American Practice in High Tension Line Construction

AMERICAN PRACTICE IN HIGH TENSION LINE CONSTRUCTION AND OPERATION. (1)

BY F. A, C, PERRINE, D, Sc.

CHARACTERISTIC of American practice is that it tends toward standards, not only in the matter of the sizes of units, speeds and manufacturing appearance, but also in the methods of producing results and in the types of engineering. While it may be true that this tendency was originally based upon a desire for cheap manufacture and interchangeability of parts, at the same time it must be understood that the present elaboration of this policy is somewhat due to the fact that in so large a country the ideas of the best men cannot be directly applied except as they may be adopted for standards. No one section of the country produces the best men necessarily, nor does any one group of engineers dominate our practice. On the contrary, the meetings of our engineering societies have taken the character of sittings of committees, where are presented many plans, and where all plans are carefully discussed and sifted. From those presented the best is chosen and becomes the standard.

Accepting these results as the standard does not imply that there is general in this country a spirit of copying or of servile imitation among the engineers. On the contrary, we feel that the result of the attitude so prevalent in American engineering at the present time, of establishing standards, has introduced a wise spirit of conservatism, and has thrown the burden of proof upon each one presenting a new idea, At the same time it has resulted in raising the character of the average engineering work throughout the country, until today good American engineering can be found, not only in the great spectacular plants near enough to the large centers of progress to have the personal attention of the most experienced engineers, but in consequence of this system of practice an equally good type of engineering can be found in the plants in the out-of-the-way deserts or mountain regions, where the local engineer of good capacity, knowing his conditions thoroughly, has relied upon the standards established by his fellows in those particulars where his own experience has been limited, and in consequence a plant is produced, not only more perfectly adapted to the particular circumstances of its surroundings, but in all details more thoroughly satisfactory than could have been designed under any other system. Our rule is that invariably one should adhere to well-established practice and introduce such modifications as are made necessary by the local conditions. This does not limit the full employment of the energies and brains of the local engineer, since, without a special consideration of outside details, there are always in every transmission plant particular circumstances which tax the ingenuity of the best. That this is the general method of American practice will be seen by anyone who consults the report of the standardizing committee of the American Institute of Electrical Engineers. The report covers, not only units, standard methods of testing and details of manufacture, but also procedure, both outdoor and in, for all types of plants, and this report in itself has resulted in a certain similarity of type where problems to be solved are similar.

The work of the transmission engineer lies in fields so essentially dissimilar that, even in spite of this general tendency, it may be difficult at first view to ascertain what is the American practice in work of this class. On closer examination one finds, however, this work falling into natural groups dependent on the length of transmission and the voltage employed, though what has been done has been materially modified by the date of erection, since during the past ten years modifications in the arts have been necessarily reflected in the types of construction.

The general groups have been somewhat decided by the manufacturers of machinery, who have presented as preferable certain available voltages. Above 2400 volts, where transmission proper really begins, the first voltage now commonly employed is 6600, which figure has been established as standard by the needs of the lighting plants in the great cities, and has been adopted by the transmission companies in place of either a higher or lower voltage, mainly because it is a standard. For this voltage direct generation at high pressure is almost invariably used. The next higher voltage now commonly employed, and now practically the first one for which step-up transformers are used, is 15,000. During the past few years this has taken the place of transmissions at 10,000, 12,000 and 13,000, and it is today the established voltage for high-tension electric railways, the general reason for its establishment as a standard being that this voltage is not more difficult to handle, as regards insulation or switching, than the three last-mentioned lower voltages, and, furthermore, that where the lower voltages have been previously established, the sphere of operation of the transmission plant has been found to be rather too much limited. There are in the Rocky Mountain region and west a great number of the older plants operating at 10,000 volts, and whenever direct high-voltage generation has been attempted, voltages of from 12,000 to 13,000 volts are used; but, at the same time, the majority of the plants which have used the lower pressures in the past today have circuits with special transformers operating at the higher figure. The next step is to 25,000 volts, which is the highest figure reached without special study of insulators, switches and lightning arresters. This voltage has been successfully handled without serious trouble during the past six years. A voltage of 35,000 is employed in a number of plants built about five years ago, and at this figure the special difficulties due to line capacity, insulator size, erratic lightning-arrester effects and switching begin to make themselves seriously felt. Above 33,000 volts the standard voltage is called 60,000, although in all plants that have heretofore been established to operate at this pressure there have been installed transformers arranged for connection to various voltages of from 40,000 to 60,000 volts, and the majority of these plants are today operating at about 50,000 volts, some of them being unable to operate at the highest pressure on account of the character of line insulators originally installed. In the choice of voltage for any transmission it is considered the best practice to establish it at the rate of 1000 volts per mile, provided the length of transmission be not above sixty miles, since above 60,000 volts no commercial work has been regularly attempted. In the table recently presented by the transmission committee of the American Institute of Electrical Engineers, the highest average voltage per mile for any one class in their report is 840; but in examining this table it must be remembered that their correspondents have reported the total length of line in service, so that, if a plant operated two lines fifteen miles each in length at 1500 volts, their table would indicate an operation of 500 volts per mile, although for each line the transmission was at 1000 volts per mile.

