GODDARD: High Voltage Porcelain Insulator Manufacture

[Trade Journal]

Publication: Electrical Review

New York, NY, United States
vol. 52, no. 6, p. 227-233, col. 1-3






The transmission of power by means of electric currents of greater or less voltage began with the introduction of the electric telegraph, and almost the first problem encountered was line insulation. It is of interest in this discussion to note that the first insulators successfully used were of porcelain. From that time to the present the progress has been rapid, the voltage having risen from two or three volts of the telegraph, to a present prospective line voltage of 150,000, and from the lowest to the highest it has always been recognized that best construction involved the use of porcelain insulators.

Fig. 6 shows sectional drawing of a high tension insulator made of several shells nested together and arranged to be fastened together by means of thin joints of Portland cement. This is the conventional design and is used for all voltages, though the size, number of shells, and shape may vary with the voltage. In general, American high voltage insulators are now made of several shells nested together, while European manufacturers still persist in turning complicated shapes from solid masses of dry clay.

The manufacture of porcelain is one of the oldest arts, but only recently has it in any sense been placed upon a scientific basis; In fact, potteries for the production of electrical porcelain exclusively, have not been In existence for more than ten years, and during that time It has been necessary to develop an entirely new system of handling pottery products.

Porcelain for electrical purposes is a mixture of ground flint or silicon dioxide and feldspar, or (K2O Al2O3SiO2) potassium aluminum silicate, raised to vitrifying temperature, that is, to a temperature sufficiently high to melt the feldspar and permit it to unite the particles of flint Into a perfectly homogeneous body of uniform electrical and mechanical strength. The production of electrical porcelain differs from the ordinary pottery product in that, in addition to presenting a symmetrical and flawless exterior, it must possess inherent electrical and mechanical strength.




Flint and feldspar occur in nature as rock which is reduced at the mine by grinding to a degree of fineness comparable to that of flour and of equal whiteness. The modern electrical porcelain potter mixes the proper proportions of flint and feldspar and again grinds the material in the presence of water in order to obtain intimate mixture. The mixture, or as it is dubbed in the factory, "clay," is separated from the excess water, immediately after leaving the grinding mills, by means of filter presses and afterward brought to uniform plasticity by means of kneading machinery.


Fig. 1. Plaster Model of 60,000-Volt Insulator Top.


Fig. 2. Conventional Power-Driven Potter's Wheel and Forming Tool.


The first step in the construction of an insulator is to build a model 16% larger than the required insulator, and from this model to make moulds of Plaster of Paris. Fig. 1 shows a plaster model of a 60,000 volt insulator top from which the mould in Fig. 2 was taken. Each shell of a multipart insulator is treated in this manner, the inside contour of the mould being that of the desired shell. Fig. 2 shows a conventional power driven potter's wheel upon the top of which is fixed the mould partly filled with clay. The whole device is rotated rapidly and a forming tool whose profile is that of the inside shape of the shell under consideration, is forced into the mass of clay. A few revolutions usually accomplish the desired end and the mould with its wet and plastic clay is then placed in a hot room of approximately 130 F. Within an hour the warm air and plaster mould have absorbed a large proportion of the water in the clay, and the embryo insulator may be removed from its mould and the rough surface, which originally rested against the mould. scraped smooth. The shell, as it may now properly be designated, is set aside for a period of ten days to two weeks in order that all water held by the clay be evaporated. If fired in this state the shell would come from the kiln hard, white and quite rough, of good dielectric strength, but with a surface which would gather and hold dirt and soot and, because of the extreme fineness of the irregularities, be practically beyond possibility of cleaning. For this reason, no other, are the insulators covered with a glassy coating of extreme smoothness, capable of being washed by the gentlest rains. The glazing process is very simple. Immediately before being placed in the kiln, the insulator shell, which in its dry state is about as strong as blackboard crayon, is dipped into a solution of clay and water and by virtue of its dryness absorbs a certain amount. On being heated the clay melts before the body vitrifies and so spreads very evenly over the entire surface. The presence of a small amount of iron In the clay is responsible for the brown color so generally used in this country although any color can be obtained by the use of the proper materials in the glaze. The firing process is entirely a matter of temperature and its complications are ail practical ones due to the large amount of material burned at one time and the difficulty of obtaining exact and uniform heat throughout the kiln.




