Page images




If the distributory system of any gas undertaking is to be as perfect as possible, it is imperative that the medium through which the distribution takes place .should be of the very best, both in kind and quality. The "efficiency " and " economy " so frequently spoken of, but not quite so often attained, clearly make this essential.

In the early history of gas lighting many very ingenious suggestions were made as to the cheapest and most suitable kinds of pipes for use in the distribution of gas. The first mains, laid by Winsor in Pall Mall, were of sheet lead bent to cylindrical form, and soldered at the edges. Murdoch's first mains were made of tinned iron and copper. In succession to these, pipes made of wood and asphalt, cement, concrete, bitumenized paper, paper impregnated with synthetic resin, slate refuse, brickwork and earthenware have been seriously suggested as suitable for use as gas mains.

In this connection, probably one of the most interesting exhibits at the Gas Exhibition held at Earl's Court in December, 1904, was a piece of 9-inch earthenware or fireclay gas main, similar to those shown in Fig. 22. This had formed part of a main which was laid for the Cambridge Gas Light Company, to whose engineer, Mr. James W. Auchterlonie, the writer is indebted for the photograph shown, and for the following brief description.

"As far as can be ascertained, the main, approximately 1,000 yards in length, was laid about sixty years ago * by Mr. John Grafton, a London engineer. The pipe is, roughly. 9 inches in internal diameter. The joints were all made with Roman cement. and the main was also coated internally to the thickness of about f-inch, and on the top and sides to the thickness of J-inch, with the same material. The blocks of which the main is built are about 12 inches long, and the bend in the foreground is in two halves, divided longitudinally. The hole on the right hand of the rule in the figure is about 1J inches diameter in the clear, and was probably used for the attachment of a service or small main connection. It is somewhat difficult to imagine that such a connection could remain gas-tight for a very long period. The ordinary cement joints of this piece of the main are quite good and sound, and have every appearance of being perfectly gas-tight."

The interest attaching to all the above descriptions of pipes, however, is now merely of an historical character, for none of them have stood the practical test of extended experience.

In a far different category must be classed the pipes associated with the name of M. Chameroy, which have been used very extensively on the Continent, and some of which the author saw being taken out of the ground when he was last in Paris. These were made of tinned sheet iron bent to the required size, and the edges strongly riveted and soldered. The pipes were then tarred and wrapped with coarse canvas, and further coated with a mixture of bitumen and sand to the depth of from J-inch to jj-inch, according to the size of the main and the local conditions. In the smaller sizes, male and female cast screws, composed of an alloy of lead and antimony, were soldered at the ends for connecting one pipe with another. Services of lead pipe are soldered to the Chameroy main and the portion disturbed re-asphalted.

* Written in 1906.

In some districts of the United States considerable use has been made of Tamarch logs, bored out to the required size, as pipes for the conveyance of gas. These have proved cheap and very durable, especially in wet and marshy soils. Pipes made of wood staves 1J inches to If inches thick, banded together with J-inch round rods, are even to-day being used in various parts of the States for water mains up to 36 and 44 inches in diameter.

At the present time, however, as far as the United Kingdom is concerned, we have to consider but two kinds of material for our main pipes, namely, cast-iron, and, under certain conditions, its young and pushing rival, steel. For all ordinary purposes cast-iron "holds the field." It is strong, durable, adaptable, easy of manipulation, and cheap. There is, however, one drawback, and that is its liability to fracture, especially in pipes of small diameter, owing to its inability to withstand heavy and suddenly applied loads. The recent great development in the use of steam road rollers, traction engines, and other heavy traffic along our thoroughfares, has, of course, greatly increased the possibility of such fracture taking place.

Improvement in cast-iron pipes.—As might have been expected, the enormous growth of the cast-iron pipe industry, during the past half century or more, has caused a great amount of time and attention to be given, on the part of ironfounders, to the improvement of the quality and workmanship of the goods placed upon the market. Whereas twenty or thirty years ago it was not at all unusual for pipes to be found of such a spongy character or so badly cast as to be almost, if not quite, porous, or so hard that it was very difficult indeed to drill and tap them, such things are almost unknown to-day. Going back a little farther still, it is of interest to note that in a discussion before the Institution of Civil Engineers in 1844 it was incidentally stated that the leakage through the pores and joints of cast-iron pipes was in some cases from 25 to 75 per cent. of the gas made. The quality of the iron to be used for pipes is of the utmost importance to the gas engineer. It should be best grey foundry pig iron and be dense, tough, perfectly homogeneous, of great tensile strength, not so soft as to be porous, yet not sufficiently hard as to crack when cut with a chisel, and easily worked with either file or drill. The casting should not be white or vitreous when cut or bioken. A solid bar of the metal of 2-inch by 1-inch section, placed on supports 3 feet apart, should be capable of sustaining a load of at least 28 cwt. in the centre, with a maximum deflection of 0-3 inch. As a test for tensile strain, a similar rod should be capable of sustaining a load of not less than 9 tons per square inch of section.


Fig. 22.— Photograph o( U-inch Earthenware Main excavated at Cambridge.

In consequence of the rather wide range of variation in the characteristics and grades of iron ore to be found in different parts of the country, it is practically impossible to specify at all minutely the chemical composition of the iron to be used. The following table, however, gives the composition of iron used by three of the principal pipe founders in the country.


The proportion of sulphur should be as low as possible. Phosphorus may vary from 1 per cent. to 1-50 per cent. For silicon the minimum limit should be 1 per cent., and it should not rise to more than 2-50 per cent. If under 1 per cent. the iron will prove too hard for the purpose, while if the 2-5 per cent. limit be exceeded the iron will probably be too soft and open in texture.

