FILTRATION AT ANDERSON
An erroneous idea seems to have gained ground that the use of sulphate of copper in purifying contaminated water is intended to supersede mechanical and sand filtration. That is not its intention. The best system of filtration guarantees only to remove ninety-eight per cent or ninetynine per cent of bacteria. So that, if there are 100,000 of these germs per c. c. present in the unfiltered water, 1,000 or 2,000 are left, among which may be some that come under the head of pathogenic. Under present conditions the results of filtration, in most cases, have been reckoned only from a quantitative standpoint, the one or two per cent of bacteria remaining being treated as a negligible quantity, the removal of ninety-eight per cent or ninety per cent of them being looked upon as the end to he obtained—all that could be secured. Under the best conditions of make of filters and methods of filtration, however, unavoidable accidents are liable to happen, any one of which may materially and harmfully add to the one per cent or two per cent of bacteria remaining after filtration. Such accidents may not he at once detected, until the death rate has increased so markedly as to cause attention to be directed to the condition of the water supply. The logical conclusion, therefore, to be drawn from such premises is, that chemical and bacteriological examinations of the water supply should be constantly made to see if it has become contaminated in the slightest degree. To carry out such a purpose intelligently a labratory and an expert analyst are necessary, such as are found at Anderson, Ind., where a 4,000.000-galloii New York Continental Jewell Filtration plant is installed, which is doing admirable work. These experiments were carried on at the time of the preliminary testing of the efficiency of the filters. The experiments were made by Severance Burrage, professor of Sanitary Science at Purdue university and Dr. S. C. Norris, city chemist and bacteriologist on behalf of the city, and C. Arthur Brown of the American Steel and Wire companv, ifl belial t of the J cwdl Filtration company The object was to undertake a series of experiments upon the effect of copper treatment in connection with mechanical filtration. For this purpose the water supply at Anderson offered exceptional opportunities for demonstrating the efficiency of copper in removing intestinal bacteria. The source of supply is the White river, into which the city of Muncie, twenty-five miles above, empties all its sewage. No falls or rapids intervene between Muncie and Anderson, and the river being icebound during the four weeks over which the test extended, there was formed a closed conduit for Muncie’s diluted sewage and that of some smaller towns situated still farther up the river. As a result of the presence of so much sewage, the water was very high in albuminoid and free ammonia, and exceptionally high in chlorin, owing to the salt water from the gas field above Muncie. It was low in turbidity, and the discoloration was slight. Owing to a small leak in the air wash-pipe—one of the unavoidable accidents already referred to—unfiltered water made its way into the pipes within a fortnight after the test began, and a valve also stuck, because raising the amount of iron sulphate proved too much for the filters to accommodate—another unavoidable accident—and no iron or copper was introduced into the water. Contaminated water, therefore, passed the filters, and, at the next washing, the clear well again contaminated the filters, near enough to the time of sampling to show bacillus coli in two of the samples. This was fortunate for the thorough testing of the value of copper, as for a short time the filters were unable to effect a high reduction of bacteria. The number of these bacteria in the river ranged at irregular intervals during the four weeks’ test from 13,000 to 155.000 per cubic centimetre—usually remaining above 50,000. The number of bacteria in the filtered water varied between 15,000 and 400 per cubic centimetre—usually remaining above 3,000. For ten days—from February 2 to February it. 1900—alum was used as a coagulant; but with such water as that at Anderson, it was not possible to get a proper coagulant with from one to three grains of alum per gallon, with or without the addition of lime. Racil1 us coli was always present in the river water and in the filtered water, and was identified by gas formation, reduction of neutral red, proportion of carbon dioxide, growth on milk, potato, gelatine, litmus agar and formation of indol. On February 11, instead of aluminum sulphate, iron sulphate containing one per cent of copper sulphate was introduced in quantities of one grain and one-half per gallon of water (0.015 grain copper sulphate per gallon—approximately one part of copper sulphate to 4,000,000 parts of water). Lime was added, two grains per gallon. The treatment was continued for four days, and during that time only once was there any indication of the presence of bacillus coli in fermentation tubes inoculated with one cubic centimetre samples of filtered water, and this occurred immediately after a leak developed in the air-pipe of the washing system that allowed unfiltered water to pass into the pipes. On the 15th of February, the amount or iron sulphate containing one per cent of copper sulphate was raised to four and onehalf grains per gallon. This quantity was found to be too great, and at midnight the amount of the coagulant was reduced to two and one-quarter grains—(0.0225 grains sulphate copper per gallon -approximately one part of copper sulphate to 2,500,000 gallons of water). The valves controling the iron solution caught on this change, and for nearly an hour no iron or copper was applied to the raw water. This allowed polluted water to reach the clear well, and at the