415 CHAPTER XIV BUSBARS AND BUSBAR CONNECTIONS Two materials offer themselves as commercially suitable for use as busbars and busbar connections in switchgear design, namely copper and aluminium. The relative properties of these are given in the table below. Copper | Aluminium Weight per unit ledath for saul con- ductivity . . 10 050 Condtuetivity for equal areas:— Electrical. oe a 1'0 0-61 Thermal . a 10 0°56 Tensile Strength (Hard drawn) i 10 0°40 Hardness (Hard drawn) . . ro 0-44 Modulus of Elasticity rm ro oss Co-efficient of Thermal Expansion Fe ro 1°39 Melting Point .. a 10 0-61 This table shows that for equal conductivity, aluminium is lighter than copper and while this is an advantage in some applications, it is not usually important in many of the indoor types of switchgear up to 33kV. Here, weight is not a deciding factor; rather it is bulk that counts, particularly in metalclad and other enclosed types, and as aluminium conductors must have about 60 per cent greater sectional area than copper conductors for the same current rating, the latter material is often preferred. In other circumstances, where space is of less consequence, aluminium finds many uses, especially for example in open busbar runs such as occur in furnace installations, electroplating plants, large battery rooms as at telephone exchanges, and vertical risers in multi-storey buildings. Aluminium is not subject to “‘progressive oxidation” in that the oxide film is continuous, very hard and adherent and is stable, thus sealing the metal from further oxidation. This is important when an installation has to operate in high temperatures, or in polluted atmospheres. Busbars and busbar connections for current ratings of 200 amperes and above, as components of electrical switchgear, are covered by B.S.159, the416 THE J. & P. SWITCHGEAR BOOK busbars or connections having air, cil, solid or semi-solid materials as the Principal insulation. When. copper is used, hard-drawn bars are preferred to cold-rolled because of their better surface finish and because hard-drawing results in increased strength. Copper has a resistivity at 20°C. of about 1°76 microhms per cm. per sq. cm. section for drawn bar or 1°72 microhms for annealed bar, while aluminium has a resistivity of about 2°87 microhms per cm, per sq. cm. section. Both copper and aluminium sections are supplied dead to size within very close limits, the tolerances being given in a series of British Standards, of which B.S.1432 and B.S. 1433 apply for flat bars and round bars in copper and B.S,2898 for these sections in aluminium. The value of small tolerances lies mainly in the fact that switch blades (for example) can be made direct from the bar as recieved without machining. CURRENT-CARRYING CAPACITY At one time the current-carrying capacity of electrical conductors was based on the rule of 1 coo amperes per sq. inch. This rule was very rough and ready and led to the uneconomical use of material, particularly where the conductors were of small section and where densities well over 1 000 amperes per sq. inch could be used. The feature which should control the current-carrying capacity of a conductor is that of temperature and to this end B.S. 159 states that the temperature rise of busbarsand busbar connections when carrying rated normal current at rated frequency shall not exceed 50°C and that this rise is based on an ambient temperature having a peak value not exceeding 40°C and an average value not exceeding 35° measured over a 24-hour period. Based on this it is clear that the temperature of any busbar or connection should not exceed go°C. Above this figure oxidation increases rapidly and may give rise to cumulative and excessive heating at joints and contacts. ‘There are very many variables which affect the temperature to which a conductor will rise. In brief they include the number and arrangement of Taminations, whether bars are on edge or laid flat, whether freely exposed to the air or surrounded by insulation, the proximity of a surrounding casing and the material of which the latter is made. ‘The amount of heat generated in a conductor is proportional to the resistance of that conductor and to the square of the current it carries, while the temperature rise depends on the rate at which the heat is dissipated, the latter taking place in varying degrees by convection, radiation and conduction. Busbars completely surrounded