ROWE: Lightning Protection of Power Lines in Guanajuato, Mexico

[Trade Journal]

Publication: American Institute Of Electrical Engineers

New York, NY, United States
p. 1239-1257, col. 1





The following is practically a summary of lightning trouble during the years 1904, 1905, and 1906, on a steel-tower long-span transmission line in the states of Michoacan and Guanajuato, Mexico.

The main transmission line, built in 1903, was the first one constructed on steel towers. These towers were of standard type, the height being 40 ft. from the top of the cross-arm to the ground. Three wires were placed upon the towers in the form of an equilateral triangle, the upper wire being supported upon a 3-in. extra heavy pipe, which formed the continuation of the tower, the other two being placed at the ends of a double channel-iron cross-arm, approximately 7 ft. in length, the sides of the triangle being 6 ft. The insulators used were 12 in. high and 14 in. in diameter. These insulators were cemented directly on cast-iron pins. The pins for the top insulators were made to screw on the 3-in. pipe; those for the side insulators had a square shank and were secured to the cross-arm by two 1-in. bolts. The conductors were 19-strand, hard-drawn copper cable, equivalent to No. 1 B. &. S. solid copper wire. There were no wooden pins or cross-arms used on the towers. For protection against the line voltage as well as any high voltages to lightning discharges, dependence was placed entirely on the insulators.

The ordinary length of span was 440 ft., but there were several spans of 1200 ft. and one of approximately 1600 ft. The line was designed for 60,000 volts at the generating end and 51,000 at the receiving end. The length of the line was approximately 101 miles.

The line is tapped at Irapuato, approximately 75 miles from the generating station and. 26 miles front Guanajuato. At this point there is a sub-station where lightning-arresters of the usual station type are installed to protect the step-down transformers. From this it will be seen that there is a chance for the line to discharge at this point as well as over the arresters at either end of the transmission line.


FIG. 1 — Showing line as originally constructed.
Fig. 1 — Showing Line As Originally Constructed.


Thus far there have been three rainy seasons, which in the part of Mexico where the line is situated, generally last from June until October. Heavy thunderstorms come up nearly every day during July and August, and there is a storm on an average of every two days through a period of four months. During the rainy season of 1904, there was considerable trouble with lightning; the next season there was less, and during the rainy season of 1906 comparatively little trouble was experienced. The nature of the troubles, the expedients tried to overcome them, and the provisions for overcoming future lightning troubles will now be considered.

The rainy season of 1904 came on with great severity in April, two months earlier than usual, and lasted until October. The injury to the transmission line from lightning was mostly continued to the puncturing of the top insulators by direct bolts of lightning, and in most cases a hole was bored through the top insulator nearly in a vertical line to the pin. These holes were approximately one inch in diameter and the sides were glazed. Usually, the insulators were badly shattered, but on putting the parts together one could generally find the glazed hole above referred to. Some side insulators were lost, but they were almost invariably injured at the time when the top insulators on the same or adjacent towers were injured. The side insulators were never punctured from the top in a vertical direction, but in some cases they showed a small puncture in a horizontal direction from the tie-wire to the pin. In nine cases out of ten, however, they were not punctured at all, but half of the top and portions of one or both petticoats on the same side were broken, presumably by the power current following lightning discharges over their surfaces.


FIG. 2 — Showing first change.  Top insulator lowered, grounded cable in place, and lightning-rods left in position.
Fig. 2 — Showing First Change. Top Insulator Lowered, Grounded Cable in Place, and Lightning-Rods Left in Position.


As the principal trouble apparently came from direct bolts of lightning striking the top insulators, it was thought advisable to erect lightning-rods on the towers on the section of the line where most trouble had occurred. At that time the placing of a grounded cable over the transmission wires was discussed, but as this could not be done before the end of the rainy season, for suitable material could not be obtained in Mexico, it was decided to put up lightning-rods and get the benefit of experience with them during the rainy season then in progress.

Lightning-rods were erected in pairs starting from the 3-in. pipe at the point where the cross-arm in pairs was attached, and projecting up on either side of this pipe to a distance of 6 ft. above the top insulator at an angle of about 30 degrees from the vertical.

By the middle of August one-half of the line giving the most trouble was equipped with lightning-rods, and there had been erected quite a few rods on towers that were considered as being particularly exposed to danger from lightning. After the erection of these rods the difficulties were very much less. although there was still some trouble on the section of the line where they were in place. In no case were top insulators punctured through from the top as on the towers where rods were not in place, but when insulators were injured they usually had a part of the top and a petticoat broken on the same side. There were also a few cases where insulators were punctured in a horizontal direction from the tie-wire to the pin. During one very severe storm, eight insulators were broken on seven towers spread out over a distance of 14 towers. As the towers were spaced 12 to the mile it is improbable that the same discharge of lightning could have caused all of this trouble, unless it was transmitted over the power lines. Of the eight insulators injured none was punctured; they were apparently damaged by the current going over their side. Although these insulators were badly injured, the service on them was continued, but at a slightly reduced voltage.

