Filtration of Water and Sewage.*
IT is six years ago since the State Board of Health of Massachusetts began experiments to supply the deficiency that was found to exist in the knowledge of the engineering profession upon the subject of filtration of water and sewage. The ends in view were the prevention of pollution of streams, the purification of domestic water supplies and the general preservation and improvement of the public health. Excellent effluents had been obtained in the experiments on irrigation by the British Association for the Advancement of Science and by Marie Davy at Baris, and on the irrigation fields in Germany, which might now be attributed to filtration; but their purity was then regarded as depending to a great degree upon the action of growing crops. In these results Marie Davy had shown a remarkable purification from bacteria. There were in 1877 various filter beds in use in England and on the Continent, mostly in connection with irrigation fields, or used, themselves, for growing crops. They were receiving from 36,000 to 90,000 gallons of sewage per acre per day, and were giving results satisfactory to the operators ; but the few published results of analyses of the effluents were less satisfactory than those obtained in the laboratories. There was not sufficient knowledge upon the subject in the profession to enable a well-educated engineer to decide whether any selected area would purify sewage applied to it, and, if it would purify, the amount to be applied and the proper intervals for application, and whether and under what conditions disease producing germs could be removed. To supply knowledge upon these points and others pertaining to the purity of water supplies, the experiment station of the State Board of Health of Massachusetts was established at Lawrence in 1887. Some of the more important general results of the six years of careful investigation will now be presented.
Up to the present time the experiments of the State Board of Health indicate that the sand of even grain that presents all of the conditions most favorable for the very complete purification of sewage has a diameter of grain of about two-tenths of a millimetre, or about eight-thousandths of an inch. The more even in size the grains are the better, but as we rarely find a sand in which some of the grains are not many times the size of others, and as the conditions for nitrification depend more essentially upon the finer grains of the mass, we have found it more intelligible to classify sands by the finer 10 per cent, of their grains, and in thus classifying sands we should say that sand presents the conditions most favorable for very complete purification of sewage, the finer 10 per cent, of whose grains have diameters equal to and less, but not much less, than 0.2 millimetre. In soils the organic matter, which is much finer than this, makes up a considerable part of the finer 10 per cent, of the particles, and this, sticking to the other particles, decreases the air and water spaces to a greater extent than would appear from the mechanical analysis ; hence, a filter bed made in the effluents was about 0.2 per cent., while from the filters with sands finer than the standard the number remaining in the effluents was about 0.001 per cent, of the number applied in the sewage. In fact we have the strongest reason to conclude that with these sands no bacteria ordinarily lived through the passage, but that under peculiar circumstances a few survived and these lived in the underdrains, and occasionally from these a few units came in the effluent to represent the hundreds of thousands applied. Sometimes the few found in the effluent were a pure culture of a single species of bacteria. It has been found that most of the bacteria that get through the filters are of kinds that live and are propagated in wet sands and underdrains. If typhoid fever germs were applied with the sewage their lateness of development upon gelatine plates allows other bacteria to grow in such numbers or to liquefy the plates to such an extent that the typhoid fever germs, if present, can hardly be distinguished ; hence it becomes necessary to select a foreign species not found in our sewage, having habits of life similar to typhoid fever germs, that can be readily distinguished from other bacteria ; such is Bacillus prodigiosus, which has been cultivated in large numbers and applied by the millions to the several filters. With sands whose finer 10 per cent, had
•An abstract of a paper read before the International Engirteerin* Congress at Chicago by Mr. Hiram I’. Mills. grains as large as 0.5 millimetre, we have succeeded in passing through a few of the Bacillus prodigiosus, and conclude that typhoid fever germs would pass through such sands in very much reduced numbers. With sands of 0.3 millimetre and less we have not succeeded in finding in the effluents any of the large numbers of Bacillus prodigiosus applied with the sewage. These results, together with the fact that typhoid fever germs continually decrease in numbers, not only in the effluents from these filters, but even, although less rapidly, in sewage-polluted waters which have not been filtered, dying out in the latter in from one to three weeks, give us confidence in concluding that filters made with these finer sands, so long as they are kept in a condition to very completely nitrify the sewage applied, will give an effluent that may with safety be turned into a drinking-water stream. Study of the working of intermitent filters shows them to be not mechanical strainers, but delicate organisms gradually growing fitted to perform the work required of them, needing time to adapt themselves to any marked change in the amount or quality of the organic matter applied to them, but regularly and for many years, performing an astonishing amount of chemical and biological purification, if the essential conditions for their action are maintained.