by insulation will have the heat removed solely by conduction in the first instance, while bars freely exposed to the air will have the heat dissipated by convection and radiation. The maximum rate of heat dissipation with rectangular sections occurs when the sections are thin thus having a long perimeter for a given cross-sectional area. When. a number of rectangular bars are arranged in parallel their current-carrying capacity is reduced becuse their surfaces will dissipate less heat, the various bars effectively shielding adjacent bars. Thus if the current-carrying capacity of a single rectangular section is known then that of a multiple bar can be determined approximately by multiplying by an appropriate factor from the following —BUSBARS AND BUSBAR CONNECTIONS a7 D.C. AC. 2 laminations . 1-8 2 laminations 2 in. wide « 1-74 30» 2 ” 3m on + 170 4 ” 2 ” 4 | + 166 5 oo» 3 » Zn + 2530 6 ” 3 ” Bm + 2120 8 ” 3 ” 4 + 209 10» 4 ” Bon 245 4 ” 4 + 230 Both the above sets of figures are given by the Copper Development Association as being applicable to bars ¢ inch thick and with 4 inch spacing between laminations. In the case of a.c., increasing the number of laminations does not tly increase the current-carrying capacity because, as Fig. 14-1 shows, in a composite bar of ten laminations, the current density in the outer strips is about four times that in the centre strip. In general, a multiple bar of four or five laminations is an economical limit from a current-carrying point of view. 1000 8 3 Ampere 3 8 1 § S Current. Bos 8 ” 4 5 6 #7 8 9 10 Number of Lamination Fic. 14-1—Alternating current distribution in a bar of ten laminations (C. F. Wagner, “Electrical World,” Vol. 79 (1922)).418 THE J. & P. SWITCHGEAR BOOK Beyond this it pays to look at other formations and to illustrate this, Fig. 14-2 shows various formations for four square inches of conductor. The comparative a.c. ratings are given above the columns and it will be noted that as between the first last columns there is a difference of 85 Per cent. 185 st 125128 18 | Le Fic. 14-2.—Comparative a.c. ratings of various conductor arrangements of equal cross-section, i.e. four square inches (Copper Development Association). These relationships must however be regarded as approximate only, as much depends on the surroundings. The comparisons for example assume conductors freely exposed to the air and if they are enclosed then very different relative ratings might apply. ‘The determination of the actual current which a given section of conductor will carry and not exceed the temperature limits noted earlier is a matter of some complication. Formulae for the determination of the direct current carrying capacity of copper conductors have been given by the Copper Development Association as follows:— Flat bars Tae = 678A%S . por? Round bars (solid) Tae = 1080A°** where Ie = Direct current in amperes. A. = Cross-sectional area in square inches. p = Perimeter of conductor in inches. These formulae assume a temperature rise of 50°C over an ambient of 30°C but it is stated that if for a 40°C ambient, an error of about 15 per cent will probably be less than those implied by other assumptions{BUSBARS AND BUSBAR CONNECTIONS 419 which, inevitably, must be made. If it is desirable to limit the temperature rise to a lower figure than 50°C, the following factors may be used:— For 30°C rise multiply by 0°757 For 40°C rise multiply by 0-887. For aluminium conductors, the following formulae have been advanced to determine the direct current carrying capacity. where Tae, A and p are as for copper. D = Diameter of rod in inches. k = Constantfas follows:— Flat bars 385 for a 40°C rise. ” 438 for a 50°C rise. Round Bae = 659 for a 40°C rise. » n= 749 for a 50°C rise. It is of interest here to note that the current carrying capacity of a bar can, without increase in temperature rise, be increased by about 20-25 per on, painting the surfaces with a dull black paint (matt finish non- ate) leaving the question of direct-current ratings, it is well to note that Thomas and Rat Rata make no endeavour to establish these ratings by the use of formulae in their book ‘‘Aluminium Busbar” (see bibliography) and make the following statement:-— “Several attempts have been made to establish the law connecting the current-carrying capacity with temperature and busbar shape in mathematical terms, notably by Melson and Booth, but it should