One quite remarkable case happened in the Guanajuato substation, due, perhaps, to a bolt of lightning striking the line about a mile away. One of the men was standing in a balcony in Guanajuato watching the lightning display, when apparently a very heavy discharge struck the tower line, located over a hill just out of sight. At the same moment the lights in the room went out, showing that the service had been interrupted. The lightning which caused this interruption of service came into the sub-station over the top wire and jumped 4.5 ft. through air to ground, at the same time going to ground over the lightning-arresters. An insulator over which the arc formed in the sub-station had half of the top cracked off, but no injury was done to the transformers or other equipment. The towers were carefully examined near where the lightning was seen to fall, but there were no broken insulators or other evidence to show that the bolt had struck the line. However, as lightning had been previously seen to strike the towers without causing trouble, it was thought that a bolt of lightning had struck the tower line, a portion of this discharge coming over the wires into the sub-station. At the place where the bolt was supposed to have struck, the towers, on exposed places, were equipped with lightning-rods.

These cases are cited to show that insulators at a distance from heavy discharges of lightning may be subjected to very severe voltages, and, from all the data at hand, it is believed that the lightning-rods were efficient in protecting the insulators from direct bolts of lightning, and that the trouble that was still present on the section of the line equipped with lightning-rods was due to discharges traveling along the transmission wires.

During the season of 1904, the line troubles were quite evenly distributed over the first 50 miles of line out of the generating station. There were only two cases of trouble outside of this section that caused injury to the line insulators. Equipping this section with lightning-rods was begun in June and finished in August. This section of the line, except for a stretch of 15 miles, was left as equipped in August, 1904, until October, 1906, or through more than two lightning seasons. As there was less trouble on this section of the line during these two seasons than during a portion of the lightning season of 1904, it is evident that the lightning-rods were at least partly effective.


FIG. 3 — Showing present appearance of line.  Lightning-rods removed.
Fig. 3 — Showing Present Appearance of Line. Lightning-Rods Removed.


After finding that lightning-rods were not a complete protection, it was decided to make the experiment of lowering the top wire and putting up a grounded cable in its place. However, on account of delay in receiving material, very little was accomplished in time for the rainy season of 1905 and the experience with lightning during that season was the same as that obtained during the rainy season of the previous year, except that there was much less trouble.

In the winter of 1905-1906, it was decided to install larger insulators on the tower line and at the same time to lower all of the top cable to a position below one of the original side insulators. On the highest point of the tower, in place of the top transmission cable, it was decided to string over the entire line a 1/4-in. 7-strand steel cable, grounded at each tower. Insulators 17 1/2 in. in diameter, 20 in. high, were adopted for the first 60 miles of line out of the generating- station, and the same insulator with a 14-in. top for the other 40 miles of line. It was unfortunate that the insulators with the 17 1/2-in. top were not ready so as to complete the line changes in time for the rains, but half of the line out of the Guanajuato sub-station was finished before the time for severe lightning storms had arrived. The other half of the line had been equipped the first year with lightning-rods, which welt still in place, the top cable being in its original position on this section, except for a distance of 15 miles.

During the rainy season of 1906, the transformers were connected in star on the high-tension side, the center of the star being grounded, and in order to detect instantly when there was a ground on the line, a series transformer was put into this grounded line and the secondary leads carried to the switchboard, where an ammeter was connected to them. With this arrangement, when an insulator broke down, or the power current followed the lightning over the insulator, the station operator could tell when there was a ground on the line, and, by shutting down at once, he was able to prevent the burning off of the transmission wire. Usually, when a ground appeared and the power was cut off, the line was found in good condition on again starting up.


FIG. 4 — Insulator perforated and broken by lightning bolt.
Fig. 4 — Insulator Perforated and Broken by Lightning Bolt.


There was no apparent trouble from lightning during the rainy season of 1906 on the section where the grounded cable was in place. On several occasions grounds came on the lines during lightning storms, and the power was immediately cut off in order to clear them. In general: these grounds could not be located, so it cannot be said that lightning did not go over the surface of the insulators on the section where the grounded cable was in place. The section where there was no grounded cable in place, but where the lightning-rods were erected, still gave some trouble, but the total amount was much less than in any preceding year.

The part of the line found to be the hardest to protect during all three rainy seasons was the middle of the 75-mile section from the generating station to Irapuato; in other words, the part of the line farthest from the location of lightning-arresters. This would indicate that arresters for discharging the line, located say every 25 miles, would be a valuable protection.

The experience of the last three seasons prompts the writer to offer the following suggestions:

1. Insulators should not be disposed upon poles or towers so that they will be in the path of bolts of lightning going to ground by the supporting structure.

