 (b) is 4.55X121=550 kv for a 121-kv arrester which will plus 10 percent is GOS kv or 35 kv in oxcoss of the insulation BIL. 10. Location and Connection ot Protective Devices ‘The protective device should be placed as close as pos sible to the apparatus it is to protect, particularly J&sa& overhead line dead ends-in a station or terminates, 8 troneformer. A traveling wave coming into the station: limited in magnitude at the arrester location to the di charge voltage of the arrester. However, a wave with the same rate of voltage rise as the original wave and with ‘a magnitude equal to the arrester discharge voltage travels un lv the station terminus where it reflects to tion its value if the line dead ends or to almost twiee its value if the line terminates in a transformer. ‘The voltage the transformer builds up at a rate twice that of the original wave until it reaches a maximum value of twi#?, q a Chapter 18 the magnitude of the arrester voltage or to whatever volt~ age magnitude can build up during the time the reflected wave travels back to the lightning arrester and a negative reflected wave travels from the lightning arrester buch Ww the transformer. Likewise, apparatus, such asa disconnect s\iteh, lo- cated ahead of the arrester is subject. to the incoming surge until the arrester discharges and its negative re ected wave returns to the switch. ‘To illustrate the effect of arrester location consider the 138-kv station shown schematically in Fig. 31 with the arrester located 100 cireuit feet beyond the disconnect, switeh and 100 circuit feet ahead of the transformer. Con- sider a traveling wave having a rate of voltage rise of 1000 kv per microsecond entering the station and an ar- rester which limits the voltage to 400 kv. Tn 0.1 microm cond after the wave reaches the switch it reaches the lightning arrester and 0.1 microsecond later, or at the end of 0.2 microsecond, it reaches the transformer where it re- fleets and builds up at a rate of 2000 kv per microsecond. ‘At the end of 0.4 microsecond after the wave first reached the switch, the incoming wave and the reflected wave from the transformer would total to 400 ky at the arrester. As shown in Fig, 31, the voltages at the switch and at the transformer would also be 400 kv. The reflected wave trom the transtormer has just reached the switch. The voltage at the arrester remains at 400 kv until the erest of the incoming wave is reached but the voltages at the switeh and transformer continue to rise at 2000 kv per microsecond until the reflected negative waves from the arrester reaches the switch and transformer at the end of 0.5 microsecond. Successive refiections occur until the wave spends itself by discharging through the arrester. As shown in Fig, 31, the voltages at the swviteh and trans- former resulting from the first reflection reaches 600 kv, or 50 percent more than the arrester discharge voltage. ‘The maximum voltage at the terminus of a ling or at a transformer at the end of a line beyond an aféster as a = dips dey / , 050Ke\, Pee us \_. torr too er DISCONNECT SwiTcH TRANSFORMER 600 + ; 5 400} — 3 wool | 2 ial i] | 3 £1 OBE Os os OCR GS Ue O*UE OF OB micRosecoNos o oe) & Fig. 31—Voltages at 138 ky substation resulting from first re- fection of traveling surge having 1000 kv por microsccond, ‘wave front. (a) At disconnect switeh Ioented 100 fast ahond af arrostor (b) At arrester. (c) At transformer located 100 feet beyond arrester. Insulation Coordination fc fe L ARRESTER LTA 500 s 8 3 8 2 5 8 g 3 3 8 ° ARRESTER VOLTAGE VOLTAGE AT P™OR T*| ° Too 200 500 LIN CIRCUIT FEET Fig. 32—Maximum voltage due to fret reflection of travel wave as function of distance from arrester and steepn: ‘wave front. 400 800 *Voltage at may reach crest of incoming surge as maximum, Voltage at T may reach twice arrester voltage as maximum. result of the first reflection of a traveling wave, may be expressed mathematically, as follows'*: de. L. eo} LEX To05 Ey up to a maximum of 2 ¢., where esurrester discharge voltage rate of rise of wave front in kv per ms, ‘L=distance between arrester and line terminus in feet. ‘The same expression ean also be used to determine the voltage at a point on a line ahead of an arrester due to a traveling wave. In this case, the voltage can reach as a maximum the crest of the traveling wave if the distance to the arrester is yreat enough or if the rate of rise of the ‘wave front is sufficiently high The curves of Fig. 22 show the voltage in excess of the arrester voltage as a function of distance from the arrester for rates of rise of wave front of 100, 500, and 1000 kv per 630 microsecond. ‘The curves can be used to determine the actual voltage at a point ahead of an arrester ur at w line terminus beyond an arrester by adding to the curve value the discharge voltage of the particular arrester involved. For example, the maximum voltages obtained at the switch and transformer in Fig. 31, by plotting the volt- time curves, could be taken from the 100 kv per micro- second curve of Fig. 32. For a distance of 100 fect, and a wave front that rises at the rate of 1000 kw por micro second, the voltage in excess of the arrester voltage is 200 ky, This added to the arrester discharge voltage, assumed to be 400 kv, gives 600 kv as the maximum voltage after the first reflection, In addition to the reflected wave phenome, itis quite possible that still higher peak voltages would exist at the Apparatus as a result of oscillations eaused by the induet= 3 Fig, 32—Installation view af CSP power transformer with protective devices, secondary circuit breaker, and metering ‘equipment built integral wich transformer. ance of the line between the arrester and the apparatus, and the capacitance of the apparatus, Furthermore, the voltage drops in the lead from the line to the arrester and in the lead from the arrester to ground, which are affected by rate of rise of surge current, add to the drop across the arrester. Any difference in ground potential between the apparatus ground and the arrester ground also adds to the voltage impressed across the apparatus insulation. In view of the above factors, it is important, particularly in stations where direct strokes may originate close to the station, that the protective devices be located close to the apparatus they are-to protect, that the leads to the devices be kept as short and direct as possible, and that the ar rester and apparatus grounds be intereonneeted and as low in resistance as possible, preferably one ohm ot less. ‘The ultimate in this respect is reached when the protec~ Insulation Coordination Chapter 18 tive device is mounted directly on the transformer. ‘This is illustrated in the installation view, Fig. 93, of » CSP transformer which has the protective devices, secondary cireuit, breaker, and metering equipment built integral with the transformer. With the line side of the arresters connected directly to the transformer terminals and the ester ground connected direetly to the transformer tank, the voltage between the winding and core is definitel ited to the discharge voltage of the arrester. To provide protection to an extended station an arrester should be located directly ahead of the disconnect switch where the Line enters the station and another arrester located directly adjacent to or on the transformer. A modification of this scheme, which is sometimes used, is to locate protector ‘tubes at the entrance to the station and conventional sta- tion type arresters at the transformer terminals. The pro- tector tubes will generally protect the switch and will limit the magnitude of surges entering the station, 11, Direct Stroke Protection Wherever it is possible for direet strakes of lightning to strike the line at or near the station, there is a possibility of exceedingly high rates of surge-voltage rise and large magnitude of surge-current discharge. If the stroke is, severe enough, the margin of protection provided by the protective The installation may, therefore, justify shielding the station and the incom- ing lines far enough out to limit the severity of surges that can come into the