be noted that their Particular formula applies only to bars that are painted perfectly black and situated in still unconfined air. Further it deals only with one conductor per circuit and hence the majority of practical cases are not covered by the formula. Ratings for aluminium alloy busbar given in this book have been established by direct experiment and not by calaculation.” Busbars and conductors for use on a.c. systems require further study to establish a current-carrying capacity. This is due to the fact that whereas direct current distribution in a conductor is controlled only by the resistance of the various parts of the conductor, giving uniform current density regardless of shape or the position of return conductor, a conductor on a.c. is subject to inductive effects as well as resistance, leading to variations in current density over the section of the conductor, and to the fact that the current density may be distorted by the current in adjacent conductors. These two phenomena are known as “skin effect” and “proximity effect” respectively. Skin effect is that effect which tends to crowd the current into the outlying parts of a conductor and is due to the back e.m.f. induced wherever an alternating flux exists by virtue of this flux cutting the conductor. emf. is Dede in a bar by its own magnetic flux and while the central portions of the conductor are cut by the flux from all parts of the bar, the420 THE J. & P. SWITCHGEAR BOOK outlying portions are not, the line linkages decreasing as the edges are Current tends therefore to crowd into the edges of a flat strip, the outer members and edges of a multiple bar, and the skin of a circular Ber, where the opp ing e.m.f. is a minimum. Proximity efiect is due to the Proximity of an adjoining or adjacent pole or phase. When a bar is cut by the flux from such adjoining or adjacent conductor, the current tends to crowd into the portion of the conductor nearest the other bar. Both effects have one result—an increase in the apparent resistance of the conductors and as J*R is the measure of the heat generated, the my % ct Coz" eee a 4 1 ¥ 2 Cross-sectional Area — Sq. Inch Fic. 14-3.—Shin effect in rectangular high conductivity copper bars at 20°C (Copper Development Association).BUSBARS AND BUSBAR CONNECTIONS 421 temperature rise of two bars of equal section carrying the same current will be higher for that bar carrying alternating curreht. For heavy current’ applications, both effects can be countered to some degree by careful con- sideration in the design and layout of the conductors, Skin effect, for example, can be reduced by adopting a form such as the “hollow square’ or by using tubular bars. In both forms as much copper as possible is located equidistant from the magnetic centre of the bar. Frosty effect may be reduced by spacing the conductors to the maximum possible or by some special arrangement to ensure that no bar is adjacent to another. ‘When conductors are close together, as is often the case in low-voltage switchgear where the electrical clearances demanded are small, and in compound-filled busbar chambers, proximity effect can be pronounced. The problem of ratings is further complicated by reason of the fact that in perhaps the majority of switchgear designs, the conductors may be enclosed in non-magnetic or magnetic enclosures, each form having its own effect on the current rating of a conductor, resulting in some reduction in the “free air” rating. An approximate value for the latter may be determined, for isolated flat bars, from the formula:— _ _lae - 5% R where R! = Apparent a.c. resistance. R =D. resistance. and R}/R is the skin effect ratio. This ratio for rectangular bars may be read directly from the curves in Fig. 14-3 against cross-sectional area and for various ratio of width to thickness. Where this ratio is large the curves a/b = 240 should be used. The formula, it should be noted, applies only to isolated (i.e. single) flat bars and factors by which the value so determined should be multiplied for multiple bars have been noted on page 417. A study of the curves reveals that for all practical purposes in single flat copper bars up to about 0°5 sq. in. cross-sectional area, the a.c. current rating can be assumed to be equal to the d.c. rating. This is true also for copper rods up to about 0°7§ sq. in. In arriving at an a.c. rating by the formula given it has been assumed