The great importance attached to the quality of the raw material applies equally to the method of casting. All pipes down to and including those of 3 inches in diameter should be cast vertically in dry sand moulds, from turned iron patterns. The reason for casting vertically is that, however great care may be exercised by workmen, there is always the danger that if the pipes are cast horizontally, or at an inclination, the metal may be unequally distributed through the section of the pipe, caused by a slight shifting of the core from its true position. It was usual at one time to cast all the smaller sizes either horizontally or at an angle; and frequently, when taking old pipes out of the ground, it is found that the metal is much thicker upon one side of the pipe than upon the other. Such pipes have often been the cause of imperfect connections, and consequential leakage, through the impossibility of obtaining sufficient depth of thread for a good attachment, when it has been necessary to drill the main upon the thin side. It is therefore usual to specify that the pipes in addition to being quite straight shall be truly circular in section with the inner and outer surfaces absolutely concentric. In order to prevent unequal contraction the pipes should not be taken from the moulds until after they have lost heat colour.

The pipes should be as smooth as possible, without sandholes, scales, blisters or lumps, and no filling in, or plugging, of faults should be allowed.

Test for uneven pipes.—A pipe in which the metal is thus unequally distributed through its section may be easily detected, by the simple device of rolling it upon a couple of level rails or lengths of wrought-iron tube. If it always comes to rest with one particular side of the pipe uppermost, it may reasonably be inferred that, when in that position, the lower portion of the pipe is the heavier, and consequently the thicker.

Pipes cast with sockets downward.—Although at some of the French ironworks the method is adopted of casting pipes with their sockets upwards, the exact opposite is the rule in the best of our English ironworks, and that for a very simple reason. When the pipe is cast with the socket downward, the latter, by reason of the weight of metal above it, is rendered very dense in solidifying. It is thus more capable of withstanding the great strain to which it is subjected in the process of jointing, and the risk of split sockets is reduced to a minimum. Few things are more exasperating than to have a


FiG. 23.—-Canting largo pipes at the Stavelcy Ironworks.

length of main well and truly laid, only to find the work rendered useless through split sockets evidencing themselves, entailing the additional expense of cutting out and rejointing, in addition to the accompanying inconvenience and delay.

Precaution to ensure density of spigot end.—It has been the custom generally for engineers to specify that pipes should be run with enough head of feed to ensure a sufficient density of the iron at the spigot end. Then any slag or impurity present, rising naturally to the top of the feed, would be removed with the surplus material when it was turned

off in the Lathe. This specification is frequently main tained even now. It is found, however, in modern practice, to be almost a superfluity of caution. By keeping a sufficiently high percentage of graphitic carbon, silicon and phosphorus in the mass, the molten iron may be made to run so easily that the slag and impurities present rise freely to the top of the ladle, and are not allowed to enter the mould at all.

Lengths of pipes.—A development which we owe probably, in the first instance, to the


Flo. 24.— Spigot end of 44-inch pipe being cut off in lathe.

American ironfounders has lately been made in the direction of casting pipes in greater lengths than was previously the case. Until a few years ago it was the common practice for pipes of 3-inch diameter up to those of 8-inch or 10-inch diameter to be cast in lengths of 9 feet. Those of 12-inch diameter and upwards were, and are still, cast in 12-feet lengths. The best pipe founders are now commencing to cast all sizes of pipes above 3-inch diameter in 12-feet lengths. The resultant economy in weight of material alone, in a long length of main, is very considerable. Suppose 1,000 yards of 8-inch main be taken as an example. If 9-feet pipes be used there would be 333-3 sockets in the length. If 12-feet pipes, the number would be reduced to 250, or 80-3 sockets less. These represent a weight of metal equal to 1 ton 11 cwts. 0 qrs. 14 Ibs., clearly saved in each 1,000 yards. In other words, as a socket is approximately equal to 1 foot of pipe in weight, there is a saving equal to 80-3 feet of main in each 1,000 yards run.

But this is only the commencement of the economy. It now (1920) costs approximately 6s. 4rf. to make an 8-inch joint in the usual way, with yarn and lead. There would, therefore, be a saving of 83-3 times 6s. 4rf. on jointing, whilst a less amount of excavation would be necessary for this purpose. And, in addition, the lessened number of joints would also imply fewer potential points of leakage, so that the unaccountedfor gas should be proportionately reduced.

Advisability of casting longer lengths in the larger sizes of pipes.—Unfortunately, the forward movement referred to appears to have spent itself for the present upon the smaller sizes of pipes. There is no doubt, however, that the increasing need of keeping down capital expenditure, and the other advantages to be derived, will inevitably act in the direction of causing engineers to specify longer lengths than those at present usually cast for pipes of 14-inch diameter and upwards. There seems to be no sufficient reason why the larger sizes of pipes should not be cast in 15-feet, or even 18-feet lengths. There would then be a reduction in the number of sockets necessary in each 1,000 yards run of main from 250 to 200 or 166-6, according to the length chosen, which would represent a very considerable weight of metal, besides the reduction in cost of jointing. If the saving thereby effected were multiplied out over a few years' main-laying operations of any large gas undertaking, it would total up to a very considerable sum.

The following table shows the economy accruing from the use of 12-feet, as compared with 9-feet pipes, worked out over 1,000 yards run of a few sizes of pipes as examples :—


Working out in the same way the saving effected by substituting 15-fcet lengths for the 12-feet lengths at present used, over three of the larger sizes of pipes, we obtain the following results :—

« PreviousContinue »