must be adopted for protecting any portion of an electric supply-line or any support, guard wire, or bearer wire of an electric supply-line which is exposed in such a position as to be liable to injury from lightning.

A direct lightning stroke will flash over the insulators and probably destroy the line at that point, but inductive surges may keep to the line and affect the apparatus in the power station, or may spark across an insulator and put the line to earth. The voltage of lightning may be inconceivably great, and it is an oscillating discharge of a variable, but always very high, frequency. It will therefore generally take the least inductive, but not necessarily the best conducting, path to earth; and lightning arrestors are designed to take advantage of this fact. They frequently consist of a sparking gap, in one form or another, which the ordinary line voltage cannot bridge over, together with a carbon or other resistance in series. Should a highpressure discharge cause an arc to strike across the gap, the line current will follow, but the series resistance limits the current and the arc dies down or is automatically extinguished, magnetically or otherwise. In order to ensure the discharge taking its allotted path, an inductive coil or 'kicking coil' is generally connected in series with the line itself on the station side of the arrestor; the oscillating discharge is thereby forced to follow the alternative path. In some places in South Africa iron inductance coils have been used instead of the more usual copper coils, and have been found more effective. The following extract from a paper by Dr C. P. Steinmetz relates to these matters :—

An interesting application of the relation between the maximum value of oscillating current and the maximum value of oscillating voltage is to the calculation of the maximum lightning discharge from a transmission line, against which the stations have to be protected. While the voltage of lightning may be practically unlimited, the maximum voltage which lightning can produce on the line is obviously limited by the disruptive strength of the line insulators against momentary voltages.

Assuming a 30 000V line, the insulators must be expected to stand continuously double voltage or 60 000V, and momentarily at least twice as much or 120 000 to 150 000V. Thus the lightning discharge current is limited to that given by the oscillating voltage of maximum value = 150 000V. At a natural impedance of the line=400 O, this gives 375A as the maximum discharge current which lightning could produce in this line, and thus the maximum current against which the lightning arrestors in the station have to protect in case of a lightning stroke on the line near the station. ' If the lightning strikes at a considerable distance from the station, obviously the discharge current at the station is much less, as the energy of the oscillation is rapidly dissipated in its passage along the line.

The commonest form of lightning arrestor consists of two bent rods of copper, fixed near the bottom at the right distance apart for the spark gap (according to the pressure) and thence diverging upwards; when the gap is bridged by a spark, due to lightning or a surge on the line, the arc caused by the line current runs up the two horns until it breaks from excessive length. One horn is connected to the line and the other to 'earth.'1 The horn gap arrestor is capable of dealing with heavy surges, but it can only operate on an over-voltage and is useless on a low-voltage, high-frequency surge. A useful setting is stated to be 1 mm. per 1000V plus 1 mm. to prevent constant discharge. This 'horn arrestor' device is sometimes used as an overhead line switch, the gap being short-circuited by a bar which can be opened by means of a long insulating rod. Multiple gap arrestors consist of a series of cylinders, with a short gap between each pair, the terminal elements connected to line and earth respectively; sometimes the cylinders are connected in series with a high resistance, and in other cases some of the gaps are shunted by a resistance. The arc is broken either by the cooling effect of the metal or bythe use of certain non-arcing metals—zinc, antimony,and bismuth.

Certain types of arrestor do not employ a spark gap directly; in the jet arrestor the line is connected straight to earth through a fine jet of water, the resistance of which is so high that the normal leakage is very small; the jet arrestor provides a permanent leak to earth for high-frequency surges, irrespective of voltage; it is particularly useful for dispelling slowly accumulating static charges. In the case of low-tension C.C. installations a tank is substituted, and conducting plates within it are adjusted to allow a small continuous leakage from pole to pole, through the water, whenever there is a likelihood of lightning troubles.

The arrestor which appears to give the best results on long lines is the aluminium cell, or electrolytic arrestor. It consists of a number of cones of aluminium, threaded on insulating rods one over the other, with a small amount of electrolyte connecting each to its neighbour, the whole being immersed in oil. A film of hydroxide of aluminium forms on both anode and kathode when such a cell is charged by alternating current, and the resistance consequently becomes so high that practically no current can pass at ordinary pressures. At about 350V per cell the film breaks down, so a sufficient number are put in series to ensure a discharge at about 50% above normal line 1 See a paper by C. C. Garrard in The Electrician, 72, 996.

pressure. When this occurs the resistance falls to that of the electrolyte, and the line is practically earthed through whatever damping resistance is inserted in series. To prevent disintegration from continual leakage through the arrestor, a horn gap is always placed in series also, and this is short-circuited daily for a few seconds in order to recharge the cell.1

On 3-phase lines it is necessary to remember that if the arrestors on all phases are connected to a common earth, and if there is in

[graphic][graphic][merged small]

sufficient damping resistance in circuit with each, the lines may be practically short-circuited. With horn arrestors this is not an uncommon mistake, and the arrestors are blamed for the consequent trouble.

