Line insulators, pins, and cross-arms all go to make up paths of more or less conductivity between the wires of a transmission circuit. The amount of current flowing along these paths from one conductor to another in any case will depend on the combined resistance of the insulators, pins, and cross-arm at each pole.

As a general rule, the wires of high-voltage transmission circuits are used bare because continuous coverings would add materially to the cost with only a trifling increase in effective insulation against high voltages. In some instances the wires of high-pressure transmission lines have individual coverings for short distances where they enter cities, but often this is not the case. At Manchester, N. H., bare conductors from water-power plants enter the sub-station, well within the city limits, at 12,000 volts. From the water-power at Chambly the bare 25,000-volt circuits, after crossing the St. Lawrence River over the great Victoria bridge, pass overhead to a terminal-house near the water-front in Montreal. In order to reach the General Electric Works, the 30,000-volt circuits from Spier Falls enter the city limits of Schenectady, N. Y., with bare overhead conductors.

Where transmission lines pass over a territory exposed to corrosive gases, it is sometimes desirable to give each wire a weather-proof covering. An instance of this sort occurs near Niagara Falls where the aluminum conductors forming one of the circuits to Buffalo are covered with a braid that is saturated with asphaltum for some distance.

Each path, formed by the surface of the insulators of a line and the pins and cross-arm by which they are supported, not only wastes the energy represented by the leakage current passing over it, but may lead to the charring and burning of the pins and cross-arm by this current. To prevent such burning, the main reliance is to be placed in the surface resistance of the insulators rather than that of pins and cross-arms. These insulators should be made of glass or porcelain, and should be used dry—that is, without oil. In some of the early transmission lines, insulators were used on which the lower edges were turned inward and upward so that a circular trough was formed beneath the body of the insulator, and this trough was filled with heavy petroleum. It was found, however, that this trough of oil served to collect dirt and thus tended to lower the insulation between wire and cross-arm, so that the practice was soon abandoned. Glass and porcelain insulators are rivals for use on high-tension lines, and each has advantages of its own. Porcelain insulators are much stronger mechanically than are those of glass, and are not liable to crack because of unequal internal expansion, a result sometimes met with where glass insulators are exposed to a hot morning sun. In favor of glass insulators it may be said that their insulating properties are quite uniform, and that, unlike porcelain, their internal defects are often apparent on inspection. In order to avoid internal defects in large porcelain insulators, it has been found necessary to manufacture some designs in several parts and then cement the parts of each insulator together.

Defective insulators may be divided into two classes—those that the line voltage will puncture and break and those that permit an excessive amount of current to pass over their surfaces to the pins and cross-arms. Where an insulator is punctured and broken, the pin, cross-arm, and pole to which it is attached are liable to be burned up. If the leakage of current over the surface of an insulator is large, not only may the loss of energy on the line where the insulator is used be serious, but this energy follows the pins and cross-arm in its path from wire to wire, and gradually chars the former, or both, so that they are ultimately set on fire or break through lack of mechanical strength. The discharge over the surface of an insulator may be so large in amount as to have a disruptive character, and thus to be readily visible. More frequently this surface leakage of current over insulators is of the invisible and silent sort that nevertheless may be sufficient in amount to char, weaken, and even ultimately set fire to pins and cross-arms.

All insulators, whether made of glass or porcelain, should be tested electrically to determine their ability to resist puncture, and to hold back the surface leakage of current, before they are put into practical use on high-tension lines. Experience has shown that inspection alone cannot be depended on to detect defective glass insulators. Electrical testing of insulators serves well to determine the voltage to which they may be subjected in practical service with little danger of puncture by the disruptive passage of current through their substance. It is also possible to determine the voltage that will cause a disruptive discharge of current over the surface of an insulator, when the outer part of this surface is either wet or dry. This is as far as electrical tests are usually carried, but it seems desirable that such tests should also determine the amount of silent, invisible leakage over the surface of insulators both when they are wet and when they are dry, at the voltage which their circuits are intended to carry. Such a test of silent leakage is important because this sort of leakage chars and weakens insulator pins, and sets fire to them and cross-arms, besides representing a waste of energy.