station, particularly in the higher volt age classifications, 69 kv and above. This ean be done by properly placed masts or overhead ground wires, The number of direct strokes per year to an unshielded substation, based on aceumulated records af direct strokes to tall objects, ean be approximated by the expression» (W+700)(L+700), 5 (5280)? ° device may be inadequate. ~APPROK:I/2 MILE. GROUND suievol — wine. o pt ro. 2 ae $s} I = Tees ~} ce Fig, 14 Typical schemes of station protection. (a) Arrester at station with no direct stroke shielding. (b) Arrester at station with shielding against direct strokes () Arrester at station with protector tubes extending out 14 mile, -~ Chapter 18 where IF and Z, are the width and length, respectively, in feot, of the substation. From this estimato of etrokee to an exposed substation, which is about one stroke every four and one-half years to a 100 feet square substation, it ‘can be reasoned that reducing the exposure to 0.1 pereent would practically eliminate the possibility of a stroke to a station. The curves of Fig. 35! were, therefore, cun- structed from extensive laboratory test data to show the configurations of masts or overhead ground wires necessary to reduce the exposure of an object to 0.1 percent. ‘The curves are plotted to show the height (L) of the shielding G20 69 Bo "tig wo 40 goto so rear ae Sow yoo A —- . aE a - | ° 3 ah z | Xe pop yt te7-——S x ‘ool i ae cy) ig. 35—Configuration of shielding object with respect to provected object for 0.1 percent exposure. (8) One shiclding mast. Dotted lines for one exposed object of height (d). Pull lines for ring of exposed objects of height (a, (©) One horizontal ground wire, (6) Two masts or twu grouiad mites, Dylled lines for masts. Full lines for horizontal wires. Insulation Coordination x, Mies Fig. 36—Areas protected by multiple masts for point expo- ‘sures of 0.1 percent, (a) Two masts with values of 2 and s taken from Fig. 35 (a) and (c) (®) Two masts separated by half the distance of those in (a). (©) ‘Three masts with (b) points obtained from Fig. 85 (c) for mid-point between two masts (@) Four masts with (b) poits obtained from Fig, 35 (e) for sid-point between twe maate, ‘masts or ground wires above the protected object as a fune- tion of the horizontal separation (X) and the height (d) of the protected object. ‘The dotted-line curves, Fig. 35 (a), showing the neces- sary configuration of a single mast protecting a single ob- ject, apply to an exposed structure having a single prom- inent projection or several projections in a limited region, such as a set of disconnects. The full-line curves, applying toa ring of objects, should be used if the live parts to be shielded are generally distributed at a given height. The configuration of the mast should be based on the most remote object. ‘The required configurations of a single horizontal ground, wire are given in Fig. 35 (b). ‘The dotted-line and the full-line curves of Fig. 35 (c) apply, respectively, to two masts and to two horizontal ground wires. The diagrams of Fig. 36% illustrate the area that can be protected by two or more shielding masts, The cross hatched area of Fig. 36 (a) is the area protected by two masts for given values of d and 1, where the radii x of the semi-circles are taken directly from Fig. 35 (a), and the separation distance S from Fig. 35 (c). If the distance between masts is decreased, the protected area is at leust equal to the area obtained by superposing the areas of Fig. 36 (a). For example, if the distances between masts is halved, the resultant protected area is somewhat as shown in Fig. 36 (b) ‘On this basis, to form an approximate idea of the width of the overlap between masts, first obtain a value of y from Fig, 35 (0), corresponding to twice the actual distance be- tween the masts. ‘The width of overlap is then equal to the value of x, obtained from Fig. 35 (a), that correspond to this y. This undoubtedly gives a conservative width of substation that ean be protected by two masts. For three imusts loeaied at the points of an equilateral triangle or for four masts located at the points of a square, 632, the protected areas are as shown in Figs. 36 (c) and (4). ‘The height of the shielding mast should be so chosen that. the b points provide 0.1 percent exposure as obtained from Fig. 85 (c) for the midpoint Lebween Gnu musts. ‘The a radii are obtained from the data for a single mast. ‘The eurves of Fig. 35 apply to stations located in regions of relatively flat terrain and low resistivity, where the effective ground plane is essentially at the earth’s surface. High values of earth resistivity lowers the ground plane, which results in less effective shielding for a given config- uration. However, most atationo are, oF if not, chould bo, provided with low-resistance grounding systems for light= ning-arrester grounds, which can also serve as shielding grounds, Where the soil resistivity is high the effective ground plane can be raised to the earth's, surface by laying counterpoise wires from the shielding masts to distances uf two or three times their height. However, in most cases, it is probably more economical to inerease the height of the masts. For application of the curves to hillside locations, the dimensions () (the shielding mast height) and (d) (the height of the protected object) should be measured per- pendicular to the earth's surface. The distance (2) be- tween the object and shielding mast should be measured along the earth's surface. ‘The lines coming into the station can be effectively shielded against direct strokes by overhead ground wires as outlined in Chap. 17. A direct stroke on a line more than } mile out from the station is limited in severity at tho station by the surge impedance and insulation of tho line and to some extent by the shunt eapacity of the station equipment. Shielding of the station and the lines approx- imately } mile out from the station, as illustrated in Fig. 34 (b), is, therefore, a desirable supplement to the lightning, arrester logated ub the station. ‘Where overhead ground wires on the incoming lines are not practical due to existing construction, additional pro- tection of the station equipment against direct strokes on the Tines near the station can be obtained by equipping each line with protector tubes at the entrance structure of the station and at each tower for a distance of approxi- mately } mile out from tho otation, coe Fig. 31 (0). How ever, shielding the station is the only way to eliminate direct strokes to the station itself. 12, Summary of Considerations Applying to Pro- tection of High-Voltage Equipment The following points can be generally concluded in the application af protective devieos to high-voltage systems, 22 kv and above. 1. Rod-gaps do not protect apparatus insulation against surges of steep wave front unless the spacing is s0 low that the gap is subject to numerous flashovers from minor aurges. 2. Protector tubes are not considered suitable for the protection of apparatus insulation although they are effective in preventing transmission-line flashovers and in decreasing the severity of surges from direct strokes near the station. The application of protector tubes involves certain limitations in system short- cireuit and recovery voltage characteristics. Insulation Coordination Chapter 18 3. Modern type lightning arresters applied properly protect station apparatus conforming to basie insu. lation levels against traveling surges. On systems having a solidly yiounded ueulial celuced rating are resters can be applied. Full-rated arresters are te. quired generally on ungronnded neutral systems or systems grounded through impedance. ‘The possibil- ities of system overvoltages should be investigated carefully in determining the minimum rating arrester ‘that can be applied economically. 