In the generation of power the revolving armature machine has almost disappeared from the new plants, and revolving field generators have become so settled in type that those produced by different manufacturers are hardly distinguishable by the casual observer. For the low-head plants using turbine wheels it is necessary to provide for a 50 per cent. increase of speed, and in the high-head plants, where impulse wheels are employed, a strength sufficient to withstand a speed increase of 100 per cent. must be allowed to provide against damage from over-speeding should the power be thrown off and the water continue to flow. The machine fulfilling these conditions and practically adopted by all the manufacturers is characteristically a revolving field machine, with the poles keyed to a cast-steel spider, the field windings being of copper strip wound upon edge, the armature being constructed of a cast-iron box girder supporting the stationary armature laminations. Almost the only departure from this type of construction for power transmission work is found in the balanced type of inductor machine, where the field is magnetized by a central stationary field coil wound with copper strip, the armature in two halves, symmetrically arranged around the central core, being of laminations supported either by cast-iron rings connected together by cold-rolled steel bars or supported by a steel shell to which the armature laminations are keyed.

Various station voltages have been employed, but, where direct generation at 6600 to 12,000 volts has not been resorted to, the practice is setting more and more to the use of about 2300 volts, this being chosen because the lower voltages require large extra station copper and the higher voltages are felt to introduce unnecessary station difficulties of insulation and switching. For switching, the present type of 2300-volt oil switch has been so well developed, by reason of the great number of plants operating at this pressure, that for handling a particular amount of energy it is both cheaper and better than any 500-volt switch on the market.

For plants operating at less than 25,000 volts the step-up transformers in use are about equally divided between the water-cooled, oil-filled types and air-blast types. Where a good supply of water is to be readily obtained, the oil-filled transformers have generally been given preference, as they can be more readily adjusted for a varying flow of water at different loads. The question of the relative fire risk from the two types has been extensively discussed, and it can hardly be said that any very definite conclusion has been finally reached, though the weight of opinion seems by far to be that the fire risk is at least not increased by the use of the oil-filled transformers, and the actual risk in either type seems to be a matter largely of installation. It is perfectly true that there have been some very serious fires, resulting in the complete destruction of power plants, where oil-filled transformers have been used, but in each case the fire has started outside of the transformers, though they themselves, by reason of being installed without reference to safety in case of fire, have furnished fuel which has augmented the conflagration. Today the conditions of installation for safety are better understood, and it now only remains to be decided whether, in the case of a fire actually arising, the oil shall be run out and the transformer cases filled with water, or the whole transformer protected either by running in an excessive amount of water through their cooling coils, or by so installing them that the transformers may temporarily be submerged to within a few inches of their tops. Actual protection of transformers by running water through their cooling coils has been found to be effective in at least one serious fire.

