Page:The New International Encyclopædia 1st ed. v. 20.djvu/419

* WATEB PURIFICATION. 349 WATER PURIFICATION. admitted to the water before it comes to the filter and is given from a few minutes to several hours' time to act before filtration takes place. ■■■flevolvinaf Shcrft Sancf Agitator SECTION OF A MECHANICAL FILTER. If the period is brief a coagiilating chamber is placed directly beneath the false bottom of each filter tank ; otherwise one or at most two larger coagulating basins, or settling reservoirs in which a coagulant is used, are emplo.ved. The coagu- lant is made of the desired strength hj proper dilution. The quantity a])plied to the water be- ing treated is projiortionetl by various ingenious devices to the changes in rate of pumpage or water consumption. Among these may be men- tioned some form of pump, driven or controlled by the (low of water through the supply pijie to the basin or filter. The amount of coagulant used varies from about one-half to three grains per gallon of water, but rarely exceeds one grain except in waters with high turbidity or that have some other unusually troublesome feature. The filtering material used in mechanical filtration is generall_v confined in wooden or steel tanks which give a filtering surface with a diameter of only some 10 to 16 feet. The depth of sand is about the same or a little deeper than in slow sand filters, but the rate of filtration runs from 90,000,000 to 125,000,000 gallons an acre. The tanks and operating meclianism are almost in- variably placed in a building, and are therefore quite free from the extremes of heat, and from rain or snow. Under proper conditions of de- sign and operation, mechanical filtration will remove almost if not quite as many of the bac- teria as the slow sand process; more turbidity and color: but less dissolved organic matter. It is more commonly applied to turbid, highly col- ored waters than to those which are objection- able on account of sewage pollution. In first cost and in labor charges, mechanical filtration has the advantage, but this is largely and some- times wholly ofTset by the cost of the coagulant, of operating the machinery and providing wash water. If the slow sand filters have to be cov- ered the capital charges may be so increased as to place them at a financial disadvantage, as com- pared with mechanical filtration, although cov- ering filter beds often materially reduces their cost of operation. The choice between the two systems depends upon such a variety of local conditions that each case should bo most care- fully considered on its merits before adopting either system. lIiSToiiicAL Sketch. The history of filtration, as applied to public water supidies, dates from 1829, when James Simpson built some filter beds for the Chelsea Water t'ompany, of London, England. Following that action, slow sand filtra- tion was first gradually, then rapidly adopted in England and on the ((intiiicnt of ICurope. It was not until alnmt ISS7 that the real action of slow sand filters was understood. I'rior to that time they were supposed to be strainers only, effecting little or no other chemical change upon the water and leaving the bacteria unliarmed. In fact, it was not so many years before this that the relation of bacteria to disease ( see Disea.se, (jEHM 'I'liEORY OF), and particularly of water- b(u-ne germs, was established. The perfection by Koch of methods applicable to the enumeration of wafer bacteria made possible the announcements by Percy Frankland and others aliout 18.S7 that filter beds renioved nearly all tlic bacteria in water. England. (lermany. and America have each played imjiortant parts in the scientific and practical development of the various aspects of the bacterial purification of wafer. The work in America has been done very largely by the Massachusetts State Board of Health, at its Lawrence Experiment Station (see annual re- ports, 1887 to date), as has been mentioned under Sewage Disposai,. At Lawrence slow sand filtra- tion has been studied in great detail and a mass of scientific data has been accumulated and valuable conclusions drawn therefrom. Other American investigations, begun a number of years later, are mentioned farther on. The largest slow sand filtration plant in the United States treats a part of the water supply of Albany, N. Y. It was designed primarily to remove bacteria from the badly polluted water of the Hudson River, and secondarily to remove turbidity and efTect such a reduction of color as might be feasible. The plant was completed in 1899 with Allen Hazen. of Xew York, as engineei'. Centrifugal pumps lift the water 24 feet or less, according to the river level, after which it passes through a .36-inch Venturi meter to and through 11 vertical perforated pipes, which afford some aeration as they discharge the water into a set- tling reservoir. The reservoir is 333 X 600 feet in plan, has a water depth of nine feet, and a capacity of 14.600,000 gallons. Eleven outlets discharge the water through proper pipes upon eight filter beds having a combined area of 5.6 acres, which, at a nominal rate of 3,000,000 gal- lons an acre, have a total- daily capacity of 14,- 700,000 gallons. From the filters the water passes to a small, equalizing clear-water reser- voir, having a capacity of 600,000 gallons. From this reservoir it flows through a steel conduit to a second pumping station, for delivery to the reservoirs and consumers. The filter beds and clear-water reservoirs are covered. The cost of this whole plant was: .$86,638 for the steel conduit to connect with the old pumping sta- tion; settling reservoir, about $60,000; clear- water reservoir, .$9,000: filter beds. $2.")5,000. Thus the filters alone cost about $17,000 per 1,000,000 gallons of rated capacity. From .Iiily 26, 1900, when the Albany purifica- tion works were put in use, to December 29, 1901,