that the conductor is at such a distance from the return conductor that the effect of current in the latter can be neglected, i.e. the rating is not affected by proximity effect. The magnitude of this depends among other things, upon the frequency and on the spacing and arrangement of the conductors. ‘his problem has been dealt with in some detail in the book “Aluminium Busbar” by Thomas and Rata and these authors state that ‘Proximity effect can safely be regarded as introducing a change of 5 per cent or less in the rating if the current is less than 2 000 to 3 ooo amperes and the voltas exceeds 1 000 volts, because the busbar will not then be too large nor will the spacing be too close.”’ In the case of rod conductors, used with considerable advantage in certain designs of switchgear, and particularly at the higher voltages, the skin effect ratio R1/R for copper is obtained from the curves Fig. 14-4. Tac422 THE J. & P, SWITCHGEAR BOOK Methed of Use— Calculate X from X=0-3065V At ‘Where A = Area of rod in sq. in. = Frequency in cycles. From X on base line project vertically to curve and from point of Intersection Project herizontally to left to read lo RY, Wren) isles than 2 use inset scale ‘This chart is adapted from Fig. 5 in “Copper for Busbars”” (Copper Development Association) Fic. 14-4—Skin effect in high conductivity copper rods at 20°C.BUSBARS AND BUSBAR CONNECTIONS 423 Based on the foregoing data, Tables 14:1 to 1 44 which follow give the current ratings for a selection of sizes of fecbagular bo rs and round solid rods.: TABLE 1431 APPROXIMATE DIRECT CURRENT RATING (INDOORS) FOR H.C. COPPER STRIPS AND BARS i | Dimensions Amperes Dimensions ms | Amperes xt" 225 | 2220 a 325 | oS im | 1080 at, | 325 ; 3txe | 1520 yxt | 470 | 4d 1920 2" xt" | 590 5x 2320 1" x} 485 ox" 2740 ey | 675 2"xt" | 1270 2* x} 860 | atxdt 1770 ayxe | 1040 ext 2250 37 xe | 1220 | sxe | 2730 4, xh | 1560 ext 3.190 ig” Ree | 1900 | B"xe | 4090 Ampere ratings are for single bars mounted. on edge, freely exposed to still air and for a temperature rise of 50°C above an ambient of 30°C (see comment on page 418). For a 30°C rise multiply by 0-757. For a “°C rise multiply by eee For a.c. ratings see formul 21. For multiple bars multiply by factors on page 417. This able extracted from Table 2 ‘Copper for Busbars” (Copper Development Association). JOINTS The first essentials for any joint that has to be made between busbar sections or between teed conductors and a busbar are (a) that it shall be mechanically strong and (b) it shall have a relatively low resistance at all times, ie. its electrical efficiency should remain stable under all service conditions. The simplest and most widely used method of making a joint is by bolting eae the sections a method which is simple and flexible, and allows the joint to be ismantled if required. Other methods include riveting, soldering or welding but in each of these the joint must be regarded as permanent.424 THE J. & P. SWITCHGEAR BOOK TABLE 14:2 APPROXIMATE DIRECT CURRENT RATING FOR H.C. COPPER RODS Diameter | Indoor | Outdoor Indoor | Outdoor amperes | amperes Diameter amperes | amperes = |e | aos |p| ast | aoe ¥ 240 340 1p" 1960 2510 r 300 485 2" 2350 3015 ” 480 645, 24" 3.190 4.080 ¥ 620 810 ac | 4030 5215 r 760 990 38” | 5020 6 435 , 910 1185 4 | 6040 7710 13” 1240 1595 Ampere ratings determined on the assumption that the conductor | is mounted horizontally, freely exposed to the air and a temperature | rise of 50°C above an ambient of 30°C (see comment on page 418.) For indoor installations still air conditions have been assumed. For a 30°C rise, multiply by 0-757 and for 40°C rise multiply by 0°887. For a.c ratings at 50 cycles see formula page 421. This table extracted from Table 3 “Copper for Busbars’’ (Copper Development Association). - - J In whatever method is employed, certain precautions must be taken to ensure efficiency, namely:— (a) The contact pressure must be ample and be maintained. (b) The conductor surface must be clean before the joint is made. (c) For aluminium bars and bimetallic joints, air and moisture must be excluded. (d) The overlap must be equal to or greater than the width of the busbar. In passing it may be noted that while bolted joints are compact and rehable, they have the disadvantage that holes must be drilled or punched through the conductors thereby reducing the effective area. This is avoided if