4. For affootively grounded aystoma inaulation levels one class below the standard have given satisfactory setvice using reduced rated arresters particularly at voltages 115 kv and above, Consideration should be given to shielding stations against direct lightning strokes. Where shielding is not practical, additional protection can be provided hy installing protector tubes at the entrance to the station and at each transmission-line tower for a dise tance of } mile from the station. VI. PROTECTION OF DISTRIBUTION TRANSFORMERS ‘The distribution transformer with its protective devices is in effect a miniature substation constituting the final voltage transformation between the generating station and ‘the individual eustomer’s premises, Because the distribu- tion transformer is small in size and comparative cost, and bbooause it is usually polo mounted, often in out-of-the-way ocations, its protective devices must be inexpensive, small in size and weight, simple, and reliable. ‘The failures of early type distribution arresters and the large amount of lightning data obtained on distribution circuits furnish proof that Uke protective devives must alsy bueve dae al to withstand severe lightning discharges. 13. General Considerations Distribution circuits are generally overhead construction and are, therefore, subject to lightning disturbances, the nature of which are discussed in detail in Chap. 16. Data collected over @ period of years with aurge-ereat, dovicos indicate that the majority of surge-current discharges on distribution circuits are relatively low in magnitude, less than 5000 amperes, but occasionally a discharge may ex ceed 100 000 amperes. More recent data collected with the Tulchronograph show that some of the surge-eurrent dis- charges that are moderate in magnitude may be long in duration, af the order of several thousand microseconds. See Chap. 16. Experience has shown that lightning disturbances are ‘more severe on rural circuits than on urban eireuits, prob- ably for two reasons. First, rural circuits are generally more exposed to lightning and, therefore, receive many more direct strokes. Second, because distribution trans- formers are less frequent on rural circuits—the drainage long-duration surges through grounded transformer wind- ings is less! “These general conclusions have been borne out by oper ating experience with distribution lightning arresters. The failure rato of early arrestors, attributed to lightning, that Chapter 18 were designed before high surge-current testing facilities svete available was comparatively high. Luter arresters, designed and tosted to withstand high ourge currents of about 100 microseconds duration have a good operating record on urban eireuits. However, when these arresters were applied more extensively to rural circuits with sparse- ly located transformers, the failure rate increased. Modern arresters, designed with more emphasis on ability to dis- charge surges of long duration, have acquitted themselves ivell an raval eirenits ‘Most distribution transformers are pole mounted, one at a location, and are used to step the voltage down from a single-phase primary circuit (2300 to 13 200 volts) to a single-phase, three-wiro, secondary circuit at utilization voltage, usually 120/240 volts. Occasionally they are used in three-phase banks to supply three-phase, low-voltage power The primary circuits, whothor cingle or three- phase, may be from a source having either a grounded or ungrounded neutral, If the neutral is ungrounded the single-phase primary consists of two of the ungrounded phase conductors, which means that the tivo primary ter- Iniuals uf Uw distribution transformer must be equally protected. If the nentral is grounded, the single-phase pri- mary cirenit will usually eonsist of one phaco wire and the neutral conductor. The neutral conduetor is usually grounded, so that only one high voltage terminal of the distribution transformer need be provided with protective equipment. The secondary will usually be three-wvire, with the snid-puint grounded. A three-phase bank of distribution transformers may be connected delta delta, star delta, or delta star, With the delta delta connection, the sceondary can have no common neutral. Sometimes, either one phase or the mid-point of ‘one of the phases of the secondary is grounded. The star delta and delta delta connection are alike as far as the secondary della eounection Is concerned. The primary neutral might ot might not be grounded. ‘The delta star connection nisually has the common noutral of the otar connected secondary grounded. in addition to surge protection, the distribution trans- former usually includes protection against internal short- ‘ireuit and secondary short-cireuit or overloads consisting cither of high-vollage [uses mounted external to the trans- former, or high-voltage fuse links and a secondary circuit, breaker mounted inside of and included az a part of the transformer ‘The distribution transformer, like larger power trans- formers contains three groups of insulation subject to volt- age stress, which should be considered in the protective teheme, namely’ 1. The insulation between the high-voltage winding and the core or tank 2, The insulation between the low-voltage winding and the core or tank, 8. ‘The msulation between the high- and low-voltage windings, ‘There are, however, two conditions that make the pro- {ection of distribution transformersand high-voltage power ttansformers differ ‘These are the difference in tho ratios Of surge strength to operating voltage and the relative ‘fleets of locating and connecting the protective devices. Insulation Coordination 633, ‘The distribution transformer (2400 to 13 200 volts) has a much higher ratio of surge strength to operating voltage, as Table I shows. As ut esunple, the ratio of the baste insulation level to the peak of the 60-cycle voltage elass- FAt73= 1275, for 2500 volt equipment, as 650 For th 1B8x-V . reason, it is permissible for the protective device in the low-voltage ratings to have a higher protective ratio than that required at higher voltages. ‘The effect of the location aud connection of the protec tive devices is more pronounced with distribution trans- Formers. Recanse lightning discharges on distribution oir- cuits and on high-voltage transmission circuits are about equal in magnitude, the actual surgo-voltage drops in the leads to the protective devices and through the ground connections of the two circuits are about equal. While these voltage diops may be ouly a portion of the discharge voltage of the protective device in the higher voltages, they may be several times the discharge voltage of the low- voltage protective device. Tt is extremely important, therefore, that protective devices on distribution circuits be located and connected properly with respect to the apparatus they are to protect. 14, Methods of Connecting Protective Devices Three schemes of connecting protective devices to pro- tect distribution transformers against lighting surges are commonly known: fication is, against 3.33, for 138-kv equipment. 1, Separate connection method. 2. Interconnection method. 3. Three-point protection method. Separate Connection Method—This method of prow tection, universally used until about 1932, is illustrated in Fig. 37. Protective devices are connected between the pri- Hw [uw | = Fig. 37—Separate connection method of protect phase transformers, 8 single- ‘mary conductors near the transformer and a driven ground at the pole. The secondary neutral is usnally grounded separately. With this connection, the protective devices are connected in serios with a relatively long ground lead whieh has considerable inductance and is usually connect ed to a driven ground, the resistance of which may be high. Tho voltage, therefore, between the primary winding aud ground is not only the discharge voltage of the arrester but, also the impedance drop of the ground lead and ground oat Insulation Coordination Chapter 1g HE 1e4 rz 4 Hh Ed 7 Sp at et iw (b) () Fig. 38—Interconnection method of protecting single-phase transformers. (a) Straight intoreonnection. () Interconnection with protective ground at tranaformer. connection, which may be several times the discharge valt- age of the arrester. Failure or flashover of the transformer insulation may occur even though the actual surge voltage aeross the arrester is only a fraction