For high-tension switching use has been made of a long arc. breaker between carbon terminals and long-inclosed fuse—a fuse drawn through a tube filled with a fine, non-conducting powder— and of oil switches. The first two types, while interrupting the circuit well, draw an arc of excessive length and produce a surging which may result in an increased potential of at least as much as 50 per cent. In consequence, these types are rapidly disappearing, except in plants operating at 15,000 volts and below, where the carbon break is preferred to the inclosed fuse, though it is common to install the two in series, allowing the fuse to operate as a safety device, but not for the purpose of switching. The type of fuse where wire is drawn through a tube filled with powder was found to operate successfully up to 40,000 volts and without serious surging on the circuit, but the powder being blown out with great force and scattered over the entire station is in consequence not allowable. The oil switches mainly employed are those with the vertical break and those with the horizontal break. The vertical break switch has the advantage that the amount of oil contained in the oil tank is relatively small, and will add to possible conflagration only a slight amount of fuel. This switch is often found on severe short circuits to blow all the oil out of the tank unless the tank is built very strongly, when it becomes necessary to insulate the plunger from the tank as it enters the switch. The horizontal break switch, while containing a large amount of oil, will for the same length of break handle about 25 per cent. more energy at any definite potential. This switch can successfully be used at 60,000 volts, and up to the present time has not been found to blow the oil from the tank. These two types of oil switch are the standard to-day, no distinct preference being given to the horizontal switch, though the writer believes that in the future this type will be used as a standard for the highest potentials.

Transmission with two-phase connection of circuits, whether using three or four wires, has for voltages above 6600 given place entirely to transmission with a three-phase connection, though three-phase transmission with two-phase distribution described by Mr. Scott at the International Congress of 1893 is very extensively employed.

The relative merits of the delta and star connection of the lines to the transformers is still somewhat in dispute, so much so that in plants of the highest voltages, where several voltages are provided, certain of the lower voltages are obtained by delta connections to the transformers, while the higher voltages are to be obtained by a star connection. In general it may be stated that up to 25,000 volts the delta connection is generally preferred, principally because with this connection a ground upon one line does not necessarily result in a short circuit, and, furthermore, the service is not necessarily interrupted in the case of the failure of a single transformer. At voltages higher than 25,000 volts the transformers for delta connection become more difficult to build and insulate. Furthermore, a single ground anywhere produces disturbances of a serious character, and inconsequence the star connection with the grounded neutral is employed, advantage being taken of the fact that a grounded neutral aids in the distribution of unbalanced loads, and furthermore, the rise of pressure which may occur from line discharge at the time of an open circuit or a short circuit is not so likely to produce serious results.

For the distribution of current through the low-tension mains it is generally the custom to transform to 2300 volts, two-phase unless either the load is mainly one of motors, or unless there are important motors of considerable size to be supplied at a distance of half a mile or more from the substation. In such cases three-phase star-connected four-wire distribution is employed, allowing the connection of distributing devices either to a 2300-volt circuit between lines and the neutral wire, or a connection to a 4000-volt delta circuit for balanced loads. This combination of circuits is found to be extremely useful where a mixed load is to be supplied at varying distances.

The high-tension lines themselves are preferably run over private right of way. Railroad rights of way were at first highly prized on account of the entire absence of trees and disturbing structures, and furthermore, on account of the fact that inspection and repairs are most easily provided for; but experience with such lines has proven that, for transmission at even so low a tension as 15,000 volts, the interference with insulation by the smoke from the locomotives, which covers the insulators, more than counterbalances all the advantages, and today such rights of way are more commonly shunned than sought. Where railroad locomotive smoke combined with sea fog is encountercd, it becomes absolutely necessary to clean each insulator at frequent periods, even though the voltage of transmission be not more than 5000 or 10,000. Along the country road this difficulty is not apparent, but in some localities farm structures and trees interfere with the transmission, so that in general it may be said that a private right of way that the transmission company can absolutely control is much to be preferred.