clamped joints are employed and because of the extra mass of metal surrounding the joint it 1s cooler in operation. The disadvantage is the extra bulk which becomes a nuisance in certain classes of gear with metal- enclosed busbars. In the design of any joint the gross contact area is not of primar importance, for it has been established that current is transferred through the joint at numerous point contacts between irregularities on the contacting ‘aces. It is well known that the number and effectiveness of these point contacts ds almost entirely on the pressure, and its distribution at the joint, whether by bolts or clamps. The effect of pressure on the contact resistance of a joint is demonstrated in Fig. 14-5, while the total pressures required per inch of width for joints to give 100 per cent efficiency areBUSBARS AND BUSBAR CONNECTIONS 425 TABLE 14:3 CURRENT RATINGS FOR NORAL D 50 SWP ALUMINIUM ALLOY RECTANGULAR: BUSBARS 1 Bar 2 Bars 3Bars | 4 Bars Size AC. AC. | AC. AC. DG. | eS eis| PO |soeis| DO |eocis| PK | so'els ‘Amperes | Amperes | Amperes | Amperes 1" xf 356] 356| 78] 715| 980] 970| 120|1100 yxt 520] §20| 1030 | roz0 | 1 380| 1 350 | 1 585 | 1535 2” x} 672 | 670 | 1315 | r 290 | 1 765 | 1 705 | 2 050] 1 940 ay xe 820| 812| 1550] 1 510 | 2 100 | 2.000 | 2 430 | 2 260 ; we" 970} 958 | 1 805 | 1 740 | 2 440 | 2 310 | 2 860 | 2 620 4” xy” 1 260 | 1 235 | 2 260 | 2 140 | 3 060 | 2 800 | 3 640 | 3 200 5” xP" | 1545 |r 505 | 2 700 | 2 sto | 3 660 | 3 240 | 4 410 | 3 700 6" af 1 B40 | 1 780 | 3 130 | 2 860 | 4 290 | 3 680 | 5 250 | 4 240 a" xe 840 | 830 | 1 560 | r 00 | 2 090 | 1.970 | 2 460 | 2 260 3" X# | 1210 | 1 180 | 2 180 | 2 050 | 2 940 | 2 660 | 3 510 | 3 030 4 XE | x 550] 1495 | 2710 | 2 480 | 3 660 | 3 150 | 4 400 | 3 560 3 ae 1.940 | 1 860 | 3 290 | 2.930 | 4 450 | 3 660 | 5 400 | 4 200 6” xB" | 2.260 | 2 120 | 3 770 | 3 340 | 5 140 | 4080 | 6 300 | 4 680 8” xB" | 2.940 | 2.750 | 4 B00 | 4 150 | 6 500 | 4.900 | 8 060 | 5 740 3° X# | 1405 | 1355 | 2 450 | 2 240 | 3 290 | 2 830 | 4 000 | 3 240 4° X<4#" | 1830 | 1740 | 3 100 | 2 720 | 4170 | 3 360 | 5 100 | 3 900 5" XB | 2.230 | 2 080 | 3-720 | 3 120 | 5 040 | 3 900 | 6 170 | 4.550 6" x4" | 2 620 | 2 420 | 4 300 | 3 500 | § 850 | 4 400 | 7 200 | 5 100 8" xh" 3 380 | 3 060 | 5 450 | 4 450 | 7 420 | 5 300] g 110 | 6 150 10" x4 4 080 | 3 640 | 6 500 | 5 000 3 H6e [een 0 goo | 6 850 | Noral D so SWP (Magnesium Silicide Alloy) is the product of Alcan Industries Ltd. For their Noral CISM (99:5% pure aluminium) busbar, multiply by 1-03. The designations are Proprietary ones for alloys corresponding to EgtE and E1E materials to BS. 2808. Ratings are based on a 50°C rise over 35°C ambient temperature in still but unconfined air. For multiple bars,the space between bars is equal to the bar thickness. Bars assumed mounted vertically on edge and a.c. ratings are based on spacings at which proximity effect is negligible. Ratings may be increased approximately 20% if the busbars are painted with non-metallic matt finish paint. This table extracted from Table 2 “Aluminium Busbar” (see bibliography).426 THE J. & P. SWITCHGEAR BOOK TABLE 14:4 CURRENT RATINGS FOR NORAL D 50 SWP ALUMINIUM ALLOY SOLID ROUND BUSBARS: Diameter D.C. amperes A.C. 50 c/s amperes F 155 155 ¥ 230 230 : 335 335 a 435 435 550 548 rf 690 685 ry" 945 925 14" 1200 1160 Noral D 50 SWP is the product of Alcan Industries Ltd. It is the proprietary alloy designation for Eg1E material to B.S. 2898. Ratings are based on a 50°C: rise over 35°C ambient temperature in still but unconfined air. A.C, ratings are based on spacings at which proximity effect is negligible. This table extracted from Table 5 ‘Aluminium Busbar” (see bibliography). given in Fig. 14-6. This, then, is the real method of judging a joint, ie., a method based on the clamping or bolting pressure. In designing to this method, however, it is essential that due rd be paid to the stresses set up in the bolts. It is usual that the latter will be of different material to the conductor and if the stress intensities in the materials of the joint are high when cold, they may be excessive when the joint is warm so that a permanent set in one part may arise. It should be borne in mind too that the bolts will not be at the temperature of the conductor since they are not directly heated by the current. Fig. 14-7 shows the approximate force per bolt for various sizes of bolts, and applies