Downcomers are generally provided with explosion doors for two purposes: First, when furnaces are stopped it is almost impossible to prevent air being drawn into the gas main at some point or other. This air mixes with the gas and produces an explosive mixture. When operation is resumed this mixture is ignited at the boiler or stove burners and an explosion follows which in some cases is violent enough to burst the main, dust catcher and downcomer unless it is relieved by suitable safety-valves or explosion doors.

Second, the furnace itself is liable to slips or explosions in which a pressure of several pounds to the square inch, ordinarily only occurring at the lower levels of the furnace, reaches the top and would do it great damage unless properly relieved.

This later theory has been combatted by Mr. Julian Kennedy who has claimed that explosions were more violent in proportion as they received freer vent. He compares the conditions which arise at such times to those occurring when the cork is removed from a bottle of charged water; if this be done suddenly a great proportion of the contents of the bottle is violently ejected, but if the cork is held back and the gas pressure allowed to pass off slowly none of the contents of the bottle are thrown out. In accordance with this theory the tendency in recent years has been to provide many less explosion doors than were formerly thought to be necessary, and the results have very largely confirmed the theory, though many furnacemen are not willing to go as far as advocated by Mr. Kennedy and do away with explosion doors entirely.

One design, made about the time of the Duquesne revolution, was intended to be the best possible regardless of expense. Some five or six outlets were provided around the top of the furnace and these were all connected with an annular main, from one or two points on which the branches of the downcomer led off. Explosion doors were arranged on the outside of this annular main, which, of course, was a few feet below the operating platform of the furnace top, and in order to provide access to it a supplementary platform was built a few feet below.

This system had almost all of the disadvantages which could be worked into one design. On the side opposite to the downcomer there was so little current of gas that the annular main became sufficiently choked up to make it virtually useless and one or two outlets on the side next to the downcomer carried off all the gas, practically as though the annular main did not exist. On account of the large size and distant location of the dust catcher and the wide stretch of its branches to reach the opposite sides of the annular gas main the slope of the downcomer was exceedingly flat and it is probable that after it was in operation a short time more than half its area was choked up and not in active use. One of the worst features of the design was the location of the explosion doors which, it will be seen, came just under the main operating platform. Owing to the choked condition of the downcomer the top pressure was high and the explosion doors were generally leaking more or less; if the gas caught fire it burnt underneath the top platform so as to damage the steel work, and if it escaped unburnt the gas came up onto the top platform in sufficient quantity to make it dangerous to stay there any considerable length of time.

It may not be amiss to say a word here as to the general design of explosion doors. These are always made hinged, but are located in two different ways; first, so that the plane of the door is horizontal when closed; second, so that it is steeply inclined or vertical when closed. The latter design is radically wrong and should never be used because the moment of the door due to its weight is little or nothing when it hangs vertical, consequently a slight pressure is sufficient to force it from its seat, but as its moment rapidly increases a much greater pressure is required to force it to open wide, as it should when it is needed at all.

Fig. 178. Elevation of Bethlehem Steel Company's furnace showing correct design of gas outlets.

The horizontal door works exactly the opposite way, exerting its maximum moment when it is closed, so that it is very easy to keep tight under ordinary conditions, but when a considerable increase of pressure occurs, forcing the door open to relieve it, the moment of the door decreases as it opens so that it operates more and more easily as it opens wider. This seems at first sight to be an insignificant detail, but as a matter of fact I have seen great annoyances caused by vertical doors, so little gas leaking from them can be objectionable either by burning and damaging the structure or by polluting the air, that the final result is to have a rigid clamp put on the door which virtually destroys its usefulness for its primary purpose, even though it may not completely eliminate the leakage.

Fig. 178 shows a design of gas outlets and downcomer which secures practically all the important advantages. The drawing shows the "E" furnace of the Bethlehem Steel Company. The outlets are four in number rising from the conical top of the furnace shell. The down-comer instead of connecting directly with the outlet openings on the furnace shell on the back or high side connects with the outlet pipes at a point several feet above the operating platform and thereby secures sufficient height to give it a slope steep enough to free itself. The explosion doors are placed on the top of these high outlet pipes many feet above the operating platform and if they leak no harm is done. If the gas burns it cannot strike anything to destroy and if it remains unburnt it pollutes the air far above the heads of those who may have occasion to be upon the operating platform.

In the early days the downcomers were made of quite thin pipe entirely unlined and these gave remarkably good service considering the relatively high temperature of the gas they sometimes had to handle, temperatures of 800° or even 1000° Fahr. in regular operation being not uncommon, while, of course, very much higher ones occurred during slips.

In modern practice, however, the downcomer is always built of very substantial steel plate and lined with circle bricks from 3 to 4 1/2 in. thick. This makes an almost indestructible construction but a very heavy one so that due care must be taken to support the weight of the downcomer at both ends. More than one case is on record in which no better support had been provided than the conical top of the dust catcher and this has buckled under the great weight thrown upon it.