Page images

at Barn Elms of mechanical filters of 7.5 million United States gallons daily capacity. They treat Thames River water after storage in a reservoir holding about 3 weeks' supply and before it goes to the slow sand filters. These filter units resemble American mechanical filters, and operate at about 160 million U. S. gallons per acre daily. The effluent looks quite clear, although it is said that bacterial removal is little over 50 per cent. The uncovered final sand filters sometimes have very noticeable algae growths in the water over them. Operating at about 2.5 million U. S. gallons per acre daily, the sand filters frequently run several hundred days without cleaning. Although algae growths seem to be the clogging element, a factor to be considered is whether there is enough organic matter in the prefilter effluent to ripen fully the final filters.

Such prefilters, used without coagulant on water stored for three weeks or so, are sufficiently well thought of so that the Board authorized in 1923 the installation of two additional prefilter plants above Hampton, of 21.6 and 38.4 million U. S. gallons capacity respectively, and generally similar to those at Barn Elms.

American experiences

Preliminary filters or scrubbers to aid in the preparatory treatment of water have been adopted at a number of places since 1900, to remove microscopic growths or to lessen turbidity where the use of coagulants was thought undesirable or inexpedient. Some prefilters were built to lessen total cost of the final filtered water, through savings effected in sedimentation basins or final filters.

Experiences have ordinarily shown no very pronounced advantages for prefilters as compared with other preparatory treatment. But prefilters have attained more prominence recently at several places where they are needed to reduce polluting matter so as not to over-tax the final filters.

Philadelphia experiences In 1902 extensive tests were made of various types of scrubbers to remove impurities prior to filtration, principally to remove turbidity. These tests resulted in installing at the Belmont plant so-called scrubbers of the Maignen type consisting of layers of sponge and coke. Coke was found more serviceable, and sponge has now been eliminated.

At the Torresdale plant, on the Delaware, mechanical prefilters were built and are still operated. Coagulants were not used until recently. Filtering material averages about 0.80 mm. effective size. The rate of filtration is nominally 80 million gallons per acre daily. At the Queen Lane plant, receiving Schuylkill River water passed through a sizeable reservoir, mechanical prefilters were built to remove turbidity and other impurities prior to slow sand filtration at a nominal rate of 5 million gallons per acre daily. Most of these prefilters were recently converted into mechanical filters for single filtration.

In recent years both rivers at the present intakes have become too much polluted for satisfactory treatment by single filtration.

The first filtration provides an effluent nearly equal in quality to unfiltered upland streams. Second filtration completes the task and renders the water safe, while chlorination gives an added factor of safety. This is the present method of protection at Torresdale and one which can be continued as long as pollution is not too great and industrial wastes do not interfere. The Schuylkill River is double filtered only in part and the opportunities for it are not as satisfactory as at Torresdale. (See Jour. Amer. Water Works Assn., 1923, vol. 10, p. 351.)

[ocr errors]
[ocr errors]

Other experiences

At Wilmington, Del., scrubbers preliminary to treating Brandywine Creek water in an unbaffled reservoir and in slow sand filters were abandoned as incapable of sufficiently removing turbidity during freshets. At New Castle, Pa., an arrangement of the mechanical filters in two groups so that the effluent from one group might be coagulated again and passed through the other group at times of unusually heavy pollution was abandoned some years ago because results did not warrant the expense and trouble involved.

Montreal experiences

This plant was installed with both prefilters and slow sand filters to be used without coaguiant or preliminary sedimentation. The purpose was to protect the slow sand filters from plankton growths which might occur in the Bay of St. Louis a new miles above the intake in the St. Lawrence. Such growths have never occurred and turbidities have not been sufficient to justify double filtration. The normal rate per acre daily in million Imperial gallons is 115 for the prefilters and 6.5 for the final filters. (See Jour. Amer. Water Works Assn., 1920, vol. 7, p. 898.)

Albany and Poughkeepsie experiences The well known slow sand filter plants at these two places were installed to treat Hudson River water which at these intakes contained substantial pollution as well as turbidity and vegetable stain. As the loading increased, prefilters of the mechanical type were installed to remove suspended matter. Recently these prefilters have been so equipped as to make them mechanical filters for treatment preparatory to slow sand filtration. Complete double filtration has no doubt proved well worth while, notwithstanding that the quality of the Hudson water has become such that steps are being taken at Albany to secure an unpolluted upland supply. (See Jour. Amer. Water. Works Assn., 1920, vol. 10, p. 97, and Jour. N. E. Water Works Assn., 1923, vol. 27, p. 238.)