to bolts tightened with a spanner or wrench of normal dimensions, when used by an average workman. The amount of overlap at joints is not dependent on the current transfer, because as previously indicated, the whole of the area is not effective in this respect. Rather, the overlap is determined by the number of bolts necessary to obtain the required pressure. The provision of a good overla helps to preserve an efficient joint, particularly if coupled with the simy precaution of smearing the surfaces of the joint with vaseline, and scratch brushing just prior to making the joint.BUSBARS AND BUSBAR CONNECTIONS 427 z | « a ze — + 4 6 E44 j #2 4 pj | | 2 72000 4000 000 8000 To000 PRESSURE ~ LB. PER SQ. IN. Fic. 14-5.—Effect of pressure on the contact resistance of a joint between two copper conductors (Copper Development Association). 30) —_—__—_ 25 20) OVERLAP-INCH ad T 500 400045005000 i000 —*(1500——«2000~=S=«2800—«s 00s TOTAL PRESSURE PER INCH OF WIDTH-POUND Fig. 14-6.—Total pressures required per inch of width for joints to give 100% efficiency flat copper strips ¥° and 4" thick (Copper Development Association).428 THE J. & P. SWITCHGEAR BOOK eee DIAMETER OF BOLT-INCH wo ‘097000 2400-3000 «004000 43001000 3300 FORCE PER BOLT-POUND Fi. 14-7.—The approximate force per bolt developed by bolts of various sizes (Copper Development Association). In some cases, where joint surfaces are very rough or where high current densities are anavoidable or where highly corrosive atmospheres exist, surfaces to be joined together may with advantage be tinned. If this is adopted the tinning should be done immediately before the joint is clamped or bolted up and it should be noted that both the electrical conductivity and the protective action of a lead-tin solder decrease as the lead content increases. For jqints between aluminium conductors a joint compound may be smeared on the faces before bolting up and one such compound has been recently developed by Aluminium Laboratories Ltd. specially for the purpose and is available under the trade name ‘“‘Densal’”.* if it is necessary to make a joint between a copper and an aluminium conductor, it has to be remembered that two dissimilar metals will be in electrical contact and if moisture is present electrolytic action will arise. The joint must therefore be made absolutely moisture proof using a grease as a joint compound or “Densal’’ as previously noted. Alternatively a bimetal connector may be introduced. EXPANSION AND CONTRACTION When there are variations of temperature there will be either expansion or contraction of the conductor and unless some provision is made to accommodate changes in length, particularly on long runs, some damage may be caused either to the conductor itself or to the supporting structure. ‘The co-efficient of the linear expansion of copper may be taken as o-oo0or7 per degree Centigrade and that of aluminium as 0-000023 per degree Centigrade. * Manufactured by Messrs. Winn & Coales Ltd. Denso House, Chapel Road, London, S.E.27.BUSBARS AND BUSBAR CONNECTIONS 429 EXPANSION PER 100 FT.INCH -B=15 010-2030 a 50 G0 706090100 110 120-130-140 TEMPERATURE-DEGREE Fic. 14-8.—Thermal expansion and contraction of copper conductors relative to their length at 20°C (68°F) (Copper Development Association). Fig. 14-8 gives curves to determine the expansion or contraction of copper conductors relative to their length at 20°C (68°F). Thus the total change in length per too feet of a copper conductor between temperatures of o°C and 40°C will be 0-4+-0-4= 0°8 inches. The expansion or contraction of a copper conductor relative to any temperature other than 20°C can be ascertained by drawing the line of zero expansion through the point of mnterertion of the curves and the ordinates corresponding to the temperature required. In the case of very short bars, it is usually not necessary to make any special provision to accommodate expansion or contraction as within the normal temperature range the amount will be small and will be taken up by a certain amount of flexibility in the supporting structure. In other cases and where long rigid runs are involved, some form of expansion joint should be introduced at intervals, some typical joints being shown in Figs. 14-9 and 14-10.430 THE J & P SWITCHGEAR BOOK FLEXIBLE CONDUCTORS FLEXIBLE CONNECTION SHOULDERED NUT ELONGATED HOLE Fic 14-9 —Forms of expansion joints Fia 14-10 —Expansion joint for 1200 ampere busbarsBUSBARS AND BUSBAR CONNECTIONS 431 THE EFFECTS OF SHORT-CIRCUIT In Chapters III and IV it has been shown how the current which can flow when a short-circuit occurs can be calculated. It was seen that the magnitude of these currents can be exceedingly high and c and clearly any busbar or connection system must in some way be poe effects are two-fold. Firstly, ae a ¢ ech oases i conductor temperature to rise considerably due to the Passage of the value of current for a short period of time and secondly, the ae effect causing high forces of attraction or repulsion to be set up between conductors. Dealing with the thermal problem first, it may be stated that this is related largely to the time the fault current is allowed to flow. What that time is will depend on the characteristics of the nearest automatic protective device which should operate to clear the fault. If this device is a circuit-breaker the clearance time will depend largely on the form of protective gear and may vary from 0:02 to o°5 or even I'0 second,* Ifon the other hand the protective device is an h.r.c. fuse exhibiting a marked “cut-off” effect then the total clearance time may be as low as (0005 seconds or less depending on the fuse link rating and the magnitude and degree of asymmetry of the fault current. Because of various indeterminate but possible extensions of the total clearance time (sticky latching mechanisms for example) it is often expedient to allow for at least a clearance time of one second and sometimes more. BS 159 does in fact state: ‘The duration of the rated short-time current shall be:— (i) 3-seconds where the ratio of the rated short-time current to the rated normal current is equal to or less than 4o. (ii) 1-second where the ratio of the rated short-time current to the rated normal current is more than 40.” In short times such as those noted it may be assumed that all the heat produced is absorbed by the conductor and that there is not time for it to be dissipated from the conductor by radiation or convection. The thermal problem then is to determine the sectional area of the conductors to carry the calculated fault current for a defined time with a temperature rise which is safe not only for the conductor itself but for any insulation which may be in contact or in near contact and for sweated joints such as at cable sockets, etc. What this safe limit is, is a matter for decision taking all factors in a particular design into account and while some designers suggest a temperature rise of 175-200°C as being reasonable, others suggest a safer limit of 100°C. This latter figure has merit because it must be remembered that it is a rise above that at which the conductor may be running and this may well be 80-90°C so that with 100°C rise, a total temperature approaching 200°C is possible. It is worth pointing out here that solder which may be used at cable sockets will tend to soften at 180°C and at higher temperatures will run out. There are therefore many arguments in favour of a reasonably low temperature rise although this can only be at the expense of more metal in the conductors. *Where LD.M.T. relays are sed for tection of ring main or other circuits (see Chapter XV), definite ranma oping ig of fap wo 9 sere may be ingore bythe alge or discrimination purposes.432 THE J. & P. SWITCHGEAR BOOK ‘An approximate formula for calculating the temperature rise in degrees C per second above an initial conductor temperature of 30°C, due to a known current is as follows:— T=k ( I )’ 10 A, where T = Temperature rise per second, degrees C. I Current in amperes r.m.s. symmetrical. Sectional area of conductor, sq, in. k = 125 for copper; 2°8 for aluminium, A more accurate formula would be: T=k. (i): (1-48) 10-® Temperature rise per second, degrees C. Current in amperes, r.m.s. symmetrical. ectional area of conductor, sq, in. Temperature coefficient of resistivity at 20°C. per °C. 000393 for copper. 0°00386. for aluminium 9°0036 for aluminium alloy. @ = Temperature in degrees C of the conductor at the instant at which the temperatures rise is being obtained. k = 1°25 for copper; 2°8 for aluminium. This formula is not easy to use because of the difficulty of ascertaining 8. Alcan Industries Ltd. have produced a series of curves, reproduced here at Fig. 14-11, from which the value—_ YA. ft can be deduced directly for a known initial temperature and a desired final temperature, for both copper and aluminium (Noral CISM). ‘These curves are based on the formula— a 1+ (Or 2a] IVvt=kaA Fi loge [: Fe (Or 220) where I = Current in kiloamperes r.m.s. symmetrical. k = 146 for copper 96 for aluminium (Noral CISM). 