« PreviousContinue »
ful scrutiny of the records of the water company with respect to the operation of the apparatus for the application of the chlorine to the supplies did not disclose any interruption in the chlorine process since December 11, 1916. Frequent analyses made by Professor G. Č. Hodges of the water company indicate that the waters as delivered to the city have been of a reasonably satisfactory sanitary quality. With all other possible sources of infection excluded, these facts indicate that the water supply from the Marcy reservoir received infection within the mains at some point within the city limits and in all probability in or near that district in which the cases have occurred.
As soon as this localization of the infection was indicated a careful study was made to determine the possible point or points at which infection could reach the water in the city mains. From the records of the water company it appeared that no mains had been opened in this section of the city since the spring or summer of 1916. One hydrant, however, had been thawed out in the middle of January and a dead end had been extended in the southwestern part of the district in the middle of December. The location, dates and methods employed in this work were such as to preclude probable infection at these times. The next point into which inquiry was made was that of the presence of auxiliary mill supplies from polluted sources for fire protection or other purposes and cross-connection with the city mains. Inquiries in this matter disclosed the fact that the Globe Woolen Mill located at the corner of Stark and Court streets in the northeastern section of the infected zone was supplied with such an auxiliary system and that during the past summer and fall extensive alterations had been made in the system. A further inquiry into the nature, character and extent of these alterations disclosed the following facts:
For several years this mill has not been in operation, but has been recently acquired by the American Woolen Company. Upon resumption of operation of this mill the insurance association required radical changes in the hydrant and sprinkler system of the mill. The survey of this mill by the insurance inspectors showed a complicated system of water pipes, hydrants, sprinklers, service pipes for general mill purposes and the city mains interconnected. For auxiliary fire service a 1,000-gallon-per-minute plunger pump was provided, obtaining its supply from a pond on the grounds of the mill, this pond being fed by Nail creek, a stream flowing from the old Chenango canal through a thickly settled portion of the city. With the old system of piping it was possible that with the fire pump in operation water from this creek might be pumped back into the city mains, as at that time there were no check valves provided between the two systems to prevent such an occurrence. About the 1st of September, 1916, work was started under the direction of the insurance association to separate the service lines of the mill from the sprinkler and hydrant system. This work continued until the second week in November. All mains supplying water for general mill purposes were disconnected from the hydrant and sprinkler system, although two connections were retained between the fire system and the city mains, one connection being from the 12-inch main in Stark street and the other from a 6-inch main in Court street. These connections, however, were provided with two gate valves between which were placed single check valves designed to shut off the city supply at such times as a greater pressure might be produced by the fire pump in the mill. Upon the completion of the work a test was made to determine the tightness of the hydrant and sprinkler lines as well as to indicate the positive action of the check valves. At the time of this test, November 16, 1916, a pressure of about 135 pounds was maintained for about 1 hour with the pump operating at a rate of about 3 revolutions per minute. This slow rate of operation indicated to the insurance inspector that the system was tight and that the check valves were operating. The pressure in the city mains is about 90 pounds per square inch, so that there was a pressure of about 45 pounds, tending to keep the check valves closed. Since this test on November 16 there is no evidence that this fire pump was again in operation until January 31, 1917. It should be noted, however, in this connection that
INCIDENCE OF TYPHOID FEVER AT UTICA. CASES PLOTTED ACCORDING TO DATES OF ONSET. during November, particularly during the latter part of the month, a great many cases of diarrhea occurred suddenly and intensively in that section of the city in which the typhoid fever cases later occurred. In view of the fact that the hydrant lines in this mill had been recently opened and in some cases relaid in trenches containing dirty water and in such manner as to give opportunity for contamination in some of the pipes, and in view of the seriously contaminated condition of Nail creek, it seems quite possible that local contamination of the water supply was received at this point in November. It is well known that check valves are extremely unreliable in operation, especially when the pressure tending to close them is comparatively small and also on account of the ease with which the valve is kept from closing by particles of gravel or bits of wood catching between it and the seat. In this particular installation the cross-connection of the two mains of the city by means of the fire system in the mill renders probable the more or less constant opening and shutting of the check valves due to varying and indeterminate changes in pressure in the city mains on the two street connections, such change in pressure tending to set up a flow of water of short duration through the hydrant system of the mill. It is very easy to see that in this manner greater or less amounts of polluted water from the hydrant system might be admitted to the city mains.