Summary Enough has been said to outline the trend of progress in keeping the load on ordinary filters within reasonable limits. Double filtration has undoubtedly proved helpful generally, though not always. The individuality of local conditions, particularly of existing filter plants, makes double filtration a special problem such that no general rules are applicable.



These are four principal reasons why it is advantageous to provide covers for reservoirs used for storing ground waters and filtered waters: (1) to keep out the light and thus prevent the growth of algae; (2) to keep out atmospheric dust, air-blown spores and seeds; (3) to protect the supply against human tampering; and (4) to prevent growths of pipe-moss in the distribution system. These may be combined in the statement that the primary object of reservoir covers is to preserve the cleanliness of the water. But there are some other reasons. Waters stored in a covered reservoir are more uniform in temperature than those stored in open reservoirs. In the winter little or no ice forms in them; in the summer the waters are not warmed by the sun. Thus in the winter there is less danger that service pipes, hydrants, and plumbing pipes may freeze, and in the summer the cooler water is more palatable.

Experience has shown that, if ground waters are stored in open reservoirs, tanks or standpipes, microscopic organisms, especially diatoms, are likely to develop, imparting to the water their characteristic odors and making the water turbid. In order that these plant growths may occur it is necessary in the first place that the waters be seeded. Open reservoirs make seeding possible, because the spores of the organisms may be carried by the wind or on the feathers of aquatic birds. When, as in the case of the old water supply of Brooklyn, N. Y., surface waters are mixed with ground water, seeding is inevitable, because surface waters almost always contain, algae, protozoa, and other microscopic organisms. Algae contain chlorophyll, a substance which when exposed to sunlight has the power of converting carbonic acid into the sugars and starches of the plant cells. Exposure of water to the light provides the necessary energy and the fact that ground waters are clear and almost free from colloidal matter enables the sun's rays to penetrate further into the water than is the case with colored or turbid surface waters, although even in clear water the intensity of the sun's rays decreases rapidly below the surface. Ground waters usually contain a sufficient supply of dissolved carbonic acid as well as iron, silica, nitrates, and other mineral substances which diatoms need.

Filtered waters exposed to the light in reservoirs are also susceptible to algae growths, but not quite to the same extent as ground waters, perhaps because colloidal substances are not entirely absent from filtered waters and light penetration is less. Nevertheless, it is the best practice to provide covers for filtered water reservoirs.

If algae, protozoa, and other microscopic organisms develop in an uncovered reservoir and enter the distribution system, they serve as food for fresh-water sponge and for several species of bryozoa which dwell attached to the inner walls of pipes and which are commonly known as "pipe moss." Such growths obstruct the flow of water and increase friction. They also break off and clog meters and water faucets. Providing a cover for a reservoir thus helps to keep the distribution pipes clean.

Uncovered reservoirs in cities are exposed to atmospheric dust, leaves, pollen, bacteria, and other wind-blown substances. In the course of time, these make a considerable deposit on the bottom, foul the water, and promote growths of organisms. B. coli are often increased by exposure to dust and the droppings of birds. In the interest of good bacterial quality, covers are therefore desirable.

Finally, reservoir covers are needed to protect the water from dirt and various substances carried in by animals or thrown in by human beings. Having gone to the expense of purifying water, it is good psychology to keep the filtered water under cover and its purity thus preserved.

In Europe filtered-water reservoirs are all covered, practically without exception. Out of 80 reservoirs for the storage of filtered water of which a record is available, 30 were covered at the time they were built new. Of the remaining 50, four were covered after their original construction, on account of troubles with dust and odors due to the growth of microscopic life; two which gave trouble had been abandoned; and only 10 of the 50 show a record of such small growths of vegetable matter that they can be said to have caused no material difficulty. Of the remaining 34 reservoirs now storing filtered water without covers, the great majority show difficulties from time to time, although they vary in point of frequency, depending upon the size of the basins and the period of time that the filtered water remains in them.

« PreviousContinue »