of the transformer breakdown voltage. When this happens, the surge gener- ally passes through to the secondary ground connection. ‘The surge is usually followed by a flow of dynamic current, until the primary fuse blows, Interconnection Method—The straight intereonnee- tion consists of connecting the protective deviees from the primary lines directly to the secondary neutral as illus- trated in Fig, 38 (a). ‘The surge voltage that can exist botween the primary winding and the secondary is dof nitely limited to the discharge voltage of the protective deviees. ‘The potential of the core and tank, because of their electrostatic coupling to the secondary winding, nor- mally rises along with the primary and secondary windings during a surge discharge aud thus limits Ue voltage be= tween the windings and core. This connection is an im- Drovement. over the conventional connection because it, eliminates the factor of voltage drop in the arrester ground lead. Operating experience supports this conclusion”. wo decided disarlvantages prevent universal applica- tion of this scheme of connection. First, the protection between windings and core or tank depends upon the tank rising in potential with the windings. Actually, practical conditions might keep this from happening. The resistance to ground of the wood pole on which the transformer is mounted may be low enough to supply the small charging current required to keep the tank at ground potential Also, as it rises in potential, the tank may flashover to a nenrhy guy wise or other graumded object. Hither eondic tion results in the application of the full potential of the surge between the windings and tank until the insulation breaks down or the secondary flashes over to the tank, The second disadvantage of this connection is that it directs the entire primary surge voltage into the eccond- aries, which is undesirable, particularly if the resistance of the secondary grounds is not low. ‘This restriction makes the straight interconnection in general unapplicable to rural circuits or other circuits that might not have the secondaries effectively grounded. A modification of the straight interconnection is shown in Fig. 38 (b). Hore, ground is made at the arrester lo- cation also. ‘The protection provided the transformer in- sulation depends upon the tank being insulated from (6) Intereonneetion with protective ground at transformer and {nwulating gop in intoreonncetion, gronndl, or, if not, upon the magnitude of voltage drop in the arrester ground lead and connection. ‘The arrester round is in parallel with the secondary ground so that the complete surge is not directed to the secondary. ‘The direet tie between the arrester ground and secondary ig undesirable unless the secondary is effectively grounded, again making the connection generally unapplicable to rural circuits Another modification that eliminates the permanent tie between the arrester ground and secondary neutral ig shown in Fig. 38 (c). An isolating gap, having low flashe over, breaks the direct tie. A rise in potential between windings during a surge discharge breake down the gap and limits the voltage between windings to the arrester discharge voltage. ‘The protection of the transformer insulation to the core or tank still depends upon the tank being insulated from ground, or, if not, upon the voltago dup ia the advester grouud lewd aul! connection being low. Three-Point Protection Method—This scheme, ile lustrated in Fig, 39 (a), definitely limits the voltage across the threo groups of insulation in the transformer inde pendently ot ground connections or resistances. ‘The pro- tective devices connected betiveen the high-voltage lines and tank definitely limit the veltage between thove part to the discharge voltage of the protective device. Like- wise, the protective device between the secondary and tank (usually a gap for 480 volts and below) limits the voltage between those parts to the breakdown voltage of the device. With the voltage between the high-voltage winding and core or tank and the voltage between the low- voltage winding and core or tank definitely limited, the voltage between the tivo windings is also limited. Referring to Fig. 39 (a), a surge coming in over & pr mary lead raises the potential of the primary winding to the breakdown of the protective device that discharge! to ground. If the arresterground impedance is high oF there is no ground at that point, the potential of the bigh= voltage winding rises above that of the core and tank until the gap T@ breaks down and limits the voltage betwee the winding and tank to the discharge voltage of the arrester plus the gap. If the voltage betiveen the tank #9 secondary exceeds the breakdown of the gap T'N, the g8P operates and discharges to tho secondary ground. The gaps, 7G and TN, while they definitely isolate the tae from the primary and secondary ground connection Chapter 18 4 uw. | TTS Mi a toes PROTECTIVE GAP (b) Insulation Coordination 685 protection to the three groups of insulation in the trans- former independently of the arrester or secondary ground ing conditions, whether tho tank ie insulated or not, oF whether the surge originates on the primary or secondary. Variations in the methods of connecting the protective devices to obtain three-point protection are shown in (b), (¢), (@), and (e) of Fig. 3 1. Protection of Three-Phase Transformer Banks ‘The shortcomings of the separate connoction mathod of protection apply equally well to the protection of three- phase transformers or three-phase transformer banks on. distribution eireuits. ‘The interconnection method is gen- erally not applicable because there is no secondary neutral unless the secondary is connected in star. Suuetimies vis phase of the secondary or the midpoint of one of the phases is grounded as shown by the broken lines of Fig. 40 ‘The three-point scheme of protection, illustrated in Fig. 40, is applicable to any winding connection. A pro- (d) te) Fig. 39—Three-point protection method (a) Three-point protection with insulating gaps. {b) Three-point peateotian simplifnd sires (6) Three-point protection with single insulating gap. (@) Three-point protection of single-phase transformer on four- vie, grvauded-neutral circu (6) Three-point proteetion with insulated tank. ing normal operation, do not greatly add to the surge Voltage impressed across the winding insulation. The three-point scheme of protection thus provides definite Fig. 40—Three-point protection applied to three-phase dis- tribution transformers, tective device is connected between each primary phase winding and tank either directly or through an isolating gap. Likewise, a protective device (air gap or coordinated secondary bushing for 480 volis aud Leluw) is connected between each secondary phase lend and the tank. The tanks of all transformers in the bank are tied together With this connection, the windings of all transformers are protected irrespective of grounding conditions or whether the surge originates on the primary or secondary circuit. 16. Surge Voltages in Secondary Cireuits Overhead secondary circuits are, to some extent, sub- ject to lightning surges originating on the secondary, ‘They ean also experience surges passing from the primary into the secondary. ‘The separate connection, Fig. 37, isolates Ue priuuary frum the secondary, Howvever, when the transformer fails or lashes over a3 a result of a pri- mary surge, the surge pasces directly on to. phase wire ‘or neutral of the secondary circuit. ‘The primary-system is also Impressed on the secondary until Uke p ¥y fuse blows. ‘The straight interconnection directs all the primary snnge on to the secondary neutral. With the modified interconnection shown in Fig. 38 (b) or with the three-point scheme of protection, some of the surge may pass on to the secondary neutral, depending upon how effectively the primary protective deviees are grounded. Experience has shown that damage eaused by eurges on secondary house cireuits is negligible. Where long expo- sures and relatively high secondary insulation may result, in damage, protection should be provided by low-voltage protective devices located at the house entrance, and con- ected between the phase wires and neutral," which is usually grounded. All grounds on the customer's premises should bo connected together. In ease the secondary ig not grounded in the eustomer’s premises, such as may be the case with a three-phase, four-wire delta circuit, the danger of damage is greater than with the usual house cirouit. Where the hazard is considered serious, it can be tliminated by connecting a protector between each phase wire and ground at the house entrance, or preferably right at the apparatus to be protected, ‘Three-phase, 440-volt circuits sometimes extend a con- siderable distance overhead to motor cireuits, thus con- sututing a hazard to the motors and essociated starting equipment. ‘They ean be protected by connecting a low- voltage protective davieo botwoon each phase wire and the frame of the apparatus, which should be grounded. Protectors connected to the secondary at the transformer will generally not provide adequate protection to the load apparatus. 