The mechanical properties of porcelain are so unvarying that a sample test has always been deemed sufficient. On the contrary it is considered unsafe to permit any porcelain to go into service without testing each piece by means of a high voltage transformer. As will be explained later, the higher voltage insulators are made up of several shells each of which is subject to a high voltage test of 50,000 to 60,000 volts. Fig. 3 shows the connections between the high voltage transformer and insulator shells to be tested. Fig. 4 shows this method as practiced. The table is insulated from the floor by means of porcelain cones in order that high voltage may be applied to the wire imbedded in the table top and designed to make contact with the pans arranged for the reception of porcelain shells. The pans are filled with water and so form one terminal of the high voltage. The other terminal is formed by water poured into the shell, attachment between it and the high voltage over-head conductor being made by means of small chains. By some controlling device in the low voltage side of the transformer, the voltage applied to the shells of porcelain is kept very near to arcing over, in most cases being about 55,000 volts. For ordinary commercial testing the spark gap in air is usually employed when it is desired to ascertain the applied voltage. The number of failures is usually 2% or 3% of the number tested. When porcelain fails electrically the short powerful power arc through the very small puncture hole amounts to practically a short circuit upon the testing transformer and the excessive heat liberated in the porcelain often boils the water and destroys the shell. Once punctured, the porcelain is ruined and can by no method at present developed, be recovered.


Fig. 3. Connection Between High-Voltage Transformer and Insulator Shells to Be Tested.


The parts of multipart insulators are united by means of pure Portland cement mixed with water only. After the cement has obtained initial set the insulators are subjected to an assembled electrical test of double line potential for a period of time sufficient to remove all doubt as to the electrical strength of the insulators, usually from three to five minutes. To facilitate cementing and testing, racks, as shown in Fig. 5, are provided so that the operations may be performed with a minimum of handling. There is one very marked advantage in this manner of applying the assembled test voltage, and it is, that the voltage may be put on before the cement is dry enough to permit handling of insulators so in case of failure by puncture of any shell, it may be separated from the others without injury, a thing absolutely impossible once the cement becomes dry.




Insulators for the lower voltages call merely for the introduction of sufficient good material, between line and supporting structure, to prevent destructive leakage. For voltages up to 20,000, little or no difficulty is experienced in securing satisfactory insulation, but above that voltage it becomes not only necessary to have good material in plenty, but it must be properly distributed in order that the danger of puncture and severe leakage be reduced to a minimum. The manufacture of good porcelain from a mechanical standpoint forbids that it be made so vitreous as to resist a puncture test of 65,000 to 70,000 volts over 1/2" or 5/8" thickness. It is vain to attempt to gain greater dielectric strength by increasing thickness since inherent cracks in thick pieces really reduce the effective electrical strength to that of 1/2" or 5/8" porcelain. Accordingly, it has come to be recognized as best practice to make no attempt to manufacture the higher voltage insulators in a single piece, but rather to secure great electrical strength by multiplicity of parts. Though entirely possible to construct 30,000 volt insulators of one piece and still safely apply a double potential test, years of experience have demonstrated that a multipart insulator is much less liable to fail entirely, when struck by stones or bullets, than are single piece insulators and so prevents complete shut-down of the transmission line. By far the larger proportion of shut-downs on the modern transmission line come from mischievously broken insulators. It is for this reason as well as for greater normal safety factor that engineers are selecting multipart insulators at higher cost, even for the lower voltages. In general, two-piece insulators are used for from 10,000 to 30,000 volts, three pieces for from 30,000 to 50,000 and four pieces for the 66,000 and 70,000 volt lines now in operation. Above 70,000 volts the underhung or suspended insulator works out most economically and, for entirely different reasons, 8 or 10 shells of porcelain are introduced.


Fig. 4. Method of Testing Insulator Shells.


Fig. 5. Testing Insulators While in Process of Cementing and Assembling.


