SPRINKLERED MILL FIRES
During the last twelve months or more there have been more destructive fires than usual in sprinklered buildings, proving the point that has always been made by all—except the underwriters, that sprinklers are by no means infallible preventatives of fire. In some of the fires that have occurred the human factor has not been taken into consideration, and the sprinkler-apparatus has failed to act, either because “some one had blundered” or because it had been allowed to g_____t into a state of disrepair, or for some other reason the work that ought to have been performed by them has been left undone or done in a most unsatisfactory manner. One or two instances may he ouoted. The first is that of a paper null at Marietta, Ga. It was a threestory, brick, of huht jointed construction, with frame storehouse adjoining, cut-off, Intt with indifferent fire-doors. lhe storehouse was tilled with waste paper, cordage and pulp raw stock. The fire started in a pile of scrap paper in the storehouse and was not held in check by the sprinklers. It spread through door-openings to the adjoining mill and was finally checked by the sprinklers in the mill and private host streams. The risk was equiped throughout with Grinnell and Hibbard sprinklers, various pipe sizes and a fair equipment, but not quite up to the modern standards. The primary supply was the town waterworks 5-in. connection to sprinklers from 6-in. main, fed one way, the dead end being over 1,000 ft. away. There were hydrants on this connection, also, two 4-in. service connections, the normal pressure being about 60 lbs. The sprinklers in the stock house where the fire started were on a dry system—Grinnell dry-valve. It was stated to have been in service at the time of the fire, with the gate-valve open. It appears that the drip-check in the intermediate chamber drip-opening was found upside down. This would allow the water to accumulate in the intermediate chamber and pass up through the valve into the system. The pipe above the dryvalve was directly exposed, and the temperature for three nights previous had been below freezing. A fair conclusion is that the water leaked into the system, and at the time of fire the sprinklers did not operate, probably because of ice in the pipes. The condition of the room where the fire started was not favorable to cpiick sprinklor contr. 1. The room was filled with wastepaper stock, and it is quite likely that the sprinklers could not at once extinguish the fire, and a large number of sprinklers may have opened, overtaxing the town water supply. The second supply was a i,ooo-gal. steam pump, which, although not in the best of condition, did excel lent service and undoubtedly was the means by which the plant was saved from total destruction. The pump draughted from a large cistern, and it is figured that about too,000 gals, were used. After the fire the fire-pump was tested, and it was found that two of the dischargevalves had liven lost, the stems being of a poor design and having worked loose. Two stems and the rubber disk were found lodged under the water-seat in the dry-valve. The lack of alarm service may have had considerable effect in that the fire was not promptly discovered. The following are the points of interest brought out by the fire: (1) Congestion and the character of the stock; (2) a weak primary supply, not an automatic secondary sttpplv to sprinklers; (3) the possibility of crippling a dry system by misplacing a drip check-valve; (4) lack of the sprinkleralarm service; (5) the lack of the private fire brigade, no town fire-alarm service, inefficient volunteer fire department; (6) unprotected floorooenings; (7) weak construction; (8) indifferent fire-doors, highly resinous wood, insufficient thickness, improperly clad, only single and not properly automatic, no sills, etc.; (9) poor type of fire-pump valves. Another failure was in an Amsterdam. N. Y., felt mill where a fire occurred in a two-story frame stockhouse. It started in a bale of jute, was discovered by an employe and was under good headway at that time. The storehouse was equiped with Grinnell glass disk sprinklers, issue “A,” dry system, with twenty h -ads on a i}4-in. pipe, spacing 10 by 10 ft. under open joints. The sprinklers were supplied by the waterworks at 72 lbs. pressure. The sprinklers failed to hold the fire in check, probably because of the small pipe sizes and generally defective equipment. A third lire was in a woolen mill at Louisville. Ky., a substantial brick structure, with stone base and metal roof. The internal construction was also substantial, with extra-heavy watertight floors; two stories were above the grade of the cornice, and three below. A retaining wall divides the first subbasement at the front, north end, which extends up to the fleor of the grade-floor, with one unprotected opening. The building was thoroughly sorinklered with two Grinnell dry-pipe systems, each quite up to the standard of 1904. This equipment is supplied by two 4-in. city connections, one on each side of the plant, which connect with large mains. The city pressure is by gravity and furnishes 55 to 60 lbs. prssure at the level of the sprinklers immediately over the point of the origin of the fire. There is, also, an additional supply from a 5,oco-gal. gravity tank, elevated 9½ ft. above the highest portion of the plant. The equipment was examined by the sprinkling inspector of the board on the Mon-‘ day preceding the fire and found in good condition and working order. The origin of the fire is unknown., Two alarms were turned in. One of the lire engines was coupled up to the Siamese connection; but by order of the chief no water was put into the sprinkler system from the engine until the seat of fire was discovered about threequarters of an hour afterwards. The engine pumped easily and freely into the system, and it was kept working until about 5 o’clock a. m., when the department thought the fire sufficiently under control to retire a number of streams and to cut out the sprinklers, including the auxiliary engine. The outside valves controling the two inlets were then also closed by the engineer of the plant under instructions from the fire department. He first shut off the post-indicator valve in the yard, thus cutting out both the city supply coming into the equipment from Frankfort avenue, and the tank supply; and almost immediately afterwards he shut off the second valve on Story avenue under the personal supervision of the assistant chief of the fire department and the captain of the salvage corps. It would, therefore, seem that, so far as the sprinkler system is concerned, the supply-valves were open, and it is evident the dry-pipe valves controling the two risers were also open, as, otherwise, the engine could not have pumped into the system, as it did freely and smoothly. The fact that the fire did not ascend the elevator-shaft, which is in close proximity to the point where the fire occurred, is an additional evidence that the sprinklers must have operated. The number of heads open were as follows: In the first subbasement, where the fire occurred, ninety-one heads, comprising the entire number on this floor, opened; on the basement, 117 heads; on the first floor, eighty-six heads; on the second floor, thirty-one heads—a total of 325 heads out of a grand total in this division of 756 heads. That the fire was able to gain the proportion it had evidently acquired at the time of its discovery is . something that has baffled even conjecture, when it is considered that the amount of combustible material stored at the point involved was comparatively small, and that the sprinkler system was also apparently in good and effective working condi tion. The area covered by the fire was relatively small, although it would app ar that at the point of conflagration it must have been unusually fierce, for the reason that it burned through heavy joists, causing the collapse of the floor above, and had also charred the supporting posts in close proximity to a depth of from I to 1 ½ ins. Other detailed reports on this fire tend to the conclusion that the water must have been shut off the sprinklers at the time the fire started and turned on later, the conditions near where the fire started apparently proving conclusively that no water issued from the sprinklers until the fire had attained considerable headway. There was a large gravity tank supply which it is stated was not used and the tank was found full after the fire, showing that this supply at least was not in service. According to the report of the adjustment committee, the physical evidence is such as to impress the committee along the lines that these sprinkler-heads, while they may have opened, discharged no water up to the time the engine-connection was made from the street. At a certain period of the fire (and it is felt that this was after the time the engine-connection was made to the sprinklers) the equipment did fine work.