Suspension Insulators

[Trade Journal]

Publication: Electrical Review

New York, NY, United States
vol. 78, no. 4, p. 135-138, col. 1-2

Suspension Type of Transmission Line Insulators


Eleventh of a Series of Articles Giving a Review of the Electric Power Situation—Grading Suspension Insulators—Static Shields and Arcing Horns—Design and Construction of Insulator Units




The suspension type of insulator was first used when line pressures of 100,000 volts were introduced. Since 1907 it has superseded the pin type of insulator for pressures above 60,000 volts and even less. With the suspension type of insulator the conductor is hung below the point of support (which is usually grounded) at the end of a string of insulator units connected to one another by metal links. The potential difference which will cause a flashover or breakdown on such a series of insulators will not be in direct proportion to the number of insulators in the string. This is due to the unequal distribution of the potential differences, which is again a question of relative capacities. The design of the individual units may appear to be good, and yet a string of such insulators, if they are not specially designed to fulfill certain requirements, may give surprisingly unsatisfactory results. A factor of importance is the ratio of the mutual capacity to the capacity to ground, which determines the potential distribution.

The distribution of the potential drop on a string of suspension insulators is easily calculated if certain assumptions are made with a view to simplifying the work. If symbols are used to denote the various voltages and currents, the formulas become somewhat complicated in appearance although actually simple in application. A clear explanation of the voltage distribution on the string of suspension units has been given by F. W. Peek, Jr.

The fact that the ratio of the mutual capacity to the capacity between the suspension link and ground is an important factor in determining the distribution of potential over the insulator string suggests the importance not only of the shape and surface area of the metal fixtures on the individual insulators but also the length of the connecting link and the spacing between insulators.




In this country it is common practice to use the same size and type of disk throughout a string of suspension units, but this necessarily leads to a greater potential difference across those near to the high-tension line than across those near to the (grounded) point of suspension. In connection with tests on strings of insulators, such as those recently conducted by Profs. Harris J. Ryan and Henry H. Henline at Leland Stanford University, the expression "voltage-duty ratio" is used to indicate the ratio between the voltages across two unit insulators in a string of suspension insulators. With a string built up of 15 standard 10-in. suspension discs a voltage-duty ratio as high as seven is possible, and this means that the unit next to the high-voltage conductor is normally subjected to seven times as great a pressure as the unit next to the point of attachment on the tower crossarm. This is unsatisfactory, and a suspension type insulator built up in this manner would be undesirable on a practical transmission line.


This illustration has not been processed yet.


The grading of insulators so as to provide for a more uniform distribution of potential throughout the string has been given much more attention in Europe than in this country. In this country we have tried and are proposing to use a metal shield between the lowest unit in the string and the high-tension conductor, and this, by preventing the concentration of the dielectric flux at the point of attachment, has the effect of improving the voltage-duty ratio. A static shield of this nature will usually take the form of a flat disk with well-rounded edges and having a diameter appreciably larger than that of the porcelain disk. An elliptical shield with its major axis in the direction of the line is also effective. Besides serving to reduce the potential difference across the units near the conductor the static shield—in common with arcing horns and rings—also tends to prevent corona formation over the porcelain surface at the point where the metal caps or links are connected to the porcelain. On a string of ten disks, the provision of a static shield may reduce the potential drop across the first insulator unit by as much as 25 to 40%.


This illustration has not been processed yet.


Although maximum voltage-duty ratio as high as seven has been mentioned as possible on a string of 15 units, this ratio would be about 2.5 on a string of ten units, and this could be reduced to below 1.5 by the provision of a static shield of the type above described. A still further improvement could be made by grading the units, i. e., using two or more different sizes or designs in the same string.




The number of units in a string will depend upon the line voltage and the size or make of unit. Lines operating at 66,000 volts are using strings with any number of units ranging from three to eight. The higher voltage lines will naturally require more units, but owing to the unequal distribution of potential drop along the series the addition of extra units to a long string will have little effect in reducing the potential difference across the first unit. Additional units in excess of the number which appears requisite for the duty demanded of them increase the safety factor of the string because they will take their share of the voltage in the event of the breakdown of other units. On 110,000-volt lines the number of units per string ranges from five to nine. On 130,000-volt lines nine or ten units are used, but as many as 12 units in a series are used at dead-ends on tension towers.