Considerations of puncture strength are the first concern of the designer. As before noted, good electrical and mechanical porcelain should be made of not more than 5/8" thickness with an ultimate electrical strength of not more than 70,000 volts. To gain greater electrical strength it is necessary to introduce other shells which continue to add a proportional dielectric strength up to about 220,000 volts, at which point the curve gradually tends to become flat and becomes very nearly so at 300,000 volts, so that it is apparently useless to utilize more than 5 shells unless some other condition is introduced. With this to start with it is a very simple matter indeed to provide economical insulators for any voltage up to the point where difficulty of manufacture begins to interfere with progress. The 10,000 and 20,000 volt insulators have little to do with puncture strength provided, of course, that the porcelain is of good quality, and so the question reduces to an estimate of necessary leakage distance and carrying capacity when subjected to an artificial rain approximating the worst conditions to which the insulator will be subjected. This is not so simple as appears on the surface, for while an insulator may behave admirably under & downpour of unprecedented intensity, it is vain to expect Nature to limit herself to a vertical, or what is technically known as a 45 rain, nor does she stop after five, ten, or fifteen minutes, but is much more likely to accompany the downpour with sufficient wind to blow the rain very nearly horizontal, and to persist for any length of time. Under such conditions the insulator either meets its Waterloo or shows its mettle in the first fifteen minutes, for at the end of that time all portions, except possibly the inside of the innermost shell, have become thoroughly wet and the surface leakage comes into play. As is well known, the persistence of any arc of definite length has to do with the voltage impressed and the current flowing, and the failure of an insulator is of course amenable to the laws which govern arcs. A designer should assume that his insulator will sometime become wet all over and should provide sufficient leakage resistance to prevent any considerable amount of current flowing, a small amount merely tending to dry off the insulator whereas a large amount would quite likely start a disastrous arc under the shells and completely destroy them by its heat before extinguishing itself. Such an insulator represents very nearly the ideal and it is regretable that so few lines are equipped with insulators of so great a margin of safety. It must not be forgotten that only once or twice a year are Insulators called upon to perform extreme duty, in fact, the insulators of most lines are for the greater part of the time working very inefficiently, probably at not over 20% of their rated capacity.

The next best thing is to provide as much dry surface as possible under average storm conditions and, to this end, it is of value to note the progressive wetting of an insulator under rain. It is customary to mount the insulator well above the cross-arm, so that all its shells are far removed from water spattering up from the arm. Assume an insulator of conventional design of three or four shells. The top becomes dripping wet at once and the film of water held on the surface carries line potential to the outmost diameter. A certain amount of rain gets by the top and strikes upon the lower shells, part spattering upward and part downward, depending upon the angularity of precipitation. That portion which spatters upward soon wets the entire under surface of the shells above and in this manner the potential is carried down to the lowest projecting shell. At this point the leakage path is broken, due to the fact that beneath the lowest shell there is no surface from which spattering may occur and consequently it remains dry, but under this condition the central shell of the insulator is carrying the total strain both as to puncture strain and against flash over. Fig. 6 shows clearly the effect of spattering, the double line indicating the surfaces covered by a film of water and thus carrying line potential down to the bottom of the lowest projecting shell. It may be further noted that between this double line, or line potential, and the pin, there is but one effective shell which has been subjected to but 60,000 volt test, and lines of very moderate potentials have strains of this magnitude between line and earth. The shells must, of course, remain wet till the storm is over or some drying-action is introduced. If the gap between the edge of the lowest shell and the pin is small enough a small arc will be established, thus completing the circuit and permitting a leakage current, formerly held in check by the dry inner surface of the shell, to flow, though limited in volume by the relatively high resistance of the pure water film covering the insulator. This leakage current heats and vaporizes the water film, its first action naturally being where the current density is greatest at the neck of each shell, where, fortunately, the surface is not subject to the direct force of the rain. Since potential is In no way concerned with resistance, the thinnest and consequently highest resistance film of water is sufficient to inaugurate the drying process, so that an insulator in service cannot be caught unawares and, if properly shaped, can care for itself.


Fig. 6. Sectional Drawing of High-Tension Insulator.


Fig. 7. A Protected Shell Which Can Not Easily Become Wet.


Fig. 8. Straight Shell Insulator.


From the foregoing it will be seen that the dielectric strength of the inner shell of any multipart insulator and the distance between its lowest edge and the supporting pin, are of utmost importance, the remainder of the insulator serving as little more than a resistance in series or leakage distance to limit the amount of current which may flow when arcing takes place. To provide against the first requisite it is merely necessary to introduce a protected shell which cannot easily become wet. This is shown in Fig. 7, as is also the great distance between pin and edge of lowest shell. What has been said applies to insulators as it is necessary to build them to meet mechanical conditions imposed, and it is the mechanical characteristics, not the electrical, which influence their cost. Were it