2. A grounded cable, strung above the transmission wires at highest point of the tower, is certainly more effective than lightning-rods in protecting the insulators and conductors from direct bolts of lightning.

3. Lightning-arresters, for discharging the line in case very high voltages are present on the wires, would be of some value if located along the line at frequent intervals.

4. On steel construction, an insulator should be used that has a high margin of safety against puncture and arcing over.

It seems at present general practice to use insulators for high potentials with practically no margin of safety, and when the conditions are severe there is almost sure to be trouble. Aside from trouble with the puncturing of line insulators by direct bolts of lightning, which was apparently prevented by the use of lightning-rods on towers, nearly, if not quite all, of the trouble on this line has come from the breaking down of insulators by the current going over their sides, or through them, from tie-wire to pin. The writer ventures to assert that, if the insulators had been much better, there would have been little trouble on the line where the lightning-rods were in place. Moreover, as a grounded cable undoubtedly affords better protection than lightning-rods, with a suitable grounded cable strung over the transmission wires, and better insulators so disposed as to be out of the path of direct bolts of lightning, even without the installation of lightning-arresters along the line there would result a line that should be practically free from lightning troubles.

NOTE.—While personal experience suggests that the use, on iron construction, of insulators with a high factor of safety will greatly reduce troubles from lightning, it was decided to change our line insulators on account of trouble from an entirely different cause. During some months of the year the insulators become covered with condensed moisture just at sunrise, due to a very sudden rise in temperature, from the temperature of about zero centigrade. This often left insulators, on the portion of the line where the temperature variation was most severe, covered with condensed moisture nearly as heavy as hoar-frost, and, during the few moments when this effect lasted, the power current was apt to arc over the surfaces of the insulators. The old insulators were by the company's engineers as the best obtainable in 1902, and the insulators decided upon in January, 1906, were the largest commercial insulators on the market. A very much better insulator was desired, but for want of time to develop something better it was necessary to adopt an insulator with a factor of safety somewhat lower than that intended to be used. The apparent freedom from lightning trouble on the part of the line equipped with larger insulators and grounded conductor is due perhaps as much to the use of better insulators as to the grounded cable.




William Hoopes: The method of treatment of this problem appears so excellent that one is at once interested in its practical application. The first conclusion reached is that the economical span-length is determined, not by the original cost of the line, but by its annual cost.

The annual cost is made up of three items:

1. Interest and depreciation on the first cost.

2. The cost of repairs and patrolling.

3. The money damage from interruptions to the service.

If lengthening the span will reduce the number of interruptions and cost of repairs, then the economical span is longer than that which gives the lowest first cost.

Inquiry into the operation of a large number of transmission lines reveals the fact that by far the larger portion of the interruptions of service is due to trouble occurring at the point of support; this applies particularly to steel-tower lines. Reduction of the number of supports does, therefore, reduce the annual cost.

The paper shows that this particular line on 800-ft. spans would cost about $300 more per mile than if on 400-ft. spans. Interest and depreciation on this at 10% would be $30 per year. Halving the number of supports would probably save much more than this.

I believe the subject has not been treated in this way in this or any other paper, so I should like to suggest to the committee that such a paper would open up a very live topic.

However, the province of Mr. Scholes' paper is really to show the least first cost of the line, and the above remarks are not strictly germane to it. Investigation of its practical application leads to the following queries:

1. Is the assumption of a uniform price per pound justifiable?

2. Does the retention of the same geometrical figure permit the design of all the towers for the least cost?

3. It is fair to assume that large foundations for higher towers cost as much per cubic yard as small foundations for low towers.

The cost of a tower to the purchaser is made up of the following items:

1. Cost of steel and transportation.

2. Cost of shop work.

3. Cost of galvanizing.

4. Cost of erection.

5. Manufacturer's profit.

If the cost of the tower is directly proportional to the weight of the steel, then all of the items of cost must vary at the same rate as the weight. The cost of the steel does vary approximately as the weight.

The paper gives the cost of a 34.5-ft. tower as $30, and of a 70.5-ft. tower as $243, which is about proportional to the cubes of the heights. The other costs should therefore be as the cubes of the heights.

Inquiry from a concern which makes a very large number of towers elicited the following. opinions:

1. As the number of parts is the same for large as for small towers, the number of shop operations will be about the same; but as the shop operations will be slower on the heavier work, the shop cost will vary about as the height of the tower.

2. The same opinion was expressed with regard to the cost of erection.

3. The galvanizing cost is approximately proportional to the superficial area, _or to the square of the height. The galvanizing was said to be a very material portion of the whole cost.

When it came to a question of manufacturer's profit the source of my information ran dry, but if it increases as the cube of the height, it would seem to afford a considerable opportunity to a resourceful purchasing agent.

From the foregoing it would seem that all the costs which go to make up the cost of the tower, other than th