17, Protective Devices for Distribution Trans- formers Three general classes of devices are used for the protec- tion of distribution transformers just as for the protection of high-voltage substations, namely, the plain air gap, the protector tube, and the conventional valve-type lightning arrester. However, the lower operating voltages, the higher ratio between insulation breakdown voltage and operating voltage, and the requirement that the device be small in size and cost, make the design and application of protective devices somewhat different. for distribution transformers than for higher voltage equipment. Plain Air Gap—Plain air gaps or fused gaps are some- times used to protect. distribution transformers. The relatively high insulation strength of the transformer makes it possible to provide a fair degree of protection to the transformer against lightning surges without having to decrease the gap spacing to a value where numerous flashovers occur as a result of minor surges. However, the device will not restore power voltage (above 480 volts) after a discharge without momentarily deenergizing the cireuit, which usually results in the blowing of a fuse, either at the transformer, or at a sectionalizing point on the line. ‘The gap spacings associated with low operating voltages are necessarily low so that unless the gap is enclosed or protected, numerous fashovers can occur as a result of birds or foreign objects bridging the gep. Double gaps of various constructions are sometimes used ‘to minimize this trouble. Since distribution transformers Insulation Coordination Chapter 1g are uflen located in remote locations, it is Important to avoid as many fuse replacements as possible. For that reason and berause of the somewhat. questionable proe tection obtained for surges of stexp wav (rot, pain aie aps or fused gaps are not extensively used to pr Gistrbation trans ormers . pees. Protector Tubes—The distribution-type protector tubo, introduced about 1031, consists cosentially of a sitll air gap, a diffuser tube, and sometimes a resistor, all con. nocted in series, ‘The series gap is just enough to insulate the tube from normal power voltage, thus eliminating a continuous voltage stress across the diffuser tube. ‘The purpose of the series resistor when used is 10 limit to approximately 500 amperes the one-half cycle of power rool 4 NEGATIVE Tvourase Arne 7 "3800 z 83 3 SE is 2.1 \ cI { 208 Per 09 _| rT SES eee OTe STS er Fig, 41—Volt-time breakdown characteristics of one type of distribution type protector tubes. current that may follow the surge discharge, thus making the application of the tube independent of the system short-circuit eurrent ‘The gap breakdown characteristics of the different volt age ratings of a typical type of protector tube are shown in Fig, 41. After the gap breaks down, the discharge voltage is equal to the are drop in the tube plus the drop across the series resistor, if one is used. ‘The series resistor is generally provided with a shunt gop that limits the voltage across the resistor to about 30 kv. If the lightning snnge is of sufficient. enrrent magnitude to build up & resistance drop of 80 kv scross the resistor the shunt g=P fiashes over and takes the resistor out of the discharge cit- cuit in which ease the discharge voltage is the are drop through the tube. Surge currents high enough to cause the shunt gap to flash over, pruduce sullicient deionizing action in the diffuser tube to cut off after the discharge without the one-half eycle of power-follow current, Although the gap breakdown voltage of the protector tube is higher than that of a corresponding valve-tyP® lightning arrester, partieularly at short time lags, the whe adequately protects modern distribution transformers rated 13800 volts and below if connected properly. Laboratory tests and operating experience have demot strated the ability of a tube to discharge severe strokes lightning. This characteristic together with its ability ‘ bes Chapter 18 2 ; eo ~ US KV) re lan Boo Lory zt * — Bao | [|_| 3 i | ex = 30) : + sot ee | | i dl | 3F0a-08 3511 TB winosecouos To aREAxoOwn Fig, 42—Gap breakdown characteristics of typical distribu- tion type arresters for 114 x 40 microsecond voltage wave withstand high momentary system voltage makes it especially well suited for anplieation on rural circuits Conventional Valve-Type Lightning Arresters— ‘The valve-type lightning arrester is the device most gen- erally used for the protection of conventional distribution, transformers, that is, transformers requiring separately mounted protective devices. The curves of Pig. 42 and Fig. 43, show respectively the gap breakdown charac- teristies and the discharge chararteristies af typieal mod- em distribution-type arresters. Operating experience of several years has demonstrated the ability of conventional valve-type arresters to provide a high degree of protection to distribution transformers. Modern construction has eliminated Uhe sacclautical diff culties experienced with early designs, which resulted in a relatively high failure rate and oceasional_radio-inter- ference complaints. Field measurements of surge-crest, magnitudes together with laboratory tests led to later designs having the ability to discharge surge eurrents of 109, rs isxvl 35 niLovo-rs~crest "000 2000. $000 3000 AMPERES 16000 “20000 Fig, 43—Discharge voltage characteristics of typical distribu- tion type arreaters for 10 x 20 microsecond current wave, Insulation Coordination 637 high erest magnitude. More recent data obtained with the fulchronograph have shown that the distribution-type arrester should also be eapable of discharging surge cur rents of long duration, Valve-type lightuiny arresters are now available that will handle either surges of high crest magnitude or long duration. Surge-Proof and CSP Transformers—Vhe surge- proof distribution transformer, containing, as a part of the transformer, devices for complete surge protection, \vas introduced in 1932. An expulsion tube arrester, known as the De-ion arrester was connected between each pri- mary terminal and tank. ‘These arresters and the coordi- nated low-voltage bushings of these transformers together provided the three-point method of surge protection which {or the first time gave the means for completely protecting all three major insulations. ‘These surge-proof transformers still required external fuse cutouts to disconnect the transformer from the line In case of secondary overload or short cirenit or internal failure. Blowing of theco fuses and sometimes failure of DEION Gaps PROTECTIVE LINK TEMPERATURE INDICATING LAMP. ‘SECONDARY BREAKER YooromareD BUSHING AP Fig, 44—Circuit diagram of CSP transformer. Te practice is to gronnd tanks, remove tank discharon exp and oan- rect tank directly to ground. Ue cutout coustituted a large share of the trouble expe~ rienced with distribution transformers caused by lightning surges. Fuses cannot always he depended on to give ade~ quate overload and short-circuit, protection, Also, the mounting of cutouts