The voltage employed to test insulators should vary in amount according to the purpose for which any particular test is made. Glass and porcelain, like many other solid insulators, will withstand a voltage during a few minutes that will cause a puncture if continued indefinitely. In this respect these insulators are unlike air, which allows a disruptive discharge at once when the voltage to which it is exposed reaches an amount that the air cannot permanently withstand. Because of this property of glass and porcelain insulators, it is necessary in making a puncture test to employ a voltage much higher than that to which they are to be permanently exposed. In good practice it is thought desirable to test insulators for puncture with at least twice the voltage of the circuits which they will be required to permanently support on transmission lines.

For the first transmission line from Niagara Falls to Buffalo, which was designed to operate at 11,000 volts, the porcelain insulators were tested for puncture with a voltage of 40,000, or nearly four times that of the circuits they were to support.

Porcelain insulators for the second line between Niagara Falls and Buffalo, after the voltage of transmission had been raised to 22,000, were given a puncture test at 60,000 volts. Of these insulators tested at 60,000 volts only about three per cent proved to be defective. These puncture tests were carried out by placing each insulator upside down in an open pan containing salt water to a depth of two inches, partly filling the pin hole of the insulator with salt water, and then connecting one terminal of the testing circuit with a rod of metal in the pin hole, and the other terminal with the pan. Alternating current was employed in these tests, as is usually the case (Volume xviii., Transactions A. I. E. E., pp. 514 to 520). For the transmission lines between Spier Falls, Schenectady, Albany, and Troy, where the voltage is 30,000, the insulators were required to withstand a puncturing test with 75,000 volts for a period of five minutes after they had been soaked in water for twenty-four hours.

There is some difference of opinion as to the proper duration of a puncturing test, the practice in some cases being to continue the test for only one minute on each insulator, while in other cases the time runs up to five minutes or more. As a rule, the higher the testing voltage compared with that under which the insulators will be regularly used, the shorter should be the period of test. Instead of being tested in salt water as above described, an insulator may be screwed onto an iron pin of a size that fits its threads, and then one side of the testing circuit put in contact with the pin and the other side connected with the wire groove of the insulator. Care should be taken where an iron pin is used either in testing or for regular line work, that the pin is not screwed hard up against the top of the insulator, as this tends to crack off the top, especially when the pin and insulator are raised in temperature. Iron expands at a much higher rate than glass or porcelain, and it is desirable to cement iron pins into insulators rather than to screw them in. There seems to be some reason to think that an insulator will puncture more readily when it is exposed to severe mechanical stress by the expansion of the iron pin on which it is mounted.

Tests of insulators are usually made with alternating current, and the form of the voltage curve is important, especially where the test is made to determine what voltage will arc over the surface of the insulator from the line wire to the pin. The square root of the mean square for two curves of alternating voltage or mean effective voltage, as read by a voltmeter, may be the same though the maximum voltages of the two curves differ widely. In tests for the puncture of insulators, the average alternating voltage applied is more important than the maximum voltage shown by the highest points of the pressure curve, because of the influence of the time element with glass and porcelain. On the other hand, when the test is to determine the voltage at which current will arc over the insulator surface from the line wire to the pin, the maximum value of the pressure curve should be taken into consideration because air has no time element, but permits a disruptive discharge under a merely instantaneous voltage.

Alternators used in transmission systems usually conform approximately to a sine curve in the instantaneous values of the pressures they develop, and it is therefore desirable that tests on line insulators be made with voltages whose values follow the sine curve. Either a single transformer or several transformers in series may be employed to step up to the required voltage, but a single transformer will usually give better regulation and greater accuracy. An air-gap between needle points is not a very satisfactory means by which to determine the average voltage on a testing circuit, because, as already pointed out, the sparking distance between the needle points depends mainly on the maximum instantaneous values of the voltage, which may vary with the load on the generator, and the saturation of its magnets. For accurate results a step-down voltmeter transformer should be used on the testing circuit.

An insulator that resists a puncture test may fail badly when subjected to a test as to the voltage that will arc over its surface from line wire to pin. This arc-over test should be made with the outer surface of the insulator both wet and dry. For the purpose of this test the insulator should be screwed onto an iron pin, or onto a wooden pin that has been covered with tinfoil. One wire of the testing circuit should then be secured in the groove of the insulator, and the other wire should be connected to the iron or tin foil of the pin. The voltage that will arc over the surface of an insulator from the line wire to the pin depends on the conditions of that surface and of the air. In light air, such as is found at great elevations, an arc will jump a greater distance than in dry air near the sea-level. A fog increases the distance that a given voltage will jump between a line wire and its insulator pin, and a heavy rain lengthens the distance still further. The heavier the downpour of rain the greater is the distance over the outside surface of an insulator that a given voltage will arc over. The angle at which the falling water strikes the insulator surface also has an influence on the voltage required to arc over that surface, a deviation from a downpour perpendicular to the plane of the lower edge of the petticoat of the insulator seeming to increase the arcing distance for a given voltage.

