gelatinous films are formed about the grains of material. As in contact filters, it is these films which play an important part in the purification effected by the filter, due to their power of removing by absorption a certain proportion of the dissolved organic matters contained in the sewage and of acting as oxygen carriers. "Largely due to the predominance of aerobic conditions within sprinkling filters, the deposited organic matter is gradually oxidized to a condition in which it has lost the power in a large measure of adhesion to the particles of filtering material. During periods of rest the oxidation of deposited matters is very rapid and coincident with the efficient drying out which is afforded the filter under favorable weather conditions. Due to these causes, when operation is again resumed, the films of stable suspended matter crack, peel, and are washed from the filters to the temporary detriment of the appearance of the effluent, but to the ultimate benefit of the filter. The removal from the filter in this manner of the deposited suspended matter means a less frequent removal of filtering material on account of clogging, as compared with contact filters, in which no such unloading takes place. As the sewage passes through the filter, constant contact of the sewage with the air is conducive to the highest degree of aerobic bacterial activity. During the active period of operation nitrates and nitrites are constantly being formed in the filters and washed out in the effluent, and serve in this way as a protecting agent against the ultimate putrefaction of the effluent, as do the considerable quantities of dissolved atmospheric oxygen which regularly escape absorption in passing through the filter. During periods of rest nitrification increases in intensity as in the case in contact filters, and the unstable organic matters which have accumulated in the filter are more or less thoroughly oxidized, depending upon the length of the resting period." (From "Report on Sewage Purification at Columbus, Ohio," by George A. Johnson.) In the above quotation reference has been made to periods of rest. In order to bring about the unloading of the filter, or removal of suspended matter, which is one of the characteristics of the sprinkling filter, it has seemed best to those at Columbus to occasionally rest each filter bed from use for a few days, as explained. In experiments conducted at the Massachusetts Institute of Technology, however, it has been found best not to rest the bed; but that the same unloading takes place voluntarily each spring, apparently with the warmer weather which induces a more vigorous bacterial growth and action. Rates as high as two or even three million gallons per acre per day have been satisfactorily treated on sprinkling filters in this country, both at Columbus and at Reading, Pa. If seems probable, however, as in the case of sand filters, that the rate should more correctly be expressed in terms of nitrogen to be oxidized than in mere gallons of fluid. Sewage from a population of 15,000 to 19,000 is treated satisfactorily on one acre of American sprinkling filters 7 feet deep. Concerning the results obtained by a sprinkling filter, the following conclusions were derived from the Massachusetts Institute of Technology experiments: "It removes about one-half the soluble organic matter, yielding an effluent which is somewhat turbid, stable, and well oxygenated. The organic matter present has been so worked over and purified by the bacteria in the filter as to be non-putrescible. Judged by the methylene blue reduction test, 93 per cent of the samples of the effluent are of such stability as to undergo no putrefactive change when kept closed up from the air for four days. Under ordinary conditions of discharge into open water such an effluent would be entirely unobjectionable. "With good distribution the trickling beds show no appreciable tendency to clog. During the greater part of the year solid matter accumulates on the surfaces of the stones throughout the bed, but when this storage reaches a certain point, usually in the early spring, the solids break away and come off in the effluent in a stable condition. In a period covering two years the total amount of solid matter coming off balanced that going on. The filtering material at the end of the experiments was in excellent condition and showed no storage of nitrogen.' It is to be noticed that the removal of suspended matter from . ᅥ sewage by a sprinkling filter is largely nominal, as there is no permanent storing of this and very little if any liquefaction; but the material removed, after being modified to a non-putrefactive form, is carried away with the effluent in flakes or patches which can either pass out with the effluent or be intercepted by a settling basin. Owing to this feature of the sprinkling filter it is necessary that the underdrainage be free and open in order that these large particles may find ready exit. In recent large FIG. 74.-SPRINKLING FILTER DRAINAGE SYSTEM. Distributor main in background. Distributor branches rest on the concrete posts shown. Broken stone is filled in up to level of tops of posts. plants, the drainage system has been so designed that it may be flushed out by a stream from a hose without removing the filtering material. In several plants, arrangement is made for intercepting the suspended matter which leaves the filter in a settling-tank before discharging the effluent into a stream. Owing to the large particles a comparatively small tank with rapid flow will serve this purpose, one having a capacity of one hour's flow being used at Columbus. The matter here collected is not readily putres 1 cible; but it is still organic, and if allowed to remain too long in the bottom of the tank it will begin to putrefy. It should therefore be removed at frequent intervals and may be used for filling in land. At Columbus advantage is taken of high water in the river to discharge this sediment directly into the stream at such periods. The amount of this deposit was found at the Massachusetts Institute of Technology Experimental Station to vary from 1.5 to 5.7 cubic yards per million gallons. One of the objectionable features of the sprinkling filter is the rapid lowering of temperature of the sewage when sprayed through cold air. The effect of winter weather is therefore greater than in the case of other filters. The Columbus filters, for instance, were started in the winter, but it was not until June or July of the following year that normal oxidation became established; whereas four to six weeks should be sufficient time for the attainment of this. In spite of this, however, no greater difficulty has been experienced in operating stationary sprinkler filters through the coldest winter weather than is found with other filters; probably because the continuous operation prevents a thorough chilling of the filter surface, which chilling is possible during the rest periods of intermittent sand and contact filters. In the construction of a sprinkling filter, the filtering material may be placed in a tank or pit or may be erected in a pile upon the surface of the ground, the material being retained within walls of concrete or masonry of open dry stone-work. The last has the advantage of affording additional opportunity for aëration and is ordinarily cheaper than the other. As the sewage simply percolates downward, there is no necessity for watertight walls. A size of particles between inch and 3 inches is believed to give the best results with sewage which has previously been well clarified. The finer the material the greater the surface area, but also the greater the tendency to clog with sediment. Some European authorities believe that uniformity of size throughout a bed is of more importance than the actual size of grain, as this gives the maximum interstitial space for the circulation of air. |