92 for aluminium alloy (Noral D 50 WSP). a = Temperature coefficient of resistivity at 20°C per °C (values as previously noted). A = Sectional area of conductor, sq. in. 61 = Initial temperature degrees C. 6r = Final temperature degrees C. t=time in seconds. As an example of the use of these curves, let us assume that the essential data is:— J] = 46 kiloamperes. t = 3 seconds. 61 = 20°C. Or = 120°C (i.e. a rise of 100°C).BUSBARS AND BUSBAR CONNECTIONS 433 eae zs LE Fic. 14-11.—Curves to determine conductor areas for a known initial temperature and any desired final temperature (Alcan Industries Ltd.). Projecting a line across from 120°C on the vertical scale to the point of intersection with the curve for copper at an initial temperature of 20°C and then from this point a vertical line to meet the horizontal scale of I/A+/t values at about 85, we get:— #. V3 = 85 Ai eon = 67938 aq, in, For aluminium, the area required would be:-— # V¥3= 55 A= S115 = 1°44 8q. in.434 THE J. & P. SWITCHGEAR BOOK In the book “Aluminium Busbar” Thomas and Rata have produced a series of nomograms from which, given the essential data as in the foregoing ean le, the conductor area may be read directly for copper, aluminium juminium alloy. aan consider now the electromagnétic problem it must first be appreciated that whereas in the thermal problem we were concerned with the r.m.s. symmetrical value of current and how long it lasted, here it is the peak current which occurs in the first half-cycle of short-circuit which determines the magnitude of the forces between adjacent bars. This peak current will have a value dependent on the degree of asymmetry and the power factor and where the latter is low (0-15 or less) then it may be 2-55 times the r.m.s. symmetrical value and for power factors of about 0-3, it will be 2-0 times. Thus we find, taking one common example, that for a known short- circuit value of 31 MVA at 415 volts the short-circuit currents will be:— Symmetrical r.m.s. amperes: 43 300 Initial peak amperes 0°15 PF — 110 415 Initial peak amperes 0-3 PF —_86 600 In all formulae to determine the forces set up under short-circuit conditions, it will be seen that the current value has to be squared so it is obvious, with the figures for the initial peak as indicated above, that the forces may be quite high. The design of the supporting structure must therefore be auch as to ensure the safety of the busbars. The formulae given in Fig. 14-12 are taken from the book “Aluminium Busbar” by Thomas and Rata, and a study of these will show that in the case of 3-phase alternating current systems, the worst condition is that of a fault between two lines (for calculation of such a fault see Chapter IV.) All the formulae given are based on the assumption that the conductors are of circular cross-section and it is necessary to apply a correction factor when conductors of rectangular cross-section are used. This factor, denoted by the letter K is el from the curves in Fig. 14-13 from which it will be seen that the ratio tb has first to be calculated and then the value of K can be read from the curve for the correct ratio a/b. Having obtained K it is simply introduced asa ouultspliet into any of formulae in Fig. 14-12 thus:— 43g 0 EB the i run A study of these curves will show that for bars spaced far apart, K approaches unity, and that while it is a maximum for very thin conductors, it is practically negligible for bars of square section. That the forces due to short-circuit current can be substantial is shown in Table 14:5 where the approximate peak force in pounds per foot run is given for different values of MVA at 415 volts and based on the assumption maximum asymmetry i.e. the r.m.s. symmetrical value of fault current as determined from short-circuit calculations has been multiplied by 2°55. This table shows that by increasing the spacing between phases an easement in electromagnetic forces is obtained and although this is desirable in many ways, it can only be at the cost of larger busbar enclosures. It may noted in passing that increasing the spacing will increase the reactance of the busbars (see Chapter ITI). Fn =