An insulator should be given an arc-over test under conditions that are approximately the most severe to be met in practice. These conditions can perhaps be fairly represented by a downpour of water that amounts to a depth of one inch in five minutes for each square inch of the plane included by the edge of the largest petticoat of the insulator, when the direction of the falling water makes an angle of forty-five degrees with that plane. A precipitation of one inch in depth on a horizontal plane during five minutes seems to be a little greater than any recorded by the United States Weather Bureau. Under the severe conditions just named, the voltage required to arc over the insulator surface from line wire to pin should be somewhat greater at least than the normal voltage of the circuit where the insulator is to be used. For the transmission line between Spier Falls and Schenectady, on which the maximum voltage is 30,000, the insulators were required to stand a test of 42,000 volts when wet, without arcing over from line wire to pin. In these wet tests the water should be sprayed evenly onto the insulator surface like rain, and the quantity of water that strikes the insulator in a given time should be measured.

When the outside of an insulator is wet with rain, it is evident that most of the resistance between the line wire and the insulator pin must be offered by the inside surface of the petticoat of the insulator. For this reason an insulator that is to withstand a very high voltage so that no arc will be formed over its wet outside surface must have a wide, dry surface under its petticoat. In some tests of line insulators reported in Volume xxi., Transactions A. I. E. E., p. 314, the results show that the voltage required to arc over from line wire to pin depends on the shortest distance between them, rather than on the distance over the insulator surface. Three insulators, numbered 4, 5, and 7 in the trial, were in each case tested by a gradual increase of voltage until a discharge took place between the wire and pin. The pins were coated with tinfoil, and the testing voltage was applied to the tie wire on each insulator and to the tinfoil of its pin. Insulators 4, 5, and 7 permitted arcs from wire to pin when exposed to 73,800, 74,700, and 74,700 volts respectively, the surfaces of all being dry and clean. The shortest distances between wires and pins over insulator surface and through air were 6⁵/₈, 6¹/₄, and 7⁷/₈ inches respectively for the three insulators, so that the arcing voltages amounted to 11,140, 11,952, and 9,479 per inch of these distances. Measured along their surfaces, the distances between wires and pins on these three insulators were 8, 11¹/₄, and 15¹/₂ inches respectively, so that the three arcing voltages, which were nearly equal, amounted to 9,225, 6,640, and 4,819 per inch of these distances. These figures make it plain that the arcing voltage for each insulator depends on the shortest distance over its surface and through the air, from wire to pin. It might be expected that the voltage in any case would arc equal distances over clean, dry insulator surface or through the air, and the experiments just named indicate that this view is approximately correct. The sparking distance through air between needle points, which is greater than that between smooth surfaces, is 5.85 inches with 70,000 volts, and 7.1 inches with 80,000 volts according to the report in Volume xix., A. I. E. E., p. 721. Comparing these distances with the shortest distances between wires and pins in the tests of insulators numbered 4, 5, and 7, which broke down at 73,800 to 74,700 volts when dry, it seems that a given voltage will arc somewhat further over clean, dry insulator surface than it will through air. This view finds support from the fact that only a part of each of the shortest distances between wire and pin was over insulator surface, the remainder being through air alone.