From what has been said concerning the voltage distribution over a string of suspension units it will be understood that when the mutual capacity is small relative to the ground capacity there soon comes a point beyond which it is useless to put more insulators in the string. Short strings of well-designed and properly spaced units will often be more effective and less costly than a longer series of insulators of which the single units may have excellent insulating properties, but may not have been' specially designed for the particular requirements. It is sometimes possible to increase the arc-over voltage of a string of insulators by reducing the distance between consecutive units, a fact that it is difficult to understand unless the importance of the proper capacity distribution has been realized. A well-arranged string of properly designed units, not necessarily similar in shape and size, is less liable to damage by lightning and kindred phenomena than a series of insulators of high voltage-duty ratio.

It is well to bear in mind that an insulator may and does behave differently when subjected to high-frequency charges such as those induced, by lightning disturbances than when the difference of potential is applied at the normal frequency of the transmission line. The distribution of potential. among a number of condensers in series should be independent of frequency, but a high-tension insulator consists of many capacities in series with resistance such as the leakage paths through the body of the material and over the surfaces and the distribution of the total potential drop across a condenser and resistance in series, even if the latter may be considered non-inductive, is not the same at high as at low frequencies. This is one of the reasons why insulators that will flash over rather than puncture on tests conducted at the ordinary line frequency will sometimes fail through puncture during atmospheric electrical disturbances.

Apart from the great advantages from the point of view of insulation which are obtained by suspending the conductor from a string of insulators connected in series, this arrangement, now generally adopted for pressures above 50,000 volts, has . the further advantage that the conductor is less liable to be affected by lightning disturbances, since at every point of support the wire is hung below the attachment to the supporting tower, which in almost every instance is a well-grounded steel structure. Another advantage is the comparative flexibility of the attachment, which very considerably diminishes the possibility of crystallization of the conductor material, such as is liable to occur when the wire is rigidly attached to the pin type of insulator. This effect is more noticeable with aluminum than with copper.




A great deal of attention has recently been devoted to the improvement of the suspension type of insulator. The parts of the early designs were cemented solidly together, with the result that the differential expansion of porcelain and metal caused cracking and consequent destruction of the porcelain after being in use for only a few years. That experts in insulator design realize the effects of expansion and other causes leading to rapid deterioration of porcelain insulators is evidenced by the experimental work that has been carried on, and the number of papers published relating to this subject during the last three or four years.


This illustration has not been processed yet.


It is proposed to discuss the question of deterioration and failure of insulators in another article. Unquestionably much trouble has arisen through unequal expansion under temperature variations of porcelain, cement and hardware in the usual type of unit in which the metal cap and link are cemented to the porcelain. When trouble was first encountered due to this cause, felt or paper gaskets were placed between the metal cap and the porcelain. In more recent designs these gaskets were removed from the finished product after the cement had set, to avoid trouble due to ionization; but various devices of the nature of elastic washers or packing between metal and porcelain are now used with a view to overcoming this trouble.

Cap and pin type suspension insulators of Westinghouse manufacture are set under steam for the purpose of accelerating the setting action of the cement and of having the setting action take place with the materials expanded. Cement which has passed the required test is sifted through a 1, 16-in. mesh sieve and thoroughly mixed, giving a mortar semi-fluid in consistency. The cap is placed, clevis down, in a rack made for this purpose. The cushioning material is placed in position and enough cement poured into the cap so that when the insulator is pressed into position some of the cement flows over the edge of the cap. The cushioning material for the pin is then placed in position, Cement added and the pin inserted. The pin is worked to drive off the trapped air, settle the cement and obtain the proper alignment. The assembled insulator is then picked up by the cap and transferred to the steam chamber, being placed in the inverted position on a rack. The air in the steam chamber while the insulators are being placed is kept moist by throwing water on the floor. After filling, the doors of the chamber are closed and the steam turned on.

In the Jeffery-Dewitt suspension insulator unit the troubles due to unequal expansion of the metal cap cemented to the porcelain are overcome by employing a flexible spider fastened to the porcelain by means of an alloy having sensibly the same coefficient of cubical expansion as the porcelain, thereby eliminating the temperature effects in the operation of the insulator. The makers of this insulator have no hesitation in submitting it to the following test: Immerse in boiling water for five minutes; remove and plunge immediately into iced water for five minutes; then back into the boiling water, and so on, repeating the process as often as desired.




In this, as in other modern designs of line insulator, there is seen a tendency to thicken the porcelain. Improved methods of manufacture