necessarily adds to the cost and com- plication of installation of the transformer. The completely self-protecting (CSP) distribution trans- former, introduced in 1993, overcame these difficultics Like its predecessor, it contained complete lightning pro- tection, provided by high-voltage De-ion arresters and low- voltage coordinated bushings arranged to give three-point protection, as shown in the circuit, diagram of Fig. 44, In addition, an internal cirewt breaker connected de- tween the low-voltage windings and low-voltage terminals protected the transformer against overload or sooondary 628, \ Taoulation Coordination Chapter 15 breakers instead of one. These are intercon nected within the transformer so as to section alize the low-voltage circuits in ease of faults or overloads, ‘The CSP transformer is completely assem= bled in the factory thus making it possible to surge test the combined transformer and pro- eeetiae tective equipment. Proof of coordination of the insulation of each CSP transformer is now siven by applying to each assembled unit a cecmarn, asa surge test equivalent to a direct steoke of cme wn esa it lightning srorcerve ume francs con, VII. SURGE PROTECTION FOR Fig, 45—Scctional view of CSP transformer for operation on Srounded-neutral circuit. short circuits. feeder Finally, protection to the high-voltage from internal transformer failures was given by internal protective links connected between the high- voltage winding and bushing. The internal breaker and protective links perform all of the functions of the fuse, so that with these transformers no external protective de- vices are required. ‘Those transformers have now almost entirely superseded the surge-proof design. The bimetallic tripping element of the breaker, which is actuated by both load current and oil temperature, is calibrated to follow closely the permissible thermal load- time characteristies of the transformer windings and pro- vides loading on the basis of copper temperature. A sectional view of a CSP transformer with two cover bushings and De-ion arresters is illustrated in Fig. 45. Of special interest is the emergency control now supplied on these transformers. This device takes care of the occa sional situation where the breaker cannot be kept closed alter having been tripped by overload, because the over- Jond persists or motor starting currents are high. If it is imperative that service be restored, even at the risk of some loss of transformer life, the breaker setting may be elevated by means of the external emergency control luaudle to permit additional overloads. ‘The necessity for the use of this device usually indicates that the load growth has exceeded the capacity of the transformer so that the unit should be replaced as soon as possible by a larger one. Recent developments include the extension of the CSP principle to include three-phase distribution transformers and both single- and three-phase completely self-proteet- ing transformers for banked accondary operation (CSPD's) ‘The latter contain all of the protective features of the CSP transformers, and, in addition, they are supplied with two ROTATING MACHINES ‘The insulation on the windings of rotating machines, such aa large or cmall motors, ae generators, and synchronous condensers, is held to a minimum because of limited space. Also, sinee the insulation is not immersed in oil, its surge strength is not much greater than’ the peak uf (ie 60-cy cle voltage breakdown. Special measures are, therefore, necessary to protect such equipment when it is connected to a sys- tem subject, to the huzards of lightning-surge voltages. Likewise, the method of grounding effects the overvoltages, during fault con- ditions and switching, which may be impressed on rotat- ing machines; theeo phenomena aro discussed in Chapters Mand 19, ‘The stress on the major insulation of any machine, that is, the insulation between the winding and frame, is de- termined mainly by the magnitude of the surge voltage to ground, whereas the stress om Uwe (urn insulation is more a function of the rate of rise of surge voltage as the surge penetrates the windin Protection of a ratat- ing machine, therefore, requires limiting the surge voltage magnitude at the machine terminals and sloping the wave front of the incoming surge. The effect of sloping the wave front is illustrated in Fig. 46. The curves of Fig. 46 (a) show tho relative volt- 4 2 TT] gL To eS ai} f - eel a Zz err Fig, 46—Distribution of surae voltage in generator winding (a) Without rotating machine protection (b) With rotating machine protection, Chapter 18 9 to ground of the Tine terminal and two intermediate points in a phase winding of a machine without, prote: tion for an incoming surge rising to crest in one-balf micré second. ‘The differences in voltages at the various points in the winding result in high stresses between turns. The curves of Fig. 46 (b) show how the stresses between turns are decreased by sloping the wave front so that the surge at the machine torminals reaches erest in twelve microseconds, Limiting the surge voltage to ground sufficiently to protect the major insulation usually requires a special lightning arrester, having a low protective ratio, con- nected between cach machine terminal and grounded —— Tung free i wsratteo ar macune® > Tcnuas of Gus wo MACHINE earaciron. STANDARO speci ARRESTER ARRESTER | 1500 Ft +} @ : Wie OF SET Sorat | ve macnn | ! came*** {yi < - | j doranoage d |] specu | I cara i LIRR Pep lamesres cron | | 1800 Fr] © . 47 —Line surge impedance and capacitor method of pro- tecting rorating machines. (a) Single machine connectéd to overhead line (b) Two or more machines connected to common bus. (6) Machine connected Uiwugl slant cable Wo overbead lines For simplicity, protective dleviees are shown on one phate only. Fach phase however must have the same protective apparatus in= stalled. ‘For citeuits below 2300 volts. s-For circuits 2800 volts and above, ***Cable lengths up to 1000 feet are considered short eables in th type of application, If the eable is over 2500 feet long ad the max chine is connected dirvetly to it, itis satisfactory to omit the eapuei- torat machine terminals. Tn this case the capacity effect of the eable is approximately equivalent to the capacitor. Tnsulusion Courdinasion way frame, Where more than one machine is connected to a common bus, one arrester connected between each phase of the bus and ground generally is arlequate if the machine frames are connected to a low-resistance ground common with the arrester ground. Sloping of the surge wave front is accomplished by let~ ting the surge, after passing through a series impedance, charge a hunt eapacitor connected to the machine ter- minals, 18, Line Surge Impedance and Capacitor Method In this scheme of protection, a special arrester is in- stalled at the machine terminal to limit the magnitude of the voltage impressed on its windings. The sloping of the surge is aceomplished by a eapacitor charged throsygh the surge impedance and reactance of the line. See Fig. 47. To limit the voltage that determines the changing rate of the eapacitor, a lightning arrester or protector tube is placed on each overhead line far enough ahead of the machine 90 that the arrester will discharge before the ed on it is modified by reflections from the capacitor. distance will depend upon the slope of the incoming surge. The farther out the arrester is located, the less will be the stress on the machine winding tor surges originating beyond the arrester, but the greater will be the possibility of a surge originating between the line arrester and the station. A distance of 1500 to 2000 fe is a good compromise betiveen the possibilities of a strok within this area and the effect of distance on the charging rate of the capacitor, ‘The rate of rise of the surge reaching the machine is alvo a function of the amount of capacitance used. The maximum permissible rate of rise depends upon the v locity of propagation of the surge in the machine winding, the number of turns per coil, the turn length and the tum, insulation. A study of many cases