In the most recent types of construction the height of pole is limited as much as possible. Where there may be some increased security from malicious disturbance in the use of high poles and decrease of line capacity may be expected, these advantages are only obtained at the expense of stability and at an increased cost. A pole thirty-five feet high set five feet in the ground .permits the safe installation of either a single three-phase line with a spread of as much as six feet by supporting one insulator on the top of the pole and the other two on the ends of a long cross-arm, or it may be used to support two three-phase circuits on opposite sides of the pole with a spread between wires of three feet by the use of two cross-arms, and at the same time such a pole permits the safe installation of telephone or other signaling circuits on brackets or cross-arms at a safe distance below the power lines. These poles should not be less than eight inches in diameter at the top and not less than twelve inches in diameter at the ground line. Variations from these dimensions may be considered as being due to special considerations based upon the location of the lines or arrangement of the circuits. It is true that such a standard pole may only be arrived at after a consideration of the wind stresses on the particular lines taken in conjunction with the spacing of the poles, but as the maximum pole spacing on transmission lines is about 135 feet, and at average wind velocities the pole dimensions may be considered safe. Extra strength required by variations of wind stress, either due to an increase in the number of wires or to a necessity of allowance for sleet, is more commonly taken care of by shortening the spans than by an increase in the size of the pole. In some cases where severe sleet conditions are to be encountered and the wires changed, it is the practice to install these poles at not more than fifty feet apart.

The material used for poles depends largely on the locality. In the southeastern states chestnut is the favorite wood; along the Canadian border and along the Rocky Mountain regions cedar is employed, while square-sawn redwood is used almost exclusively on the Pacific Coast. With increase in voltages and consequent increased trouble from insulators, a demand has arisen for a pole-line construction that will permit a decrease in the number of insulators and allow an increase in the size of each. This has been accomplished by the use of galvanized iron towers not less than forty feet from the ground line to the wires and spaced about 500 feet apart. One plant in Mexico has recently successfully installed this method of construction. A second in the same country has contracted for its material and a number of plants in the United States are contemplating its use. The question of the life of wooden poles depends not only upon the character of the wood and its condition when cut, but also upon the local conditions of atmosphere and soil. In some places the poles which are available have no longer life than about five years, and, in the extreme, wooden poles cannot be greatly depended upon for a longer period than fifteen years, though the redwood poles installed along the lines of the transcontinental railroads west of the Rocky Mountains have in many instances given a life up to thirty-five years and are still found to be in good condition; but these poles are set into a soil strongly impregnated with alkali in a country where rains are few and the air generally dry. Nothing is known as yet of the life of the galvanized iron towers except from the windmill practice, where towers which have been galvanized after all punching and machining is done are found to be in good condition after a period of ten to fifteen years.

The cross-arms in use are almost invariably made of pine without treatment other than painting. These arms are let into the pole from one to two inches, being held by bolts through the pole and arm, and when long are additionally supported by braces. Even with higher poles wood arms are used, the general feeling being that there is less probability of the circuit being completely disabled should an insulator break and the line fall, if it falls upon a wooden rather than a steel arm. At the same time au experiment in the use of wooden braces has not been found to result in any advantage. In consequence, flat, galvanized-iron braces, established a number of years ago as standard by the telegraph and telephone companies, are now almost universally employed in the construction of transmission lines. With increase in spans and voltages the insulators are increasing in size. This condition will probably in the future demand a strength of arm greater than can be obtained by the use of wood. This problem, however, has not as yet obtained a definite solution.