FILTERATION OK Water.—The principles of the purification of sewage by intermittent filteration are applicable to the purification of polluted water supplies, with this important difference. Water supplies have not sufficienfcorganic matter to combine readily with all of the oxygen the water absorbs from the air, consequently contain more or less free oxygen; hence, they do not require so frequent renewal of air in the sand, and much larger quantities can be passed through the filter and there find a sufficient additional amount of air to combine with that nitrogenous matter they contain, to nitrify it through longer periods of continued use than the still more polluted water which we call sewage. Wc may in a very general way distinguish between the two by saving sewage contains fifty times as much nitrogenous organic matter as a polluted water supply. The question may then arise, can we satisfactorily purify by (iteration fifty times as much water as sewage upon the same area? It may prove to be practical in the future, but at present we will give attention to results that have been obtained when filtering about half this amount. One of the first steps in considering what sand to use in a filter is to determine how much water will flow through it with the available head. Mr. Allen Ilazen, the chemist in charge of the experiment station, found the quantity that would flow through the sands ordinarily used for filteration to be, at a constant temperature directly proportional to the acting head, and inversely proportional to the depth or distance through the sand. 1 le also found it directly proportional to the square of the maximum diameter of the the finer 10 per cent, of the sand grains, ( ailing the maximum diameter in millimetres d, the head in feet h.the distance flowing through the sand 1, and the quantity in 1,000,000 gallons per acre per day he found the following formula to satisfactorily express the results of his experiments with water at about 50 deg. Fahr.:—(J=8oo hd2 -j-l. The sand which we have found capable of very completely purifying the largest quantity of sewage, having d equal to 0.2 millimetre, would, when 1 foot deep and with 1 foot acting head, allow 32,000,000 gallons per acre to flow through it in 24 hours, and the same quantity would flow through any other depth if the acting head were equal to the depth. There are practical considerations that make if desirable to depend upon a less head than the dist*ice of the water passes through the sand. If there be one foot acting head for a passage of five feet through the sand, the maximum quantity that would flow through this sand in 24 hours would be 6,400,000 gallons per acre, but the whole 24 hours would not be available if the water be applied intermittently. With the Merrimac river water used in the experiments, it was found that filling the sand with air once a day was sufficient, in which case we can count upon a steady flow through the sand for two-thirds of the 24 hours, thus passing 4.300,000 gallons per acre per day when the sand is clean ; but this quantity would be still further reduced for the water has some organic matter and some mineral matter in suspension at all times, and during the spring freshets the amount of mineral matter is very large. A1 of this suspended matter tends to clog the interstices of the upper layer of the filter, and to allow less water to pass ; hence the sand of the filter must be coarse enough to allow the desired quantity to flow through when the upper layer is obstructed, and not make the intervals between scrapings of the surface to remove this deposit loo inconveniently short. With the water used, the sand we are considering allowed 60,000,000 gallons of water to pass through one acre between the times when it was necessary to scrape the surface and remove about one-eighth inch of sand and sediment and replace with clean sand. When filtering 2,000,000 gallons per acre daily, the deposit hail to be removed once a month in the six months from May to October, and more frequently during the spring and fall freshets. With coarser and finer sand the amount passed between scrapings was nearly in projwrtion to the size of the sand grains. Sands, the larger grains of the finer 10 per cent, of which are from 0.1 to 0.3 millimetre in diameter, and whose capacity to pass water through a depth equal to the acting head, ranges from 8,000,000 to 72.ooo.1xx1 gallons per acre in 24 hours, have given excellent results when filtering 2,000,000 gallons per acre per day, and the mean of these sizes has also satisfactorily purified 3,000,000 gallons per acre per day. With still coarser sands up to 0.5 millimetre in diameter, good results have been obtained when filtering 1,000,000 gallons per day; but the results were variable, and the largest size mentioned has not given assurance that sewage polluted water can be constantly purified from its disease germs. As before indicated there is in drinking water nearly enough absorbed oxygen to combine with the nitrogenous organic matter to burn it up or to nitrify it, v^hen it is brought in contact with nitrifying bacteria under conditions favorable for their action. Filling the srrtrt^vith air-once a day furnishes the additional quantity of oxygen necessary for very complete nitrification.