Museum of the City of San Francisco
By Subject
By Year
The Gift Shop

Report to the National Board of Fire Underwriters
on the San Francisco Conflagration.

from Engineering News
August 9, 1906

The conflagration which destroyed part of San Francisco, Cal., on April 18 to 21, 1906, has been studied for the Committee of Twenty of the National Board of Underwriters by Mr. S.A. Reed, Consulting Engineer, and a report of his study has now been submitted by the latter. The report contains much matter of great interest in fire protection, but as it is a document of considerable length we were able to reproduce here only brief extracts from it. Much of the descriptive information concerning the actions and results of the fire resides in the photographic views appended to the report, of which there are no less than 136, but for these as well as the full text we must refer to the report itself.

It will be remembered that the San Francisco conflagration began almost immediately after a violent earthquake, which occurred about 5:13 a.m., of Wednesday, April 18, 1906; the fire continued, as a conflagration, for about three days, and in that time swept over an area of 2,381 acres (=3.72 sq. mi.), which included the business district and a large part of the dwelling district. The property loss has been estimated as at least 80 per cent of the property value of the city before the fire.

The report of Mr. Reed gives a brief explanation of how the fire was limited or stopped:

At and within fighting range from the waterfront, the Inlet, and [Mr. John Center] Centre's Tank on the south, a successful defence was made on account of the ability to get water, the opportunity for safe retreat and for reinforcements, and finally on account of the assistance of pumps on vessels. The inland limits of the conflagration were determined by a variety of causes. First, on the south and southwest, by a windward position, and a thinning out to scattered low-lying frame construction. Second, on the west, near the line of Van Ness Ave., by a desperate stand made by the Fire Department in co- operation with men from the U.S. Army and Navy, with extensive dynamiting and back-firing. The essential causes, however, were, first a change of wind, and, second, the presence of water all along this line, from the Buchanan St. mains and from the military and naval hose-lines stretched 2,000 ft. from the Bay down Van Ness Ave., and supplied by the Quartermaster's tug. There were also numerous cases of individual defence by citizens which contributed to the check on the southwestern and southern borders.
It is readily understood that the work of fighting the conflagration , after the fire once got beyond control, was a forlorn hope, and had little relation to the termination of the conflagration. But even the earlier efforts of the Fire Department were ineffective because of the virtual destruction of the water distribution system by the earthquake. Another factor, also, the simultaneous outbreak of fires at many different points (not proved, but a reasonable conclusion), made the early defence difficult; yet, Mr. Reed says:
Judging from the successful work which was done during this period (the early hours of Wednesday – Ed.) on fires which occurred in places in the third [water district] services, where some water could be obtained, and having in view the lightness of the wind and its direction, it is a reasonable presumption that had water been abundant the Fire Department might have obtained control by noon of the first day.
This makes it of interest to note that conclusions are presented in the report as to the possibility of a system of water supply and distribution which will survive such violent attack as that of the April earthquake:
The real issue will continue to be whether any system of underground pipes can be relied upon to survive a shock equal to this one. If not, then the most perfect distributing system and the most reliable, abundant and powerful sources will be useless * * * Even a mere perfect system of gates, whereby crippled portions may be cut out, would not meeting the requirements of a panic-proof as well as earthquake-proof water service. During the first hour of the fire there were 80,000,000 gals. of water in the three reservoirs within the city limits, being rapidly drawn down, of course, by the breaks, but an any rate the point is clear that a shortage of water was not the difficulty There is no escaping the conclusion that even if San Francisco had had the most perfect water system that can be devised, under ideal management, and with conduits and reservoirs undamaged, yet the breaks of underground distributing pipes, even at a few times, occurring as they would all at once at the critical time viz., the outbreak of the fire, would have been followed by results not materially different. Saltwater mains would have been equally crippled. Steel pipe would not have survived better than cast-iron , breaks involving sheer drops of 3 and 4 ft. in the street grade. There is no doubt that the liability to extensive damaged may be reduced considerably by observing certain lessons of the event. Omitting, for the present, any consideration of apparently well- founded theories as to the existence of a fault or line of weakness running through the solid formation on a line from Lake Merced to and through San Andreas Lake, which line of weakness should be avoided by future conduits, and confining ourselves merely to considerations which are quite obvious from observations of damage done within the city limits – not only to water pipes, but also to sewer and gas pipes – it is clear that a large part of the liability to damage may be avoided by deep foundations for pipe lines where laid through soft formation. It will also doubtless be possible to make future house connections in a manner less liable to fracture in case of an earthquake. Still, in spite of all these possible improvements, there will be a residuum of liability whose seriousness cannot be ignored. The bleeding of water systems by broke house connections is a familiar difficulty during conflagration; but they usually occur progressively, as one building after another is wrecked by the spread of the fire, and it is conceivable that a highly organized gate manipulation would keep pace, to a certain degree, with the wastage of water; but the case is different where most of the damage occurs instantly. The particular difficulty of broke house connections can, of course, be met by a separate fire main system, but the latter would itself still be liable to break. The conclusion cannot be avoided that security on the subject of water supply reasonably approximating that of other cities, the regular pipe system will have to be supplemented by a supply of some kind independent of any underground pipe system.
It is clear, from this alone, that San Francisco must depend in large part on real fire- resisting construction of its buildings (and proper guarding of the inflammable contents) for its safety against conflagration. Therefore, if the lessons and recommendations which the great fire yielded have value anywhere, the have greatest value in San Francisco itself. But, of course, their importance to fire-protection is quite independent of any localized value.
Engineering News
Vol. LVI. No. 6
August 9, 1906

Return to top of page