November 1980 Foreign Insulators

Reprinted from "INSULATORS - Crown Jewels of the Wire", November 1980, page 4

The name "Alice Springs" is really a misnomer, because the deep pool of water after which the place is named is not a "spring" but a soak. The town, a panorama of which is given below, is in the centre of the continent on the Overland Telegraph Line between Adelaide and Darwin. The locality was given its name by a party engaged on the construction of the Overland Telegraph Line in 1871 when they sighted the big, deep waterhole in the bed of the dry creek or river now known as the River Todd.

The Todd only flows after each rain, and at this place there is a large outcrop of granite rocks. The swirling of the quickly flowing water keeps this big hole washed out and leaves it full of water. Quite good soakage water can be obtained anywhere in the creek at about six feet down, and this soakage keeps the hole full of water, and because the hole does not dry out, it was probably thought by the party that there must be a spring at that place. There are, however, many springs in the MacDonnell ranges, but none at the particular spot after which the old Telegraph Station was named.

Although the continent was crossed for the first time by the explorer (McDouall Stuart) in 1862, less than ten years later a telegraph line had been completed practically along the route taken by him. In the late 1860's there was great rivalry between Queensland and South Australia as to who should have the honour of linking their telegraph system with the cable being laid from Singapore to Darwin. Each wanted the other Australian colonies to support the project from their particular point of view. When in June, 1870, no agreement had been arrived at and the Cable Company seemed to favour the land line being constructed from Darwin to Brisbane, the South Australian Government, evidently realizing that some drastic action was necessary if they were to have the line in their State, made an offer to the Cable Company to build a telegraph line from Adelaide to Darwin (1975 miles) and have it completed by the time the Company had completed their cable, this being estimated at eighteen months from the time of the offer. The offer, which meant that the telegraph line had to be constructed at the rate of 110 miles per month, was accepted, and heavy penalties for non-completion of the line in the time were provided for in the agreement.

Actually it was not until about the middle of August that year that the construction was commenced. The route was divided into three sections -- from Port Augusta to latitude 27 deg. S. (about 60 miles north of where Oodnadatta is now, a distance of 550 miles), from 27 deg. S. to 19 deg. 30 mins. S. (approximately where the present town of Tennant Creek is, about 570 miles), and thence to Darwin, approximately 650 miles. The first section was in more or less settled country and provided very little difficulty, the northern section had some difficulties but not very great, but the centre section was in practically unknown country and therefore was the most difficult. Each section was subdivided into many sub-sections, and a party allotted to construct each sub-section. A small exploring party went ahead of each main party and marked out the route to be taken. The equipment of each party included 15 horse wagons, 17 bullock drays, one bullock wagon, five express wagons, 165 horses and 200 bullocks. A depot was established at the Finke River (about 830 miles from Adelaide) for the provision of fresh meat for the men working on the adjoining sections, and 2000 sheep were sent there. It must be remembered that all the material, provisions, etc., had to be hauled from either Port Augusta or Darwin by horse, bullock vehicle, or camels, and some idea of the difficulties experienced can be realized by the fact that it took Harvey's party, who constructed one of the central sub-sections, eight months to reach the beginning of their section.

It was far too big a job to be done in the time, and when the period had expired (December, 1871), there were still many gaps in the line. A delay had also occurred in the cable construction, and although not far off completion, the cable was not completed on the contracted date. A compromise was reached regarding the infliction of penalties, which were considerably reduced but not entirely abolished, and the South Australian Government redoubled its efforts, but it was not until 22nd August, 1872, that the last gap was closed and telegraphic communication established between Australia and England. The total cost of the line was £479,154.

The original line was a 7/14 stranded iron wire conductor, and although most of it was removed and replaced by a 400 lb. G.I. conductor many years ago, there are still some small sections of the original wire in use.

During October, 1938, it became necessary to remove a small piece of the original wire in connection with the establishment of a Telephone Office at Finke, and it was found that the old wire was in perfect condition and not showing any signs of deterioration. Several types of insulators appear to have been used. One type was of porcelain, about 4-1/2 inches across at the bottom, but having a metal top, two inches in diameter, screwed on to the porcelain.