For plants operating below 25,000 volts much use has been made of glass as a material for insulators. Glass has been for many years the standard insulator material in American telegraph and telephone practice, and in spite of many experiments that have been tried with porcelain, it is still considered the best and cheapest material for this service. However, in transmission work one of the great advantages claimed for glass in telephone and telegraph practice disappears. The engineers of these companies claim that it is important to provide against dark, narrow spaces within the insulators, on account of the fact that they form the homes of insects. The transparency of the glass largely obviates this difficulty. Where large insulators are used, such as are employed by transmission companies, the spaces within the insulators are well lighted from below, and the transparency of the material is not important. Glass is comparatively fragile, and for transmission work it has nothing to recommend it except low first cost and cheap inspection; these, to be sure, are very often overpowering advantages when the voltage is low enough, so that the particular form of insulator used gives a large factor of safety, and in consequence up to 15,000 volts glass insulators are generally preferred unless there are special climatic conditions which render them liable to fracture. Many series of tests have shown conclusively that the porcelain insulator has a greater mechanical strength, is less liable to surface leakage, has a safe dielectric strength, and in addition that it is exceedingly difficult to so break the head of a porcelain insulator as to allow the wire to fall away from it. The one disadvantage of porcelain is that there is an uncertainty as to its solidity, and that itis only possible to ascertain its solidity by most careful high-voltage tests. The question of the form of high-voltage insulator is as yet in high dispute, operating engineers being inclined to a design where the petticoats are very long and comparatively close together, so that great creeping distance be given over the surface of the insulator, between line and line and between line and pin, comparatively little regard being placed on the flashing distance. Engineers of the manufacturing companies, however, incline toward one of a much more open type of large diameter and with few petticoats. This latter form undoubtedly gives the greatest sparking distance, has the least dark spaces within it and is more readily cleaned by rain storms. It is also important that such an insulator be operated at high voltage without noise, and, as there is a definite loss of energy whenever the insulators on a line are noisy, it may be safely predicted that the open type of insulator is to be the one that will be in the future considered as the standard.

While insulator size may be largely determined by form, at the same time we may in general note that up to 10,000 volts, insulators, whether of glass or porcelain, have a minimum diameter of about five inches. A seven-inch insulator can successfully be used on voltages as high as 25,000, a thirteen-inch insulator is sufficient up to 40,000 volts, while at 60,000 volts it does not seem safe to install insulators having less diameter at the top than fourteen inches, A greater size would unquestionably invariably be used for these high voltages if the problems of the manufacture of porcelain and support of the insulator were altogether solved.

For support of insulators on cross arms it is necessary to provide that the lowest petticoat be raised above the cross-arm as much as the radius of the insulator, and, as the strain comes on the extreme top of the pin, it is obviously difficult to successfully support the largest size of insulators by means of the common pin in cross-arm construction. By using carefully selected woods this has been successfully accomplished for insulators up to November, eleven inches in diameter, but at 40,000 volts in bad weather such insulators carry enough current over the city to char a wooden pin. Accordingly practice has settled to the use of iron pins in plants operating above 25,000 volts. At this voltage and below the wooden pin can be successfully used and indeed forms a certain protection to the line by reason of the fact that the pin itself is a safe insulator, and is only in danger of being burned where the insulator is punctured. Above this voltage, however, only metal pins can be employed, not only on account of the large size of the insulator, but also on account of the fact that there is much burning of wooden pins. The manner in which these pins are burned has attracted considerable attention, having presented some problems which are exceedingly interesting. There is no doubt but that the effect is due to leakage over the surface of the insulator, but it is extremely interesting to note that in some cases the pin is actually charred, whereas in other cases there is an apparent disassociation of something in the wood, and peculiar salts are left behind, either reduced from the atmosphere or from the material of the wood itself. This matter was discussed by C. C. Chesney in a paper read before the American Institute of Electrical Engineers.

The materials that may be used for pins are locust and eucalyptus. The latter wood is decidedly preferred in the plants west of the Rocky Mountain region and where it is readily available, as the wood has been found to be as strong as hickory, dense and readily handled when thoroughly seasoned and dried. For the largest sized pins, however, as has already been said, no wood is entirely satisfactory, and in consequence use is made of malleable cast iron or cast steel.

As regards conducting material, it may, of course, be said that the only materials at present available are copper and aluminum. For a number of years there has been a discussion of the possible use of iron for short lines on high-potential plants, since the smallest copper wire that may successfully be strung is unnecessarily large under such circumstances. This procedure, however, has not obtained the approval of any of our electrical engineers. The copper wire is invariably uninsulated in high-tension work, since it is correctly believed that no insulation is a true protection, and the frank nakedness of the bare wire is a warning, and in consequence a safeguard to those who are compelled to work near the line.