has indicated that the ‘maximum rate of rise should be limited to a value such that, if the terminal voltage continues to rise, it will not, equal the test voltage of the machine in less than 10 micro- seconds. Considering a minimum practical line surge impedance and a practical machine surge impedanee, requires at least { microfarad of eapacitance. It is inde- pendent of rated circuit voltage because the machine test voltage and the voltage limited by the line arrester are proportional to rated circuit voltage. However, in the construction of capacitors there is a limit to the minimum capacitance that can be obtained economically on a, standard unit. For example, the standard 6900-volt unit coulains } microfarad, whereas the 13 800-volt unit con- tains only mierofarad. Tn an ungronnded machine, heease of the possihility of reflections from the neutral point, the voltage may double at the neutral. To limit the voltage at the neutral as recommended, it is necessary to hold the rate of rise ot the surge entering the machines to } the recommended value, by using at least }, inotoad of } mierofarad. In Table 0 are given the recommended capacitances for various volt- age classes from 650 to 13800 volts. For 11500 and 13 800-volt classes, two standard }-microfarad units are recommended for ungrounded machines, whereas + micro- farad is sufficient for a grounded usuchine, Below 11 300 640 ‘TanLe 9—RECOMMENDED CAPACITANCE VALUES FOR LINE Sunt: Isirevancie Miurnov uF PROTECHING RUFAtING MACHINES cm | Numer af | Meafaende Great | suaniard Gate] “in ach How | comet | per Phe |stnia tt oo | my 1 2 we | i os co | ay i 08 no | oramaa | 2 ismo | ‘Geonden | i tsa | Unguniea |? *Ugcinesetral considered grounded fgrounded through 3 reistr of $0 ches rio, ieactanssgroundat inachises ssouid be conaered ungrounded, volts, the standard unit contains } microfarad so no in- crease is required for ungrounded machines. Tn some applications it may be expedient to use an arrester from the neutral point of the machine to ground. For all machines larger than 1000 kva, special station type arresters should be used at the generator terminals For machines of less than 1000 kva station type arresters ‘may not be justified economically. Special line-type ar- reetera can be used. In all cases, standard line-type arresters or protector tubes are placed out on every overhead line entering the station at generator voltage. For 2300-volt circuits and above, these arresters should be located approximately 1500 feet from the station, For voltages below 600 volts, these arresters can be located within 500 feet of the station. ‘The possibility of a lightning stroke fo the line near the station can be minimized by overhead ground wires placed over this part of the line as indicated in Fig. 47 (b) spacing and location, with respect to the phase wires, is important and the application requires special study. There isa limit, however, to what can be done with ground wires, Tt should be stressed that the lightning strokes terminating on low-voltage circuits are just as severe as those terminating on high-voltage cireuits. Therefore, to get_good protection with overhead ground wires, the equivalent of w Ligh-vollaue Tine, widh harye spacing dl increased surge insulation from line wires to ground con- ductors should be used for the first 1500 or 2000 feet from the station. Properly applied ground wires then are ex- pensive on low-voltage circuits, especially if considerable Toney has to be spent to decrease the ground resistance. In many cases, it is more economical to use the choke-coil scheme of surge protection. If the machine is connected to the overhead line through a short cable, Fig. 47 (c), the protection at the machine should be the same as discussed above. In addition there should be a set of line-type arresters on the line 1500 to 2000 feet from the cable pothead and another set at the cable pothead. If several eables are connected to the bus or machine and their total length exceeds 2500 feet the cable capacity acts to slope the wave front, and the eapac- itor can be omitted at the machine, 8 Insulation Coordination Chapter 18 19. Choke Coil and Capacitor Method ‘The most, complete protection of rotating machines connected directly to overhead lines is obtained when lightning arresters are used to limit the magnitude ot the incoming surge, and lumped inductance and capacitance sure used to limit the slope of the incoming surge. With this scheme of protection, the machine is given full pro. tection for all surges, even for direct strokes to the over. head line close to the station. Special lightning arresters are paralleled with the re- quired amount of capacitance and tied to the generator terminals or station bus. See Fig. 48. A standard arrester is applied on the line side of the choke coil to limit the ES Ui WBE ao TaN MACHINE! CHOKE cow sstanoaRo SPECIAL voltage that determines the charging rate of the eapacitor, ‘The rate of rige of the terminal voltage depends upon both the amount of inductance and amount. of capacitance used. ‘The values that should be used to limit the rate of tise to the maximum permissible value are given in Table 10, A study has shown that the minimum capacitance to use in conjunction with a 175-microhenry choke is } ‘Tanta 10—Recowsemsnen Caracerawen Varsies ron CHOKR Colt AND Capacitor Meio oP PROTECTING ROTATING MACHINES Machines | Number of Standard Capacitors Cirouit eal et Phase Neutral pe Volesee | Connection |~ T 115 uh Choke | 350 4h Choke 650 any 2400 Grounded 1 2400 | Ungrounded 2 1 ‘1100 Grounded 2 . 160 Ungrounded 2 1 4800 Grounded 1 a 4800 | ‘Ungrounded 2 2 6900 Grounded | 2 6000 Ungrounded 2 1 11 500 Grounded 2 1 AL 500 Ungrounded 4 2 13 &00 Grounded 2 1 13 son ingrantnded 4 2 oe peviral considered grounded {grovel troughs centr of 60 0b ‘Heetiucearounied machines sould be conandered ungrounded Chapter 18 microfarad for grounded machines. If the inductance is increased to 350 microhenries, } microfarad would be sufficient For all machines above 1000 kva, special station-type arresters should be used at, the generator terminals: ‘whereas below 1000 kva special line type arresters can be used. In all cases, standard line or station-type arresters are located on the line side of the choke eoil. Station-type arresters are recommended, but, where it is not felt that thoy can be justified, line-type arresters ean be used; the dogree of protection expected will dictate the arrester to use This scheme is more expensive, but gives decidedly more reliable protection because the area close to the sta tion is fully protected 20. Machines Connected to Overhead Lines Vhrough Transformers Experience with machines connected to overhead lines, throngh transformers has indicated that damage to the machines from lightning surges on the overhead lines is rare if adequate arrester protection is provided on the high-voltage side of the transformer. However, surges coming in over an overhead line may produce high volt~ ages on the low-voltage side of the transformer, even if i ¥ Fig. 49—Protection of rotating machines connected through ower transformers to overhead lines. the high side is adequately protected with arresters. ‘The surge is transmitted through the transformer by both clectroatatie and electromagnetic coupling. ‘The cleetro static coupling depends tipon the transformer capacitances between windings and to ground, and is independent, of reactance and turns ratio. ‘The electromagnetic coupling depends upon the turns ratio, reactance and size of the trausformer, as well as Une bank connection, Ub is, whether star-delta, star-star, ete Where additional protection to the rotating equipment is desired, it ean be obtained by connecting a special ar- rester and one standard capacitor unit to each phase ter- 7 aie GROANS WIRETIF SE} — 7 i POWER TRANSFORMER, ! ! rovarins. | Sto sation wagune | fe annestca | ' + | ! | 1 | 1900 FF i STANDARO src 7 Lgtnetrire anmearen Lannesten ‘50—Protection of rotating machines connected direct to and through transformers to overhead lines, Insulation Coordination 6a minal of the machine, A standard lightning arrester is, of course, required on the line side of the transformer to protect the transformer. See Fig. 49. ‘When the machine is connected to an overhead line hoth direct and through transformers, the special arrost= ers and capacitors should be applied at each machine terminal the same as when the machine is connected to the overhead line only. Standard line-type arresters should be located from 1500 to 2000 feet out on the direct connected line. Standard station type arreaters should be installed on the high side of the transformer. See Fig. 50. 