There are numerous cases on record where the failure of check valves designed to separate polluted auxiliary supplies from the city supply have failed to operate properly with resultant outbreaks of typhoid fever. Amongst the better known of such cases are those at Lowell, Mass., Philadelphia, Pa., and Auburn, N. Y. In Utica itself several years ago a small outbreak of typhoid occurred among the employees of the International Heater Company due to the fact that a valve separating a polluted supply from the city supply was either open by accident or was leaking. Another case of the unreliability of check valves also occurred in Utica. During the summer of 1916 a representative of the Underwriters' Association found that the check valves in one of the mills of the eastern part of the city were fastened open.
Subsequent to the collection of this information regarding the Globe Woolen Mills it was deemed advisable to make an investigation of other mills in the western part of Utica in order to determine the extent to which auxiliary fire systems were in use in this portion of the city, this investigation being made by Mr. C. M. Baker, assistant engineer of this Department. For convenience the facts of this investigation are summarized in appended table B.
The facts given in this table show that such auxiliary supplies and crossconnection occur to considerable extent and it should be pointed out that in every case where auxiliary fire supplies from polluted sources are separated from the city mains by check valves only there is every possibility of contamination of the water in the city mains and possibly infection.
Various samples of the city water were collected at the time of the investigation as well as samples of water from Nail creek, the Erie canal and of ice from the Erie canal. The results of the analyses of these samples made by the Division of Laboratories and Research are shown in the appended table. Subsequent to the outbreak of diarrhea in November several samples of water were collected in the western section of the city and the results of the analyses 'of these samples are also given in the appended table. On December 1 several samples were also collected in this district by Professor Hodges and the results of his analyses indicate considerable contamination of the water. The analyses of samples collected on December 8 also indicate active and possibly local contamination of the supply, and the analyses of samples taken at the time of the recent investigation indicate also a small amount of active contamination of the water in this district.
To summarize, it may be stated that the investigation has eliminated contact (with four exceptions), carriers, raw foods and milk as agents in the causation of the outbreak, leaving therefore only the locality of residence and the public water supply as common factors in the majority of the cases. The fact that those sections of the city receiving water from the Deerfield
and Southern reservoirs, as well as certain sections receiving Marcy water, have been practically free from cases of typhoid, indicates a local infection of the water in the mains, analytical results having shown that active contamination has actually occurred. Several mills in or bordering the infected district have auxiliary water supplies from polluted sources separated from the city mains by check valves. In view of the unreliable character of such check valves it seems clear that these auxiliary supplies are the most probable source of the infection of the water in the inains. It is impossible, however, with our present knowledge to state definitely the particular mill or the particular time from or at which the infection was introduced into tlie mains, owing not only to the lack of information on the part of the mill owners but also to the fact that information of this kind is very difficult to obtain on account of the tendency on the part of those in authority not to assume responsibility. The entire epidemiological evidence, however, points strongly to localized infection of the city water, and the facts relating to the existence and operation of check valves between auxiliary mill supplies from polluted sources leave no doubt as to these cross-connections being the indirect means through which the infection occurred. I would therefore recommend :
1. That the city and water company's authorities make a careful and thorough inspection of all mills and other establishments in the city to determine the presence of all cross-connections between private supplies from sources other than those of the public supply.
2. That steps be taken, as this investigation progresses, to correct all conditions which may tend to allow the introduction of polluted water into the city mains through such connections.
3. That, in order to determine the presence of contamination from these or other sources, the water company make a survey by means of laboratory analyses of samples taken from various points in the distribution system.
4. That the proper city authorities pass an ordinance requiring the complete physical severance of all auxiliary supplies from the general city supply, except in those cases where the sources of such auxiliary supplies have been found to be safe against the possibility of contamination.
It is believed that the rearrangements of sprinkler and hydrant systems necessary to carry out this fourth important recommendation can be made without excessive cost, without increasing the fire risk and without impairing the efficiency of either the auxiliary or the city services.
Chief Engineer ALBANY, N. Y., March 1, 1917