Copper is used either soft, hard-drawn or stranded. For transmission work, where the wires are smaller than three-tenths inch in diameter, use is not made of soft-drawn wire, and it may be stated that the standard in American practice is to use the soft-drawn wire only for large, low-potential circuits where the small change in conductivity due to the hard drawing is an important factor. Up to three-tenths-inch hard-drawn copper may be considered standard. Between three-tenths and four-tenths-inch diameter the practice is evenly divided between solid hard-drawn wire and strand. Larger than four-tenths-inch strand is almost invariably employed. Some use has been made of solid aluminum, but, as the material must be handled with great care, it has been found generally to be the better practice to employ aluminum strand, which is more readily installed and more reliable after installed.

Preference between aluminum and copper is almost entirely a matter of price for transmission lines. It is true that aluminum is stronger in reference to its weight for different conductivities than copper, but at the same time it is materially larger, and the resultant transverse wind stress on the line is greater. For short lines, delivering a small amount of power at voltages of 40,000 or above, aluminum is decidedly to be preferred, since it is found that at these voltages a wire less than one-quarter inch in diameter will discharge through the air, and this discharge may result in a considerable loss of energy. Accordingly, it is not possible at these voltages to successfully use wires less than three-tenths inch in diameter, no matter what the amount of energy or the distance. Accordingly, where the amount of energy and the distance may result in the loss not being the determining factor, aluminum is much preferable for the reason that at a definite size it is materially cheaper than copper. Where salt sea fog is to be encountered, both aluminum and copper are acted upon. The action on aluminum is greater than the action on copper, and in consequence copper must necessarily be used. Where such conditions are not encountered aluminum is an entirely safe material, provided it is not exposed to the elements in contact with any other metal. The joints, therefore, must either be made of aluminum of the same quality as the wire, or the joints must be carefully insulated so that no moisture will penetrate. Aluminum must be strung with careful reference to the temperature at the time of erection, since its coefficient of expansion is very large, about three times the coefficient for copper, and experience in the erection of copper lines will result in an unsafe aluminum line. Careful tables have been prepared as to temperature, span and sag, and when these tables are followed no apprehension need be felt as to the safety of the line.

The most difficult problem at present encountered in the construction of high-tension transmission lines is that presented by the lightning arresters. For voltages up to 25,000, the non-arcing types of lightning arresters, either with or without series resistances, may be successfully used. Above this voltage and where large amounts of energy are available, these arresters are found to be short-lived, and up to the present time no thoroughly satisfactory arrester has been presented, which does not, when interrupting the ground circuit after a discharge, injure the insulation of the line and transformers. The horn form of lightning arrester developed in Germany has been found to operate with invariable success so far as the lightning arrester itself is concerned, but, as it is interrupting the ground circuit, it draws a large arc, and oscillations are produced on the line, which in many cases have been found to have more serious result than the discharge they were installed to remove. Condensers in parallel with the lightning arresters and ingenious arrangements of arresters and resistances have been used with some success, but none of these plans may be considered to be entirely satisfactory for the highest potentials operated from the largest generating plants.

In the operation of such lines every effort is made toward maintaining continuity of service. Such lines are carefully patrolled, even when it becomes necessary to build a special runway for the patrolman, and it is remarkable with what certainty these experienced men can predict the hours of life of a failing insulator and provide for voluntary interruption of the service in time to remove the imperfection. Duplicate lines for long-distance work is in an invariable necessity, though by far the best protection that can be offered for service is the supply of current from different power stations over lines following different routes. The present tendency is toward the consolidation of plants, not only for the purpose of decreasing the general operating expense, but more particularly providing continuity in the case of the most serious accidents. No difficulty is experienced in operating in parallel plants widely separated, and where a number of plants are feeding into the same network, two certain plants are assigned the regulation of the entire system, others feeding the circuit being allowed to operate their machinery at full load continuously. The line capacity offers the most serious problem in determining regulation where the loads vary widely, but this quality becomes important only for great variations of load, which, as the plants are increasing in size and load, are disappearing. When proper care has been given to the installations of the lines and where duplicate lines and plants are provided for, care in operation and patrol of the lines has resulted in success both from the engineering and financial standpoint.

A paper presented before section D (electric power transmission), at the International Electric Congress, St. louis, September 13, 1904.