Old types of insulators used on the original Overland Telegraph Line and a piece of the original iron wire. The two outer insulators are the metal armoured type and the hole. through which the wire was passed can be seen in the top of the centre insulator.

A metal plate bolts on to the metal top, and two holes, through which a wire could be passed were formed when the plate was screwed down. Apparently only one hole was used, but the tightening of the bolt held the wire firmly between the plate and the metal top of the insulator. There were thus no tie wires necessary with this type. Another type is similar to the present day trunk line insulators, a little smaller, but completely covered with a metal armour. The metal armour is shaped exactly like the insulator, and the wire was tied to this similarly as is done on present day porcelain insulators. The porcelain was set into the armour by a kind of cement, and a thread was provided in the porcelain for the spindle. A number of these insulators are still to be seen lying along the line, and although over 66 years old, do not show the slightest sign of rust or deterioration.

In many instances the line did not take a direct route between various points, but followed creeks and watercourses. The reason for this was that most of the poles were cut from the timbers growing along these watercourses, and also that it was necessary to follow them in order to obtain water. However, the white ants soon showed their presence, and although there are some of the original butts still to be seen besides the present iron poles, most of the wooden poles had very short lives, perhaps only of a few years' duration. In 1880, re-poling with Siemens and Oppenheimer poles was commenced in places, but it was not until 1898, when a 265 lb. copper conductor was added that the line was fully iron-poled. During the erection of the copper wire and the final iron poles, the line route was considerably straightened, now following a more direct route and not keeping to the watercourses. No more wires have been added since then, but the methods of telegraphy used have kept abreast of the times and enabled the growing volume of traffic to be handled satisfactorily.

Originally the messages were repeated by hand at several stations along the route, and in between these stations were many others at which linemen were located. These latter were placed at points where water could be obtained, and varied in distance from. 95 to 180 miles apart. As the country became opened up and telegraph systems improved, it became possible to abolish many of these stations, and one by one they have passed out of the Department's control. Some are now police stations, some are cattle station homesteads, and others railway stations. Today the only repeater stations apart from Port Augusta, in the circuits, are Alice Springs on the copper line, and Marree, Powell Creek and Alice Springs on the iron. Besides these the only other stations remaining in the Department's hands are Tennant Creek, Daly Waters and Katherine.

In time, hand repeating gave way to "pole changer" repeaters, and about 1926 relay repeaters were installed at the three repeater stations.

Nearby the pool which the first party had mistaken for springs, the Alice Springs Telegraph Station was built. For many years it was a lonely outpost, receiving its mail only once every six weeks or two months. At first it came by packs and camels from Port Augusta, later from Marree, still later from Oodnadatta, and in 1927 the railway line was completed to Stuart Town, two miles south of Alice Springs Telegraph Station. Although Stuart Town was surveyed in the late 1890's, it did not take shape until the completion of the railway. With the growth of the town it became necessary to establish an official office, and in 1932 the old telegraph station at Alice Springs was closed and a new post office opened in what was originally Stuart Town, but which now had its name changed to "Alice Springs."

It is not often that we get a good downpour of rain, but we usually manage one good one each year. On 19th February, 1938, from two to five inches of rain fell from Alice Springs to Oodnadatta. Rivers ran bankers, and at one time the Alberga was flowing 16 feet above the railway bridge and over half a mile wide. Communication with Adelaide was completely cut off for several days, and the train service was not restored until six weeks later. Miles of railway line was washed away and many sections of the telegraph line carried away in the floods. Where the telegraph lines cross the Alberga there were two 28 ft. iron beam poles, but after the water had subsided they were found in different places about a quarter of a mile downstream, one end of each just showing through the mud and the other end under between six and seven feet of mud and silt. Some food supplies were brought to Alice Springs by aeroplane, but by the time the train service had been restored, there were many items which were unprocurable in the town, despite that rationing had been resorted to.