In the illustration (fig. 81) G represents any number of generators serving one 3-phase line, and reactances or choke coils are represented by R on all three lines from the switchboard; for the purpose of illustration three types of arrestor are portrayed on the three phases,

1 For a fuller description of protective devices, see a paper by Kenelm Edgcumbe on over-pressure protective gear for high-tension circuits [Electrical Review, 75, 712 et seq.).

viz. an electrolytic arrestor in series with a Siemens horn gap; a multiple cylinder arrestor in series with a non-inductive resistance; and a plain horn gap. Actually the three arrestors would be all of the same type on any particular line. It will be seen that they are placed at the beginning of the line, so as to protect all the apparatus on the generator side.

Some engineers have little faith in lightning arrestors,1 especially those having a spark gap, whether for actual lightning protection or to guard against surges in the line. Various protective devices of a different nature are employed alternatively, as, e.g., the Moscicki condenser and the Merz-Hunter and Merz-Price devices. The former have been installed in various stations in Rhodesia, and are fully considered in a paper by Wragg (Trans. South African I.E.E., June 1915). The condensers are sensitive to high-frequency surges of any voltage; they will take care of rapid surges; they have no permanent leak to earth and can be installed where water is not available for jets. Mechanically they appear to be faulty at present. As regards the latter devices, A. P. Trotter (Electrical Adviser to the Board of Trade) makes the following remarks :—

There is one thing I should like to mention about the Merz-Hunter, the MerzPrice, or any other gear. The gear depends, as is well known, upon a small arma ture that can be attracted. If that works freely, and does its work, it is all right; but those who have had experience of working with automatic gear know that sometimes, if the gear is set fine enough to do the work intended, it operates too often. If these gears operate every time that a bird pecks at an insulator, or a bit of wet straw blows across the line, and also exhibit nervous tendencies and cause trouble, the obvious thing would be to tie them up with a bit of string. {Journal I.E.E., 52, 219.)

A considerable amount of literature exists on the subject of abnormal pressure rises in circuits due to resonance (§ 90); to switching operations, e.g. the sudden breaking of an inductive load (§ 47); to the sudden switching on of a condenser, in the form of an unloaded cable; or to the instability of arcs and sparks (§ 123). Reference may be made particularly to articles by C. C. Garrard in The Electrician, 75, 350 et seq.

A continuous steel earth-wire (often barbed) is generally run along the top of the poles of a transmission line, and serves to protect the conductors to some extent from lightning. If underground cables are used to carry aerial lines across a street, as is sometimes the case, the overhead earth-wire should be brought down also, and wound

1 See remarks by P. V. Hunter, Journal I.E E., 52, 302. on to the cable sheathing; and £arth plates should be provided at each side of the crossing. The wire should in any case be earthed at every second or third pole if possible. Where no good natural earth can. be found in which to bury the coil of earth-wire, an iron pipe is driven into the ground and either coke or salt filled in; it should be placed where rain will run in. The earth connection for all types of protective devices should be as short as possible to be effective, and especially for condensers. The following extract is of interest in connection with these considerations :—.

Lightning cannot directly affect an underground cable; therefore, as an overhead line cannot be underground, the next best thing is to maintain a groundpotential zone by the continuous earth-wire.

Further protection of switch-gear will be found expedient to deal with surges, and this may take the form of spark gaps in series with carbon rods; the gaps are formed by metal cylinders about 1 mm. apart in numbers according to the voltage of supply. They come into action when a high-pressure surge is set up, the principle being that the excess voltage breaks across the gaps, and while the cylinders cool the arc, the carbon rods limit amount of current.

In cases where the authorities insist on cables being underground, such as at crossings, no lightning arrestors beyond the ordinary pole lightning conductors are necessary, as the short length of cable underground is quite able to look after itself; but the overhead earth-wire must be continuous and, if carried out by bonding to the armouring of the cable, earth plates should preferably be sunk at each side of the crossing. These refinements whilst ensuring an efficient line need not entail great expense, and the benefits to be derived are valuable. (Burrows Assoc. M.E.E., January 1911.)

485. Lightning Conductors.—Reference has already been made in § 223 to the protection of installations from lightning, and the devices used are further described in § 484. A word may here be added as regards the protection of buildings from lightning.1 In order to ensure absolute protection a complete network of wires, connected with earth at many points, would be made to cover the whole structure; this is indeed practically done in the case of magazines. Ordinarily one or more lightning conductors are used, according to the size of the building, and the nearer the arrangement corresponds to a complete screen the better will the protection be. Iron or steel wire or stranded rope is considered to be more effective than copper, and is also less liable to be cut away and sold. Copper strips of large section, ending 7 or 8 ft. above ground level, were once a common sight on public buildings in India. If copper is used it is a disadvantage to use highconductivity metal, such as is used for wiring, as it increases the surges

1 A very instructive address on this subject, by A. Hands, is reported in Electricity, 30, 126, 185, 191.

« PreviousContinue »