The fact that the dry part of the surface of an insulator and the air between its lower wet edge and the pin or cross-arm offer most of the resistance between the line wire and the pin and cross-arm is plainly brought out by the results of the tests above mentioned, in the cases of insulators numbered 4 and 7. While 73,800 volts were required to arc from line-wire to pin when the entire insulator was dry and clean, the arc was formed at only 53,400 volts during a moderate rain-storm, in the case of No. 4 insulator. With insulator No. 7 the arcing voltage was 74,700 when the entire surface was clean and dry, but the arc from wire to pin was started at 52,800 volts during a moderate rain. No. 5 insulator seems to present an erratic result, for when dry and clean the arc jumped from wire to pin at 74,700 volts, and yet during a moderate rain no arc was formed until a voltage of 70,400 was reached. For each of the seven insulators on which tests are reported as above, the voltage required to arc from line wire to pin was nearly or quite as great during a dry snow-storm as when the insulator surface was clean and dry. When the insulators were covered with wet snow their surface insulation broke down at voltages that were within ten per cent above or below the arcing voltages during a moderate rain in five cases. With two insulators the arcing voltages, when they were covered with wet snow, were only about sixty per cent of the voltages necessary to break down the surface insulation between wire and pin during a moderate rain.

When the outside surface of an insulator is wet, as during a moderate rain, it seems that the under surface of the insulator, and the distance through air from the lower wet edge of the insulator to the pin or cross-arm, make up most of the insulation that prevents arcing over from the wire to the pin or cross-arm. It further appears that it is useless to extend the distance across the dry under surface of the insulator indefinitely without a corresponding increase of the direct distance through air from the lower wet edge of the insulator to the wood of cross-arm or pin. Insulator No. 7 in the tests under consideration had a diameter at the lower edge of its outer petticoat of seven inches, and was mounted on a standard wooden pin. The diameter of this pin in the plane of the lower edge of the insulator was probably about 1¹/₄ inches, so that the radial distance through air from this edge to the pin must have been 2⁷/₈ inches approximately. During a moderate rain the surface insulation of this insulator broke down and an arc was formed from wire to pin with 52,800 volts. The sparking distance between needle points at 50,000 volts is 3.55 inches, according to Volume xix., A. I. E. E., p. 721, and must be shorter between smooth surfaces, such as the wire and pin in question, so that nearly all of the 52,800 volts in this case must have been required to jump the 2⁷/₈ inches of air, leaving very little to overcome the slight resistance of the wet outside surface of the insulator. On this insulator the surface distance from wire to pin was 15-¹/₂ inches, while the shortest breaking distance was only 7⁷/₈ inches, so that the distance across the dry under surface of the insulator must have been 15¹/₂ - (7⁷/₈ - 2⁷/₈) = 10¹/₂ inches approximately. It is evidently futile to put a path 10¹/₂ inches long across dry insulator surface in parallel with a path only 2⁷/₈ inches long in air, as an arc will certainly jump this shorter path long before one will be formed over the longer. The same line of reasoning applies to No. 3 insulator in this test, which had a diameter of 6³/₄ inches, a surface distance from wire to pin of 13 inches, and a minimum distance of 7¹/₄ inches, and whose surface insulation broke down at 48,600 volts during a moderate rain. The necessity of increasing the distance between the lower wet edges of insulators and the pins and cross-arm, as well as the distance across the dry under surfaces of insulators, led to the adoption of the so-called umbrella type for some high-voltage lines. In this type of insulator the main or outer petticoat is given a relatively great diameter, and instead of being bell-shaped is only moderately concave on its under side. With an insulator of this type mounted on a large, long pin, the lower edge of the umbrella-like petticoat may be far removed from the pin and cross-arm. Beneath the large petticoat of such insulators for high voltages there are usually one or more smaller petticoats or sleeves that run down the pin, and increase the distance between it and the lower edge of the largest petticoat.

Insulators on Transmission Lines.

The inner petticoat or sleeve that runs down over the pin and sometimes reaches nearly to the cross-arm, of course becomes wet on its outside surface and at its lower edge during a rain; but between this lower wet part of the inner petticoat, or sleeve, and the lower wet edge of the larger outside petticoat, there is a wide, dry strip of insulator surface. A result is that an arc over the surface of the outside petticoat can reach the wet edge of the sleeve only by crossing the strip of dry under surface or jumping through the air.

The same type of insulator is used on the 60,000-volt lines between Electra and San Francisco and between Colgate and Oakland, each insulator having an outer petticoat 11 inches in diameter and one inner petticoat or sleeve 6¹/₂ inches in diameter. This inner petticoat runs down the pin for a distance of 7¹/₂ inches below the outer petticoat. Slightly different pins are used for mounting the insulators on the two transmission lines just named, so that on the former the distance through air from the lower edge of the outer petticoat to the cross-arm is 11 inches, and on the latter the corresponding distance is 11¹/₂ inches. On the Electra line the lower edge of the inner petticoat of each insulator is about 3¹/₂ inches, and on the Colgate line about 4 inches above the cross-arm.

