The time required to completely expel the volatile constituents from the charge in a gas retort varies very much, because there are great diversities in the composition of the different kinds of coal employed. Some varieties of coal, such as cannel, are easily decomposed, and the operation may be complete in about three hours; while other kinds may require double that time. The quantity of gas procurable from a given weight of coal also varies according to the kind of coal made use of. Thus, while a hundredweight of cannel may give 430 cubic feet of gas, the same weight of Newcastle coals will yield but 370 cubic feet. The nature of the gases given off from a retort will be different at the different stages of the operation.
The scene presented by the retort-house of a large gas manufactory, when viewed at night, is a singular spectacle. The strange lurid gleams which shoot out amid the general darkness as the retorts are opened to withdraw the coke, and the black forms of the workmen partially illuminated by the glare, or flitting like dark shadows across it, form a picture which might engage the pencil of a Rembrandt. In Fig. [348a] is depicted the retort-house at the Imperial Gas Works, King’s Cross. Here the retorts are arranged in several tiers—the coal being brought, and the coke withdrawn, by the aid of an iron carriage running on rails parallel to the line of furnaces.
Fig. 348.—The Retort.
In the process of heating, a proper regulation of the temperature is of the highest importance. It is found that when the retorts are heated to bright cherry-red, the best results are obtained. At a lower temperature a larger quantity of condensable vapours are given off, which collect in the gasholders and distributing pipes as solid or liquid, and occasion much inconvenience, while the quantity of gas obtained is decreased. On the other hand, if the temperature be too high, some of the gases are decomposed, and the quantity of carbon contained in the product is so much diminished as seriously to impair the illuminating power. Again, every second the gases after their production remain in the red-hot retort diminishes their light-giving value; for those hydro-carbons on which the luminiferous power of the gas depends, are then liable to partial decomposition; a portion of their carbon is deposited on the walls of the retort in a dense layer, gradually choking it up, while the liberated hydrogen does not add to the illuminating but to the heating constituents of the gas. A plan has been patented by Mr. White, of Manchester, for rapidly removing the illuminating gases from the retort by sweeping them out by means of a current of what has been termed “water gas.” This water gas is produced by causing steam to pass over heated coke, and is a mixture of carbonic acid, carbonic oxide, and hydrogen. Though only two of these are combustible gases—and even they do not yield light by their combustion, and, by adding to the bulk of the gas, serve rather to dilute it—yet it has been found that in some cases twice the amount of light is obtainable by White’s process than the same weight of coal supplies when treated in the ordinary manner.
Fig. 348a.—Retort House of the Imperial Gas-Works, King’s Cross, London.
The hydraulic main, as already mentioned, being kept half full of tar into which the lower ends of the dip-pipes descend, prevents the gas from escaping through the stand-pipes when the lid of a retort is removed for the introduction of a fresh charge. The hydraulic main is from 12 to 18 in. diameter, and the dip pipes pass into it by gas-tight joints. Various forms of purifiers are in use besides the simple one already mentioned. Some of these have arrangements for agitating the gas with a purifying liquid by mechanical means, the motion being supplied by a steam engine.
The gasholder, as it sinks in the water of the cistern, presses with less force on the contained gas, and unless this inequality of pressure were counteracted there would be very unequal velocities in the flow of gas from the burner. The equality of pressure is obtained by making the weight of the chains by which the gasholder is suspended equal to half the weight the gasholder loses in the same length of its motion. Gasholders are also constructed without chains or counterpoises, as these are found to be unnecessary where the height of the gasholder does not exceed half its width. In such cases, especially when the vessel is very large, the difference of pressure at the highest and lowest position is quite inconsiderable, and nothing more is necessary than that upright guides or pillars be placed to preserve the vertical motion of the vessel. Another improvement, which enables a lofty gasholder to be used without increasing the depth of the tank, consists in forming the gasholder of several cylinders, which slide in and out of one another like the draw-tubes of a telescope. Each cylinder has a groove formed by turning up the iron inside the rim, and at the top of the next cylinder the edge is turned outwards so as to drop in the groove or channel, which thus forms a gas-tight joint, for it is of course filled with water as it rises. The pressure is, however, more accurately regulated by an apparatus called the governor, through which the gas passes in before it enters the mains. The construction and action of the regulator will be understood from Fig. [349], where A represents a kind of miniature gasholder, inverted in the cistern, B. From the centre of the interior of the bell hangs a cone, C, within the contracted orifice of the inlet-pipe. If this cone be drawn up, the size of the orifice, D, is reduced, and, on the other hand, by its descent it enlarges the opening through which the gas passes outward. By properly adjusting the weights of the counterpoise, E, such a position of the cone may be found that the gas passes into the mains at an assigned pressure. Suppose, now, that from any cause the pressure of gas in F increases, that pressure acting upon the inverted bell, A, causes it to rise and carry with it the cone, which, by narrowing the orifice of the outlet, checks the flow of gas. Similarly, a decrease of pressure in the mains would be followed by the descent of the cone, and consequently freer egress of gas. In hilly towns it is necessary to fix regulators of this kind at certain heights in order to equalize the pressure. It is found that a difference of 30 ft. in level affects the pressure of gas in the same main to about the same amount as would a column of water one-fifth of an inch high, the pressure being least at the lowest point.