21. Machines Connected to Overhead Lines ‘Through Feeder Regulators or Current Limit- ing Reactors If the machine is connected to the ovorhoad line through a feeder regulator or current limiting reactor, the protec- tion should be the same as for a machine directly con- nected to the overhead line. In addition, a standard line TT ave ROTO TIRE TE TREDI Gorgwal acHINE | REACTOR OR REGULATOR oO | Apkre 1800 FP Fig. 51—Protection of rotating machines connected through current limiting reactor or feeder voltage regulator to over- head lines, = (A reactor, when present, is equivalent approximately to the effect ‘of using a choke evil, Arreater A may be omitted if the eurront limiting reactor has an inductance of over 350 microhenries.) type arrester should be added on the line side of the regu- lator or reactor. See Fig. 51 Te dhe iuductauce of he curreut Tnniting reactor is greater than 350 microhenries, the arrester on the line can be omitted. 22. Characteristics of Special Lightning Arresters Special low-breakdown lightning arresters for connee- tion in parallel with capacitors at the machine terminals (or at tho bus if two or more machines are connected to 2 common bus) are available in either the station type or line type. ‘The characteristics of typical arresters of both types are shown in Table 11. The station type arresters have a somewhat lower breakdown voltage, a lower dis- charge vollage st high surge currents, and a higher surge current discharge capacity. For the best. protection, station-type arresters should be applied to machines of all ratings. However, it is recognized that applications involving small sized ma- chines may not economically justify the most expensive protection. For that reason, line-type arresters are gen- erally applied on machines 1000 kva and below, and station-type arresters on machines larger than 1000 kva Either arrester must be applied on the basis that the power voltage from line to ground across the arrester 12 ‘Tantm 11—PRRFORMANCE CHARACTERISTICS OF SPECIAL LIGHT- ‘NING ARRESTERS FOR SURGE PROTECTION OF ROTATING MACHINES Gap Breakdown Front of : Wave | Wave Vote cats] Bont af Type | ta | cacycte| of Stand. | 10 Micro | us Wave Re Time] ard ATBE | seonds to | 500 Amp ate of [Breakdown Rive, | Rv Crem” Ke Crest ¢ ) 8 | 2 9 Sect} 8 Joe | om | om Staion] 9 fot | a0 | am Tw | 2 | | a | a 5 |» | m |e 0510 ai | 2s speat| 8 | 4 3 . line | 6 | x4 x | oe Tye | 9 | at nol on J | om a | {5 at 45 6 Breakdown when uel with epadion under any normal or fault condition does not exceed the arrester phase-leg rating, Vill. SUMMARY ‘The problem of coordinating the insulation of electrical equipment has progressed through years of research from a subject only vaguely understood to a sound engineering, practice based on well defined principles and known facts, Basic insulation levels have now been established that fx the lower limits of insulation surge strength to definite values that can be demonstrated by standardized test methods. Protective devices are available that will pro- vide a high degree of protection to insulation meeting the basic levels. Effective schemes have been devised for pro- tecting insulation that requires special consideration. ‘This progress was made possible only by cuoperation between the manufacturers and users of electrical equip- mont. in obtaining invaluable information on the nature of lightning and its effects on equipment in service. Con- tinued cooperation will undoubtedly produce additional information that will make possible further improve- ‘ments to the methods of coordinating insulation, REFERENCES 1. Standard Basie Impulse Insulation Levels, A Report of the ‘Joint Committee on Coordination of Insulation A.LE.E, E.E. and N,E.M.A, E,E.L. Publieation No. H-8, N.E.M.A, Publica tion No, 100, A...E.B. Pransactions, 1041. ALL-B.E. Lightning Reference Book 1918-1035. Coordination Session Toronto Convention 1990. Seven Papers ALL, Transactions 1180, 4, Recommendations for High-Voltege Testing, E.E.1-N.E.M.A. Subcommittee Report, A.J-E.E, Transactions, Vol. 58, page 598. Insulation Coordination 10. aL. 12, 13. M. 15. 16. 18, wv. 2. 21 2, 2, 26, ar. a1. 2 Chapter 18 Flashover Voltages of Insulators and Gans, A.LE.R. Sieom. mittee Report, A-1.E.8. Transactions, Vol. 58, page 882. Flashover Characteristios of Insulation, P. H. MeAuley, Ble. lwieFournal, uly 1036, Protection of Power Transformers Against Tightning Surges, A.LE.E. Committee on Electrical Machinery, Transformer Sub. commitice, ‘Technical Paper 41-79, Report on Apparatus Bushings, A.LE-R, Joint Committee on Bushings, Technical Paper 41-76. ‘The Control Gap for Lightning Protection, Ralph Higgins and HL, Roden, A.L-E.E. Transactions, 1936, pege 102). Line ‘Type Lighting Arvester Perforuaee Characteristies, ALER. Lightning Arrester Subcommittee, Technics! Paper 41-138. Station-Typo Lightning-Arrester Performance Characteristics, ALLEL, Lightning Arrester Subeormittee, Elecrial Engineer fino, June 1940, Selection of Lightning Arrestor Ratings, R. D. Evans and Ed. ward Beck, Westinghouse Engineer, February 1942 A Traveling Wave Frimer, award Beck, Eecric Journal, March 1982 to October 1982 inclusive ‘Transmission Tapped for Distribution by New Unit, George 8, Van Antwerp and HS. Warford, Electrical World, 1938, Shiekding of Substations, C. F. Wagner, G. D. MeCann and ©.M Tear ATRR Transactions, Vol. 61 Lightning and Lightning Protection i Distribution Systems, R.C. Bergvall and Edward Beck, .f.E.B. Transactions, Vol.59, Distribution Transformer Lightning-Protection Practice, L. G. Smith, A.LE.E. Transactions, Vol. 57. Surge Proof Distribution Transformers, J. K. Hodnette, Ble tric Journal, February 1982 ‘Surge Protection for Distribution Transformers, J. K. Hodnette, Blealie Journal, March 1036, Direct-Stroke Protection of Distribution Transformers, H. V. Putman, Elecric Journal, February 1937. Protection of Hotating A.C. Machines Against traveling Wave ‘Voltages Due to Lightaing, W. J. Rudge, Jr, R. M. Wieseman and W. W. Lewis, A.L.E.E Transactions, Vol. 52. Surge Protection for Rotating Machines, J. F. Calvert, A. C. Monteith, and E. Beck, Bletrie Journal, March 1933. Application of Arresters and the Selection of Insulation Levels JOH, Foote and J.R. North, Blecrical Engineering, June 1087. ‘Application of Station-Type Lightning Arresters, A. C, Mon teith and W. G. Roman, Blecri Journal, March 1988. ‘The Insulator String, R. W. Sorensen, A.E.E. Transactions, 19834, Vol. 53 Failure of Disk Insulators on High Tension ‘Transmission Sy tems, H. D. Panton, A.E.E, Transactions, 1722, Vol ‘Trausunonion of Electric Power xt Balen Tigh Voltages, Phi Spora, A.C. Monteith, A.I-B.E. Pronsactions, Vol. 66, 1947- ‘Corona Considerations on High-Voltage Lines and Desiga Pes- tures of Tidd 500-kv Test Lines, C.F. Wagner, Anthony Wag: ner, E.L, Peterson, 1. W. Gross, 4...E.B. Transactions, Vol. 68, ass, ). Transformers and Lightning Arresters—Tidd 500-ky Test Lines, F. A. Lane, J. K. Hodnette, P, 1. Bellaschi, Edward Beck, ALLELE. Transactions, Vol. 9, 194 Flashover Charaeteristies of Rod Gaps and Insulators, A.LE.E- Committee. Flectrical Engineering (A..E.E, Transaction’), Vol. 56, June 1987, pages 712-14, Lightning Performance of 200-kv Transmission Lines—II, ALEE. Committe, ALE. Transactions, Vol. @5, 1946, panes 70-5. ‘The Charaeteristios and Performance in Service of High-Voltag® Foreelain Insulators, J. S. Forrest, Journal, Tastitution of Ble trical Engineors (London, England), Vol. 89, part 11, No. 7 February 1942, pages 00-80. He

You might also like