When the waters had subsided sufficiently to allow movement, Mr. E. Colson, of Bloods Creek Telephone Office (the nearest resident to the Alberga River), offered to attempt repairs to the lines, and set out on camels to do the job. Progress was slow, as it is a fair day's ride to do 30 miles by camel, especially in wet country, and stoppages were frequent in order to clear debris away from the lines. In addition, many detours had to be made to get across creeks and rivers which were still flowing, and he estimated that to traverse the 100 miles of line from his place to Oodnadatta, he traveled 160 miles. At the Alberga crossing he and three natives worked most of one day in water up to their arm-pits, rigging up temporary supports for the line. These were tripods about 10 feet high, and were constructed from timber cut from the trees growing nearby. Insulators were tied on to them, and the lines in turn tied to the insulators. Altogether it took eight days to travel and repair the 105 miles of line between Bloods Creek and Oodnadatta.

On the south side of the Alberga no wire was left on the poles (which were washed out and bent all shapes) for a quarter of a mile, the wire evidently having been caught and carried away by debris which was washed down. Many miles of railway line and two bridges were washed away, and it took nearly five months to repair the permanent way. It is. interesting to note that none of the railway bridges carried away have been replaced. Instead, the line is laid on a built-in rock foundation in the bed of the creek or river. Excavations are made to some depth in the river bed and stone or rock firmly packed therein and the line laid on top of the stone. This method was first tried out when the Finke bridge carried away about six years ago, and has proved to be quite effective for, at the most, the damage amounts only to a short length of railway line being twisted, instead of an irreparable costly bridge.

With the extension of the practice of dialing over trunk lines, troubles have been experienced on account of the large variation in the insulation resistance of the aerial circuits used. Some dialing lines, in wet weather, have an insulation resistance of 10,000 ohms or even less, and become unworkable, and in 1936 it was decided that the standard trunk insulator should be re-designed in order to effect, if possible, an improvement in this state of affairs. It was recognized that, although there are several sources of loss in a telephone insulator, by far the most important one at voice and telegraph frequencies is surface leakage between wire and pin, and the object of the designers was, therefore, to reduce this effect.

With a clean insulator the surface leakage, even in wet weather, is small, but, as dirt accumulates, a semi-conducting path is formed on the surface if the atmosphere is moist, particularly if the dirt includes salt or carbon. The resistance of this path depends upon its length and cross-sectional area in the same way as the resistance of any conductor depends upon these variables, and it is therefore desirable to make the length of the path as great as possible whilst, at the same time, keeping its cross-section area as small as possible, i.e., by keeping the circumference of the insulator as small as possible.

This result is partially achieved in many insulators by using two skirts. The trunk insulator which has been standard in Australia for many years is one design which incorporates this feature, and so is the B.P.O. trunk insulator. These two designs are shown in Figs. 1 and 2. Examination of them shows that the B.P.O. insulator provides a longer leakage path between wire and pin (32 cms. as compared with 23 cms.), and at the same time the average diameter of the leakage surface is considerably less (4.60 cms. as compared with 7.54 cms.). Thus, if each of the insulators has an equal thickness of similar deposit upon it, the surface resistance of the B.P.O. insulator will be nearly 2.3 times as great as that of the Australian trunk insulator. In addition, exposure tests conducted at the Research Laboratories indicated that the external surface of the Australian insulator which curves outwards is more likely to accumulate dirt than the straight vertical sides of the B.P.O. design.

It was therefore decided to use the B.P.O. insulator as the basis for design, but to attempt to make it suitable for use with wooden spindles which, in Australia, are more economical than steel spindles. The use of wooden spindles would necessitate increasing the diameter, for it would not be practicable to obtain a sufficiently strong wooden spindle to fit the thread of the B.P.O. design.

This increase of diameter would tend to reduce the surface leakage resistance and also would increase the weight of the insulator; it would therefore be undesirable. Instead of abandoning the wooden spindle the designers decided to depart from the B.P.O. practice and eliminate the second skirt. They argued that, although by doing so they would reduce the length of the leakage path, they would also reduce the average diameter of the insulator, and thus the two effects would partly cancel out. Furthermore, they believed that the outer surface, which is continually washed by rain, is the most important part, and that the elimination of the inner skirt would not be as serious as might be imagined. To obtain the maximum benefit, however, a wooden spindle of minimum practicable diameter was designed for use with the new insulator, the final design of which is shown in Fig. 3. It is hoped that, when tests have been completed, this insulator will become the Australian Standard Trunk Insulator.