next washing the filters were again contaminated, and samples from two of them developed typical bacillus coli. The following five days bacillus coli developed but once, and this was immediately following the reappearance of the leak in the air-pipe. Iron sulphate containing one-half per cent copper sulphate was now applied at the rate of 1.5 grains per gallon—(0.0075 grains copper sulphate per gallon—appoximately one part of copper sulphate to 8,000,000 parts of water). Bacillus coli was eliminated during the two days that this mixture was used. Iron sulphate was substituted during one day containing only onefourth copper sulphate (0.00375 grains copper sulphate per gallon—approximately one part of copper sulphate to 15,000.000 parts oi water). Tins amount was insufficient to eradicate completely the bacillus coli, and two of the filter samples contained this organism. Pure iron sulphate was then used at the rate of three grains per gallon, and bacillus coli developed from samples of each filter. Iron sulphate containing one per cent copper sulphate was substituted for the greater part of the pure iron salt, and in the following samples (as shown in the accompanying table) no bacillus coli developed in the filtered water. From the preceding experiments there seems to be no doubt that the filtering of polluted water of this character through the layer of coagulum of iron and copper which forms on the filterbed brings the bacteria borne in the water into contact with the precipitated copper for a sufficient length of time to destroy bacillus coli, and, as bacillus typhi is still more sensitive to the action of copper, it, too, must necessarily be removed from the filtered water. These tests seem to meet the objection to the use of copper sulphate in water supplies—namely, that there is a considerable chance of appreciable amounts of copper reaching the consumer. Considering the harmlessness of copper, this is a theoretical rather than a practical objection, and is answered in the present instance. The copper is all precipitated and the insoluble coagulum of iron and copper is caught upon the filters. It may he added that Karl F. Kellerman, of the Bureau of Plants, Washington, D. C, was deputed to investigate the test at Anderson. He arrived there before the test was begun and remained until it was concluded. Dr. George T. Moore, late of the same bureau, from whose researches in and experiments on the treatment of water with sulphate of copper so much has resulted, also visited the plant during the test.
The filtration plant was installed by the New York Continental Jewell Filter company. It consists of four filters, each thirty-five feet by ten feet. They are of concrete, and their capacity is 400,000 gallons. Notwithstanding the fact that the water of the White river is contaminated by sewage from Muncie and other cities and towns above Anderson, the water, which is taken from a point one mile and one-third above the city, when distributed to the consumers is now practically free from bacteria. The plant embraces all the essentia! improvements and appliances that have resulted from recent experiments in filtration, and contains substantially all the necessary appliances and machinery used in the filtration plant at Little Falls, N. J. Before it was decided to filter the water, a new source of supply was sought. Deep wells were suggested, but were found inadequate and impracticable. The Kilbuck stream was then suggested: but the expense rendered that scheme prohibitory, a further objection to it lying in the fact that the water so obtained must be filtered. The White river was then fallen back upon again as a last resource. Chicago analysts and those of the Indiana State board of health, who had investigated the water under the most unfavorable conditions, reported that, impure and polluted as it was, it could be made fit for use. It was, therefore. resolved to filter it. Sand filtration was thought of: but, as its initial expense was so very heavy, the idea was abandoned, and mechanical filtration was decided upon. Bids were received from four concerns, that offered to supply such a plant. Their bids were as follows: $16,500; $16,690; $24,000; $27,700 (the New York Continental Jeweli Filter company). One proposition was from a company that had no municipal plant in operation within reasonable distance of Anderson that could be inspected. A second was from a company that had been in business only a few years, and whose plant, after it had been examined by Anderson officials, was not considered suitable for such water as that city had to deal with—being reasonably satisfactory for mountain water where a coagulant is hardly ever necessary, but not adequate to filter that of White river. I11 constniction it was cheap and simple; hut it omitted many devices and appliances essential to a complete plant. The proposition of a third was for a new and untried system, which had never been constructed or used by any municipality. In the inventor’s own city it had not been approved, and it had been rejected, after having been experimented with, under most favorable conditions, in two other very large and important cities. The proposition of the New York Jewell filter company was in every respect the most suitable, and, because it was so complete and so thoroughly up-to-date in every way, was chosen, in spite of the fact that that company’s bid of $27,000 was the highest. Incidentally it may be noted that the citizens of Anderson have publicly ‘ expressed their full persuasion that rigorous steps should, and will “soon be taken by the legislature and courts to prevent the pollution of streams, and that there is a growing disposition all over the country to compel cities and industrial institutions to disse f their sewage and polluted matter othcrwis than by dumping it into streams.”