Insulators on Transmission Lines.

The CaÑon Ferry line is carried on insulators each of which has three short petticoats and a long separate sleeve that runs down over the pin to within 1¹/₂ inches of the cross-arm. This sleeve makes contact with its insulator near the pin hole. The outside petticoat of each insulator on this line is 7³/₄ inches above the cross-arm and 6¹/₄ inches above the lower end of the sleeve. Both the main insulator and the sleeve, in this case, are of glass.

White porcelain insulators are used to support the 50,000-volt Shawinigan line, and are of a recent design. Each of these insulators has three petticoats ranged about a central stem so that their lower edges are 4¹/₂ inches, 9 inches, and 13 inches respectively, below the top. The highest petticoat is 10 inches, the intermediate 9³/₄ inches, and the lowest 4¹/₄ inches in diameter. The height of this insulator is 13 inches, compared with 11¹/₄ inches for those used on the Electra and Colgate lines and 12 inches for the combined insulator and sleeve used on the CaÑon Ferry line. When mounted on its pin, this insulator on the Shawinigan line holds its wire 16¹/₄ inches above the cross-arm, compared with a corresponding distance of 14¹/₂ inches on the Electra, 15 inches on the Colgate, and 13¹/₂ inches on the CaÑon Ferry line. The two upper petticoats on each of these insulators are much less concave than the lowest one, and the edges of all three stand respectively 11³/₄, 7¹/₄, and 3¹/₄ inches above the cross-arm. From the edge of the top to the edge of the bottom petticoat the direct distance is 8¹/₂ inches.

Of the three transmission lines above named that operate at 50,000 to 60,000 volts, that between Shawinigan Falls and Montreal leads as to distances between the line wire and insulator petticoats and the cross-arm. On the Santa Ana line, where the voltage is 33,000, the insulator is of a more ordinary type, being of porcelain, 6³/₄ inches in diameter, 4⁷/₈ inches high, and having the lower edges of its three petticoats in the same plane. Each of these insulators holds its wire 8⁵/₈ inches above the cross-arm, and has all of its petticoats 3¹/₂ inches above the cross-arm. Unlike the three insulators just described, which are mounted on wooden pins, this Santa Ana insulator has a pin with an iron core, wooden threads, and porcelain base. This base extends up from the cross-arm a distance of 3¹/₈ inches, and the wooden sleeve, in which the threads for the insulator are cut, runs down over the central bolt of the pin to the top of the porcelain base, which is ⁵/₈-inch below the petticoats.

The 30,000-volt lines from Spier Falls are carried 10³/₄ inches above their cross-arms by triple petticoat porcelain insulators. Each of these insulators is 8¹/₂ inches in diameter, 6³/₄ inches high, and is built up of three parts cemented together. A malleable-iron pin cemented into each insulator with pure Portland cement carries the outside petticoat 7¹/₂ inches and its lowest petticoat 4¹/₄ inches above the cross-arm. When the voltage on the Spier Falls lines was raised from about 13,000 to 30,000, the circuits being carried in part by one-piece porcelain insulators, a number of these insulators were punctured at the higher pressures, and some cross-arms and poles were burned as a result. No failures resulted on those parts of these lines where the three-part insulators were in use. The second pole line between Niagara Falls and Buffalo was designed to carry circuits at 22,000 volts, or twice that for which the first line was built. Porcelain insulators were employed on both of these lines, but while the 11,000-volt line was carried on three-petticoat insulators, each with a diameter of 7 inches and a height of 5¹/₂ inches, the 22,000-volt line was mounted on insulators each 7¹/₂ inches in diameter and 7 inches high, with only two petticoats. The older insulator has its petticoats 2 inches above the cross-arm, and the lower petticoat of the new insulator is 3 inches above the arm. These two insulators illustrate the tendency to lengthen out along the insulator axis as the voltage of the circuits to be carried increases.

For future work at still higher voltages, the advantage as to both first cost and insulating qualities seems to lie with insulators that are very long in an axial direction, and which have their petticoats arranged one below the other and all of about the same diameter, rather than with insulators of the umbrella type, like those on the Electra and Colgate lines.