The length of the surface leakage path of this insulator is slightly less than that of the old trunk insulator (18 cms. as compared with 23 cms.) but its average diameter is considerably less (5.08 cms. as compared with 7.54 cms.) and therefore its theoretical surface leakage resistance is greater than that of the old design. Furthermore, as stated above, a large proportion (57 per cent.) of its leakage path is exposed to the cleansing action of rain. With the old trunk insulator, this proportion is 87 per cent. and with the B.P.O. insulator it is 32 per cent. The designers hope that this feature of the new insulator will largely compensate for the higher theoretical leakage resistance of the B.P.O. insulator.

After the design had been completed, it was decided to conduct field tests by installing the insulators on the Melbourne-Geelong trunk route but, before this was done, the decision to scrap this route made other plans necessary, and it was decided to equip certain lines between Melbourne and Dromana which were known to have very low insulation resistance during periods of high humidity. It was. intended to equip one circuit with the new insulators made of porcelain, and one circuit with B.P.O. porcelain insulators, and to compare the results obtained with the standard Australian porcelain insulators erected on the same poles. It was then thought advisable to make some of the new insulators in glass in order that the value of this material for the purpose could be examined. The Australian Glass Go. agreed to co-operate but pointed out that the new design, although suitable for porcelain, was not ideal for glass because of the cooling stresses likely to be set up in the body of the insulator. They submitted a design which gave the same leakage path but which, they considered, would be better for glass, and the Department purchased sufficient of these insulators to equip a circuit between Melbourne and Dromana. Fig. 4 shows a glass insulator of the design tested.

[At the time these tests were being made (1936) Australian Glass Company's insulators would be marked Agee (AGEE). I don't recall seeing any insulators so marked with this design. It is not a CD 130.7, though similar. Maybe it was discarded, and later evolved into the 130.7 we all know, and marked A.G.M. M.A.]

The installation of the insulators on the Melbourne-Dromana circuits was completed in March, 1938, and each of the circuits was connected to the Research Laboratories by an underground cable pair between the Glenhuntly cable head at the Melbourne end of the circuits, and the Laboratories, and a recording galvanometer was arranged to record the insulation resistance of each of the circuits. The method of connection did not affect the usefulness of the circuits for their normal traffic, and they were not, therefore, withdrawn from service. Before tests commenced, the insulators on these lines were cleaned by spraying them from the ground.

In July, 1938, the first analysis of the results was made. As was known, the insulation of the various lines varied considerably with the humidity, and in order to compare the results, it was therefore necessary to compare the readings at equal humidities, and for the purpose of these tests, the readings at high humidities are, of course, the most important because it is under these conditions that lowest insulations occur.

Some variations occurred in the day by day readings, but the lowest insulation resistances recorded at any time were as under: --

These readings are the overall readings between wires of the lines which are approximately 46 miles long; the lowest readings at other humidities were in about the same order. For example, at 80 per cent. they were:--

These results are very satisfactory and, so far, confirm the designers' belief in the single skirt design. It would be unwise to be unduly optimistic until further results have been obtained and the effect of weathering has been ascertained, but on the results so far, the Department has felt justified in purchasing large supplies of trunk insulators to the new design. This will, of course, introduce complications such as are always met when an old standard is replaced by a new, but it is felt that the advantages to be obtained are very real and far outweigh the objections.

The reason for the low value of the glass insulator is not very clear at present, but it is possibly due to the presence of cracked insulators on the line. During installation, some of these insulators cracked and a number actually broke. Whatever the reason, however, the results so far reflect some inferiority of the glass insulators installed for test, which may or may not become more pronounced as the tests proceed.

It is possible that an extra groove may be added to the design of Fig. 3, but unless future tests cause opinions to be modified considerably, it is probable that no drastic alterations to it will be made and that it will become the new Departmental Standard, and will be known as Insulator, Trunk, L.S. (Long Skirt) to distinguish it from the present Trunk Insulator. The new wooden spindle for use with it has already been designed, and designs for the necessary steel spindles are in course of preparation.